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 NM/01 Workround for PCRE version dependancy in version reporting
12 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
13 This fixes portability to compilers other than gcc, notably
14 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
16 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
17 makefiles for portability to HP-UX and POSIX correctness.
19 PP/01 Permit LOOKUP_foo enabling on the make command-line.
20 Also via indented variable definition in the Makefile.
21 (Debugging by Oliver Heesakkers).
23 PP/02 Restore caching of spamd results with expanded spamd_address.
24 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
26 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
27 Improves build reliability. Fix from: Frank Elsner
29 NM/02 Fix wide character breakage in the rfc2047 coding
30 Fixes bug 1064. Patch from Andrey N. Oktyabrski
32 NM/03 Allow underscore in dnslist lookups
33 Fixes bug 1026. Patch from Graeme Fowler
35 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
36 Code patches from Adam Ciarcinski of NetBSD.
38 NM/04 Fixed exiqgrep to cope with mailq missing size issue
41 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
42 is logged, to avoid truncation. Patch from John Horne.
44 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
45 Patch from Jakob Hirsch.
47 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
48 SQL string expansion failure details.
49 Patch from Andrey Oktyabrski.
51 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
52 Patch from Simon Arlott.
54 PP/09 New lookups functionality failed to compile on old gcc which rejects
55 extern declarations in function scope.
56 Patch from Oliver Fleischmann
58 PP/10 Use sig_atomic_t for flags set from signal handlers.
59 Check getgroups() return and improve debugging.
60 Fixed developed for diagnosis in bug 927 (which turned out to be
63 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
64 Patch from Mark Zealey.
66 PP/12 Bugzilla 1056: Improved spamd server selection.
67 Patch from Mark Zealey.
69 PP/13 Bugzilla 1086: Deal with maildir quota file races.
70 Based on patch from Heiko Schlittermann.
76 TF/01 Failure to get a lock on a hints database can have serious
77 consequences so log it to the panic log.
79 TF/02 Log LMTP confirmation messages in the same way as SMTP,
80 controlled using the smtp_confirmation log selector.
82 TF/03 Include the error message when we fail to unlink a spool file.
84 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
85 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
86 for maintaining out-of-tree patches for some time.
88 PP/01 Bugzilla 139: Documentation and portability issues.
89 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
90 Handle per-OS dynamic-module compilation flags.
92 PP/02 Let /dev/null have normal permissions.
93 The 4.73 fixes were a little too stringent and complained about the
94 permissions on /dev/null. Exempt it from some checks.
95 Reported by Andreas M. Kirchwitz.
97 PP/03 Report version information for many libraries, including
98 Exim version information for dynamically loaded libraries. Created
99 version.h, now support a version extension string for distributors
100 who patch heavily. Dynamic module ABI change.
102 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
103 privilege escalation vulnerability whereby the Exim run-time user
104 can cause root to append content of the attacker's choosing to
107 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
110 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
111 If dropping privileges for untrusted macros, we disabled normal logging
112 on the basis that it would fail; for the Exim run-time user, this is not
113 the case, and it resulted in successful deliveries going unlogged.
114 Fixed. Reported by Andreas Metzler.
120 PP/01 Date: & Message-Id: revert to normally being appended to a message,
121 only prepend for the Resent-* case. Fixes regression introduced in
122 Exim 4.70 by NM/22 for Bugzilla 607.
124 PP/02 Include check_rfc2047_length in configure.default because we're seeing
125 increasing numbers of administrators be bitten by this.
127 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
129 PP/03 Bugzilla 994: added openssl_options main configuration option.
131 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
133 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
135 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
137 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
138 without a peer certificate, leading to a segfault because of an
139 assumption that peers always have certificates. Be a little more
140 paranoid. Problem reported by Martin Tscholak.
142 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
143 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
144 NB: ClamAV planning to remove STREAM in "middle of 2010".
145 CL also introduces -bmalware, various -d+acl logging additions and
146 more caution in buffer sizes.
148 PP/09 Implemented reverse_ip expansion operator.
150 PP/10 Bugzilla 937: provide a "debug" ACL control.
152 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
154 PP/12 Bugzilla 973: Implement --version.
156 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
158 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
160 PP/15 Bugzilla 816: support multiple condition rules on Routers.
162 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
163 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
164 ignore trailing whitespace.
166 JJ/02 prevent non-panic DKIM error from being sent to paniclog
168 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
171 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
172 Notification from Dr Andrew Aitchison.
174 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
175 ExtendedDetectionInfo response format.
176 Notification from John Horne.
178 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
181 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
182 XSL and documented dependency on system catalogs, with examples of how
185 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
188 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
189 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
190 configuration file which is writeable by the Exim user or group.
192 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
193 of configuration files to cover files specified with the -C option if
194 they are going to be used with root privileges, not just the default
197 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
198 option (effectively making it always true).
200 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
201 files to be used while preserving root privileges.
203 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
204 that rogue child processes cannot use them.
206 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
207 run-time user, instead of root.
209 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
210 Exim run-time user without dropping privileges.
212 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
213 result string, instead of calling string_vformat() twice with the same
216 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
217 for other users. Others should always drop root privileges if they use
218 -C on the command line, even for a whitelisted configure file.
220 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
222 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
228 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
229 $data_path, and $header_path variables; fixed documentation bugs and
232 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
233 exipick to access non-standard spools, including the "frozen" queue
236 NM/01 Bugzilla 965: Support mysql stored procedures.
237 Patch from Alain Williams
239 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
241 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
242 Patch from Andreas Metzler
244 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
245 Patch from Kirill Miazine
247 NM/05 Bugzilla 671: Added umask to procmail example.
249 JJ/03 installed exipick 20100323.0, fixing doc bug
251 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
252 directory. Notification and patch from Dan Rosenberg.
254 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
256 TK/02 Improve log output when DKIM signing operation fails.
258 MH/01 Treat the transport option dkim_domain as a colon separated
259 list, not as a single string, and sign the message with each element,
260 omitting multiple occurences of the same signer.
262 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
263 Bugzilla 985, 986. Patch by Simon Arlott
265 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
266 Patch by Simon Arlott
268 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
269 MBX locking. Notification from Dan Rosenberg.
275 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
277 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
279 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
281 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
283 NM/04 Bugzilla 915: Flag broken perl installation during build.
289 TK/01 Added patch by Johannes Berg that expands the main option
290 "spamd_address" if it starts with a dollar sign.
292 TK/02 Write list of recipients to X-Envelope-Sender header when building
293 the mbox-format spool file for content scanning (suggested by Jakob
296 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
297 (http://www.dcc-servers.net/) support via dccifd. Activated by
298 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
300 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
301 by Mark Daniel Reidel <mr@df.eu>.
303 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
304 When building exim an external PCRE library is now needed -
305 PCRE is a system library on the majority of modern systems.
306 See entry on PCRE_LIBS in EDITME file.
308 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
309 conversation. Added nologin parameter to request.
310 Patch contributed by Kirill Miazine.
312 TF/01 Do not log submission mode rewrites if they do not change the address.
314 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
316 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
317 log files in place. Contributed by Roberto Lima.
319 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
321 TF/03 Bugzilla 615: When checking the local_parts router precondition
322 after a local_part_suffix or local_part_prefix option, Exim now
323 does not use the address's named list lookup cache, since this
324 contains cached lookups for the whole local part.
326 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
327 Robert Millan. Documentation is in experimental-spec.txt.
329 TF/04 Bugzilla 668: Fix parallel build (make -j).
331 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
333 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
334 Patch provided by Jan Srzednicki.
336 TF/05 Leading white space used to be stripped from $spam_report which
337 wrecked the formatting. Now it is preserved.
339 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
340 that they are available at delivery time.
342 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
344 TF/08 TLS error reporting now respects the incoming_interface and
345 incoming_port log selectors.
347 TF/09 Produce a more useful error message if an SMTP transport's hosts
348 setting expands to an empty string.
350 NM/06 Bugzilla 744: EXPN did not work under TLS.
351 Patch provided by Phil Pennock.
353 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
354 Patch provided by Richard Godbee.
356 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
357 acl_smtp_notquit, added index entry.
359 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
360 Patch provided by Eugene Bujak.
362 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
363 accept(). Patch provided by Maxim Dounin.
365 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
366 Patch provided by Phil Pennock.
368 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
370 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
371 Patch provided by Brad "anomie" Jorsch.
373 NM/14 Bugzilla 622: Added timeout setting to transport filter.
374 Patch provided by Dean Brooks.
376 TK/05 Add native DKIM support (does not depend on external libraries).
378 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
379 Patch provided by Graeme Fowler.
381 NM/16 Bugzilla 851: Documentation example syntax fix.
383 NM/17 Changed NOTICE file to remove references to embedded PCRE.
385 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
388 NM/19 Bugzilla 745: TLS version reporting.
389 Patch provided by Phil Pennock.
391 NM/20 Bugzilla 167: bool: condition support.
392 Patch provided by Phil Pennock.
394 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
395 clients. Patch provided by Phil Pennock.
397 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
398 Patch provided by Brad "anomie" Jorsch.
400 NM/23 Bugzilla 687: Fix misparses in eximstats.
401 Patch provided by Heiko Schlittermann.
403 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
404 Patch provided by Heiko Schlittermann.
406 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
407 plus update to original patch.
409 NM/26 Bugzilla 799: Documentation correction for ratelimit.
411 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
412 Patch provided by David Brownlee.
414 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
416 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
418 NM/30 Bugzilla 888: TLS documentation bugfixes.
420 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
422 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
423 Unlike the original bugzilla I have changed all shell scripts in src tree.
425 NM/33 Bugzilla 898: Transport filter timeout fix.
426 Patch by Todd Rinaldo.
428 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
429 Patch by Serge Demonchaux.
431 NM/35 Bugzilla 39: Base64 decode bug fixes.
432 Patch by Jakob Hirsch.
434 NM/36 Bugzilla 909: Correct connect() call in dcc code.
436 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
438 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
440 NM/39 Bugzilla 911: Fixed MakeLinks build script.
446 TK/01 Add preliminary DKIM support. Currently requires a forked version of
447 ALT-N's libdkim that I have put here:
448 http://duncanthrax.net/exim-experimental/
450 Note to Michael Haardt: I had to rename some vars in sieve.c. They
451 were called 'true' and it seems that C99 defines that as a reserved
452 keyword to be used with 'bool' variable types. That means you could
453 not include C99-style headers which use bools without triggering
454 build errors in sieve.c.
456 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
457 as mailq or other aliases. Changed the --help handling significantly
458 to do whats expected. exim_usage() emits usage/help information.
460 SC/01 Added the -bylocaldomain option to eximstats.
462 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
464 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
466 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
472 PH/01 Another patch from the Sieve maintainer.
474 PH/02 When an IPv6 address is converted to a string for single-key lookup
475 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
476 dots are used instead of colons so that keys in lsearch files need not
477 contain colons. This was done some time before quoting was made available
478 in lsearch files. However, iplsearch files do require colons in IPv6 keys
479 (notated using the quote facility) so as to distinguish them from IPv4
480 keys. This meant that lookups for IP addresses in host lists did not work
481 for iplsearch lookups.
483 This has been fixed by arranging for IPv6 addresses to be expressed with
484 colons if the lookup type is iplsearch. This is not incompatible, because
485 previously such lookups could never work.
487 The situation is now rather anomolous, since one *can* have colons in
488 ordinary lsearch keys. However, making the change in all cases is
489 incompatible and would probably break a number of configurations.
491 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
494 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
495 conversion specification without a maximum field width, thereby enabling
496 a rogue spamd server to cause a buffer overflow. While nobody in their
497 right mind would setup Exim to query an untrusted spamd server, an
498 attacker that gains access to a server running spamd could potentially
499 exploit this vulnerability to run arbitrary code as the Exim user.
501 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
502 $primary_hostname instead of what libspf2 thinks the hosts name is.
504 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
505 a directory entry by the name of the lookup key. Previously, if a
506 symlink pointed to a non-existing file or a file in a directory that
507 Exim lacked permissions to read, a lookup for a key matching that
508 symlink would fail. Now it is enough that a matching directory entry
509 exists, symlink or not. (Bugzilla 503.)
511 PH/03 The body_linecount and body_zerocount variables are now exported in the
514 PH/04 Added the $dnslist_matched variable.
516 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
517 This means they are set thereafter only if the connection becomes
520 PH/06 Added the client_condition to authenticators so that some can be skipped
521 by clients under certain conditions.
523 PH/07 The error message for a badly-placed control=no_multiline_responses left
524 "_responses" off the end of the name.
526 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
528 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
529 (without spaces) instead of just copying the configuration text.
531 PH/10 Added the /noupdate option to the ratelimit ACL condition.
533 PH/11 Added $max_received_linelength.
535 PH/12 Added +ignore_defer and +include_defer to host lists.
537 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
538 way in which PCRE > 7.0 is built.
540 PH/14 Implemented queue_only_load_latch.
542 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
543 MAIL command. The effect was to mangle the value on 64-bit systems.
545 PH/16 Another patch from the Sieve maintainer.
547 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
549 PH/18 If a system quota error occurred while trying to create the file for
550 a maildir delivery, the message "Mailbox is full" was not appended to the
551 bounce if the delivery eventually timed out. Change 4.67/27 below applied
552 only to a quota excession during the actual writing of the file.
554 PH/19 It seems that peer DN values may contain newlines (and other non-printing
555 characters?) which causes problems in log lines. The DN values are now
556 passed through string_printing() before being added to log lines.
558 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
559 and InterBase are left for another time.)
561 PH/21 Added message_body_newlines option.
563 PH/22 Guard against possible overflow in moan_check_errorcopy().
565 PH/23 POSIX allows open() to be a macro; guard against that.
567 PH/24 If the recipient of an error message contained an @ in the local part
568 (suitably quoted, of course), incorrect values were put in $domain and
569 $local_part during the evaluation of errors_copy.
575 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
576 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
579 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
580 issue a MAIL command.
582 PH/02 In an ACL statement such as
584 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
586 if a client was not listed at all, or was listed with a value other than
587 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
588 the condition was not true (as it should be), so access was not denied.
589 The bug was that the ! inversion was incorrectly passed on to the second
590 item. This has been fixed.
592 PH/03 Added additional dnslists conditions == and =& which are different from
593 = and & when the dns lookup returns more than one IP address.
595 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
596 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
598 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
599 FSYNC, which compiles an option called disable_fsync that allows for
600 bypassing fsync(). The documentation is heavily laced with warnings.
602 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
604 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
605 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
606 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
607 including adding "make clean"; (3) Added -fPIC when compiling the test
608 dynamically loaded module, to get rid of a warning.
610 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
611 message fails, move_frozen_messages = true and ignore_bounce_errors_after
612 = 0s. The bug is otherwise harmless.
614 PH/07 There was a bug in the dovecot authenticator such that the value of
615 $auth1 could be overwritten, and so not correctly preserved, after a
616 successful authentication. This usually meant that the value preserved by
617 the server_setid option was incorrect.
619 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
621 PH/09 Installed PCRE release 7.0.
623 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
624 run for batched SMTP input. It is now run at the start of every message
625 in the batch. While fixing this I discovered that the process information
626 (output by running exiwhat) was not always getting set for -bs and -bS
627 input. This is fixed, and it now also says "batched" for BSMTP.
629 PH/11 Added control=no_pipelining.
631 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
632 patch, slightly modified), and move the expansion of helo_data till after
633 the connection is made in the smtp transport (so it can use these
636 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
638 PH/14 Added log_selector = +pid.
640 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
642 PH/16 Add ${if forany and ${if forall.
644 PH/17 Added dsn_from option to vary the From: line in DSNs.
646 PH/18 Flush SMTP output before performing a callout, unless control =
647 no_callout_flush is set.
649 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
650 was true (the default) a successful delivery failed to delete the retry
651 item, thus causing premature timeout of the address. The bug is now
654 PH/20 Added hosts_avoid_pipelining to the smtp transport.
656 PH/21 Long custom messages for fakedefer and fakereject are now split up
657 into multiline reponses in the same way that messages for "deny" and
658 other ACL rejections are.
660 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
661 with slight modification.
663 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
664 draft, changing the syntax and factoring some duplicate code".
666 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
667 for deliveries of the second and subsequent messages over the same SMTP
670 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
671 ${reduce, with only minor "tidies".
673 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
675 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
676 expansion side effects.
678 PH/27 When a message times out after an over-quota error from an Exim-imposed
679 quota, the bounce message says "mailbox is full". This message was not
680 being given when it was a system quota that was exceeded. It now should
683 MH/03 Made $recipients available in local_scan(). local_scan() already has
684 better access to the recipient list through recipients_list[], but
685 $recipients can be useful in postmaster-provided expansion strings.
687 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
688 in the case of a MAIL command with additional options following the
689 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
690 were accidentally chopped off.
692 PH/29 SMTP synchronization checks are implemented when a command is read -
693 there is a check that no more input is waiting when there shouldn't be
694 any. However, for some commands, a delay in an ACL can mean that it is
695 some time before the response is written. In this time, more input might
696 arrive, invalidly. So now there are extra checks after an ACL has run for
697 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
698 pipelining has not been advertised.
700 PH/30 MH's patch to allow iscntrl() characters to be list separators.
702 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
703 returned in the SMTP response when smtp_return_error_details was false.
706 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
707 instead of the C I/O that was originally supplied, because problems were
710 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
711 Exim which did not show up earlier: it was assuming that a call to
712 SSL_CTX_set_info_callback() might give an error value. In fact, there is
713 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
714 was a macro that became an assignment, so it seemed to work. This has
715 changed to a proper function call with a void return, hence the compile
716 error. Exim's code has been fixed.
718 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
721 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
723 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
725 PH/37 If a message is not accepted after it has had an id assigned (e.g.
726 because it turns out to be too big or there is a timeout) there is no
727 "Completed" line in the log. When some messages of this type were
728 selected by exigrep, they were listed as "not completed". Others were
729 picked up by some special patterns. I have improved the selection
730 criteria to be more general.
732 PH/38 The host_find_failed option in the manualroute router can now be set
733 to "ignore", to completely ignore a host whose IP address cannot be
734 found. If all hosts are ignored, the behaviour is controlled by the new
735 host_all_ignored option.
737 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
738 homing or because of multiple MX records with /mx) generated more than
739 one IP address, and the following item turned out to be the local host,
740 all the secondary addresses of the first item were incorrectly removed
741 from the list, along with the local host and any following hosts (which
742 is what is supposed to happen).
744 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
745 whoever called Exim into the -H file. In the case of the daemon it was
746 behaving confusingly. When first started, it used values for whoever
747 started the daemon, but after a SIGHUP it used the Exim user (because it
748 calls itself on a restart). I have changed the code so that it now always
751 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
752 message are rejected with the same error (e.g. no authentication or bad
753 sender address), and a DATA command is nevertheless sent (as can happen
754 with PIPELINING or a stupid MUA), the error message that was given to the
755 RCPT commands is included in the rejection of the DATA command. This is
756 intended to be helpful for MUAs that show only the final error to their
759 PH/42 Another patch from the Sieve maintainer.
761 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
762 Eximstats - Fixed some broken HTML links and added missing column headers
764 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
765 columns for Rejects, Temp Rejects, Ham, and Spam rows.
767 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
769 PH/43 Yet another patch from the Sieve maintainer.
771 PH/44 I found a way to check for a TCP/IP connection going away before sending
772 the response to the final '.' that terminates a message, but only in the
773 case where the client has not sent further data following the '.'
774 (unfortunately, this is allowed). However, in many cases there won't be
775 any further data because there won't be any more messages to send. A call
776 to select() can be used: if it shows that the input is "ready", there is
777 either input waiting, or the socket has been closed. An attempt to read
778 the next input character can distinguish the two cases. Previously, Exim
779 would have sent an OK response which the client would never have see.
780 This could lead to message repetition. This fix should cure that, at
781 least in a lot of common cases.
783 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
784 advertised in response to EHLO.
790 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
791 fixed by 4.65/MH/01 (is this a record?) are fixed:
793 (i) An empty string was always treated as zero by the numeric comparison
794 operators. This behaviour has been restored.
796 (ii) It is documented that the numeric comparison operators always treat
797 their arguments as decimal numbers. This was broken in that numbers
798 starting with 0 were being interpreted as octal.
800 While fixing these problems I realized that there was another issue that
801 hadn't been noticed. Values of message_size_limit (both the global option
802 and the transport option) were treated as octal if they started with 0.
803 The documentation was vague. These values are now always treated as
804 decimal, and I will make that clear in the documentation.
810 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
811 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
814 MH/01 Don't check that the operands of numeric comparison operators are
815 integers when their expansion is in "skipping" mode (fixes bug
816 introduced by 4.64-PH/07).
818 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
819 child addresses, Exim now panics and dies. Previously, because the count
820 is held in a short int, deliveries were likely to be lost. As such a
821 large number of recipients for a single message is ridiculous
822 (performance will be very, very poor), I have chosen to impose a limit
823 rather than extend the field.
829 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
830 leftover -K file (the existence of which was triggered by #402).
831 While we were at it, introduced process PID as part of the -K
832 filename. This should rule out race conditions when creating
835 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
836 processing considerably. Previous code took too long for large mails,
837 triggering a timeout which in turn triggers #401.
839 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
840 in the DK code in transports.c. sendfile() is not really portable,
841 hence the _LINUX specificness.
843 TF/01 In the add_headers option to the mail command in an Exim filter,
844 there was a bug that Exim would claim a syntax error in any
845 header after the first one which had an odd number of characters
848 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
849 callout verification, Exim cached a "reject" for the entire domain. This
850 is correct for most verifications, but it is not correct for a recipient
851 verification with use_sender or use_postmaster set, because in that case
852 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
853 case of MAIL FROM:<> rejection from other early rejections (e.g.
854 rejection of HELO). When verifying a recipient using a non-null MAIL
855 address, the cache is ignored if it shows MAIL FROM:<> rejection.
856 Whatever the result of the callout, the value of the domain cache is
857 left unchanged (for any other kind of callout, getting as far as trying
858 RCPT means that the domain itself is ok).
860 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
863 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
864 manifest itself as EPIPE rather than ECONNECT. When tidying away a
865 session, the daemon ignores ECONNECT errors and logs others; it now
866 ignores EPIPE as well.
868 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
869 (quoted-printable decoding).
871 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
872 later the small subsequent patch to fix an introduced bug.
874 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
876 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
878 PH/08 An error is now given if message_size_limit is specified negative.
880 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
881 to be given (somewhat) arbitrary names.
883 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
886 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
887 miscellaneous code fixes
889 PH/10 Added the log_reject_target ACL modifier to specify where to log
892 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
893 hostname. This is wrong, because it relates to the incoming message (and
894 probably the interface on which it is arriving) and not to the outgoing
895 callout (which could be using a different interface). This has been
896 changed to use the value of the helo_data option from the smtp transport
897 instead - this is what is used when a message is actually being sent. If
898 there is no remote transport (possible with a router that sets up host
899 addresses), $smtp_active_hostname is used.
901 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
902 tweaks were necessary in order to get it to work (see also 21 below):
903 (a) The code assumed that strncpy() returns a negative number on buffer
904 overflow, which isn't the case. Replaced with Exim's string_format()
906 (b) There were several signed/unsigned issues. I just did the minimum
907 hacking in of casts. There is scope for a larger refactoring.
908 (c) The code used strcasecmp() which is not a standard C function.
909 Replaced with Exim's strcmpic() function.
910 (d) The code set only $1; it now sets $auth1 as well.
911 (e) A simple test gave the error "authentication client didn't specify
912 service in request". It would seem that Dovecot has changed its
913 interface. Fortunately there's a specification; I followed it and
914 changed what the client sends and it appears to be working now.
916 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
919 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
920 address A is aliased to B and C, where B exists and C does not. Without
921 -v the output is "A verified" because verification stops after a
922 successful redirection if more than one address is generated. However,
923 with -v the child addresses are also verified. Exim was outputting "A
924 failed to verify" and then showing the successful verification for C,
925 with its parentage. It now outputs "B failed to verify", showing B's
926 parentage before showing the successful verification of C.
928 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
929 look up a TXT record in a specific list after matching in a combined
932 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
933 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
934 they consult the DNS. I had assumed they would set it the way they
935 wanted; and indeed my experiments on Linux seem to show that in some
936 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
937 To be on the safe side, however, I have now made the interface to
938 host_find_byname() similar to host_find_bydns(), with an argument
939 containing the DNS resolver options. The host_find_byname() function now
940 sets these options at its start, just as host_find_bydns() does. The smtp
941 transport options dns_qualify_single and dns_search_parents are passed to
942 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
943 of host_find_byname() use the default settings of RES_DEFNAMES
944 (qualify_single) but not RES_DNSRCH (search_parents).
946 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
947 spool_read_header() do less string testing, by means of a preliminary
948 switch on the second character of optional "-foo" lines. (This is
949 overdue, caused by the large number of possibilities that now exist.
950 Originally there were few.) While I was there, I also converted the
951 str(n)cmp tests so they don't re-test the leading "-" and the first
952 character, in the hope this might squeeze out yet more improvement.
954 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
955 flag allowing group syntax was set by the header_syntax check but not
956 turned off, possible causing trouble later; (2) The flag was not being
957 set at all for the header_verify test, causing "group"-style headers to
958 be rejected. I have now set it in this case, and also caused header_
959 verify to ignore an empty address taken from a group. While doing this, I
960 came across some other cases where the code for allowing group syntax
961 while scanning a header line wasn't quite right (mostly, not resetting
962 the flag correctly in the right place). These bugs could have caused
963 trouble for malformed header lines. I hope it is now all correct.
965 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
966 with the "reply" argument non-NULL. The code, however (which originally
967 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
968 but it didn't always do it. This confused somebody who was copying the
969 code for some other use. I have removed all the tests.
971 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
972 feature that was used to support insecure browsers during the U.S. crypto
973 embargo. It requires special client support, and Exim is probably the
974 only MTA that supported it -- and would never use it because real RSA is
975 always available. This code has been removed, because it had the bad
976 effect of slowing Exim down by computing (never used) parameters for the
977 RSA_EXPORT functionality.
979 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
980 authenticator to fail if there's a tab character in the incoming data
981 (there should never be unless someone is messing about, as it's supposed
982 to be base64-encoded). Also added, on Timo's advice, the "secured" option
983 if the connection is using TLS or if the remote IP is the same as the
984 local IP, and the "valid-client-cert option" if a client certificate has
987 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
988 authenticators. This can be used for authorization after authentication
989 succeeds. (In the case of plaintext, it servers for both authentication
992 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
993 if any retry times were supplied.
995 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
996 connection, where there is no client IP address to check. In this
997 situation, the verify now always succeeds.
999 PH/25 Applied John Jetmore's -Mset patch.
1001 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1003 PH/27 In a string expansion for a processed (not raw) header when multiple
1004 headers of the same name were present, leading whitespace was being
1005 removed from all of them, but trailing whitespace was being removed only
1006 from the last one. Now trailing whitespace is removed from each header
1007 before concatenation. Completely empty headers in a concatenation (as
1008 before) are ignored.
1010 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1011 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1013 PH/29 [Removed. This was a change that I later backed out, and forgot to
1014 correct the ChangeLog entry (that I had efficiently created) before
1015 committing the later change.]
1017 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1018 address errors (4xx response to RCPT) over the same connection as other
1019 messages routed to the same hosts. Such deliveries are always "forced",
1020 so retry times are not inspected. This resulted in far too many retries
1021 for the affected addresses. The effect occurred only when there were more
1022 hosts than the hosts_max_try setting in the smtp transport when it had
1023 the 4xx errors. Those hosts that it had tried were not added to the list
1024 of hosts for which the message was waiting, so if all were tried, there
1025 was no problem. Two fixes have been applied:
1027 (i) If there are any address or message errors in an SMTP delivery, none
1028 of the hosts (tried or untried) are now added to the list of hosts
1029 for which the message is waiting, so the message should not be a
1030 candidate for sending over the same connection that was used for a
1031 successful delivery of some other message. This seems entirely
1032 reasonable: after all the message is NOT "waiting for some host".
1033 This is so "obvious" that I'm not sure why it wasn't done
1034 previously. Hope I haven't missed anything, but it can't do any
1035 harm, as the worst effect is to miss an optimization.
1037 (ii) If, despite (i), such a delivery is accidentally attempted, the
1038 routing retry time is respected, so at least it doesn't keep
1039 hammering the server.
1041 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1042 in ${readsocket because some servers need this prod.
1044 PH/32 Added some extra debug output when updating a wait-xxx database.
1046 PH/33 The hint "could be header name not terminated by colon", which has been
1047 given for certain expansion errors for a long time, was not being given
1048 for the ${if def:h_colon_omitted{... case.
1050 PH/34 The spec says: "With one important exception, whenever a domain list is
1051 being scanned, $domain contains the subject domain." There was at least
1052 one case where this was not true.
1054 PH/35 The error "getsockname() failed: connection reset by peer" was being
1055 written to the panic log as well as the main log, but it isn't really
1056 panic-worthy as it just means the connection died rather early on. I have
1057 removed the panic log writing for the ECONNRESET error when getsockname()
1060 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1061 runs only) independently of the message's sender address. This meant
1062 that, if the 4xx error was in fact related to the sender, a different
1063 message to the same recipient with a different sender could confuse
1064 things. In particualar, this can happen when sending to a greylisting
1065 server, but other circumstances could also provoke similar problems.
1066 I have changed the default so that the retry time for these errors is now
1067 based a combination of the sender and recipient addresses. This change
1068 can be overridden by setting address_retry_include_sender=false in the
1071 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1072 remote server are returned as part of bounce messages. This was not
1073 happening for LMTP over a pipe (the lmtp transport), but now it is the
1074 same for both kinds of LMTP.
1076 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1077 in header lines that were in fact CNAMEs. This is no longer the case.
1079 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1080 and queue runs started by the daemon processed all messages. This has
1081 been fixed so that -R and -S can now usefully be given with -q<time>.
1083 PH/40 Import PCRE release 6.7 (fixes some bugs).
1085 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1087 PH/42 Give an error if -q is specified more than once.
1089 PH/43 Renamed the variables $interface_address and $interface_port as
1090 $received_ip_address and $received_port, to make it clear that these
1091 values apply to message reception, and not to the outgoing interface when
1092 a message is delivered. (The old names remain recognized, of course.)
1094 PH/44 There was no timeout on the connect() call when using a Unix domain
1095 socket in the ${readsocket expansion. There now is.
1097 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1098 be meaningful with "accept".
1100 SC/01 Eximstats V1.43
1101 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1103 SC/02 Eximstats V1.44
1104 Use a glob alias rather than an array ref in the generated
1105 parser. This improves both readability and performance.
1107 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1108 Collect SpamAssassin and rejection statistics.
1109 Don't display local sender or destination tables unless
1110 there is data to show.
1111 Added average volumes into the top table text output.
1113 SC/04 Eximstats V1.46
1114 Collect data on the number of addresses (recipients)
1115 as well as the number of messages.
1117 SC/05 Eximstats V1.47
1118 Added 'Message too big' to the list of mail rejection
1119 reasons (thanks to Marco Gaiarin).
1121 SC/06 Eximstats V1.48
1122 Mainlog lines which have GMT offsets and are too short to
1123 have a flag are now skipped.
1125 SC/07 Eximstats V1.49 (Alain Williams)
1126 Added the -emptyok flag.
1128 SC/08 Eximstats V1.50
1129 Fixes for obtaining the IP address from reject messages.
1131 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1132 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1133 whitesspace changes from 4.64-PH/27
1135 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1138 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1139 are found, allow negative numbers in numeric criteria)
1141 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1143 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1144 to match changes made in 4.64-PH/43
1146 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1148 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1149 "domain" is actually an IP address, and force a failure. This locks out
1150 those revolvers/nameservers that support "A-for-A" lookups, in
1151 contravention of the specifications.
1153 PH/48 When a host name was looked up from an IP address, and the subsequent
1154 forward lookup of the name timed out, the host name was left in
1155 $sender_host_name, contrary to the specification.
1157 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1158 restricted to single-key lookups, Exim was not diagnosing an error if
1159 * or *@ was used with a query-style lookup.
1161 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1163 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1164 long ago, but noone interested enough thought of it. Let's just say that
1165 the "1.1" means that there are some new functions that weren't there at
1166 some point in the past.
1168 PH/51 Error processing for expansion failure of helo_data from an smtp
1169 transport during callout processing was broken.
1171 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1172 tested/used via the -bh/-bhc/-bs options.
1174 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1175 bug, fixed in subsequent PCRE releases).
1177 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1178 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1180 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1186 SC/01 Use a glob alias rather than an array ref in eximstats generated
1187 parser. This improves both readability and performance.
1189 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1190 Don't display local sender or destination tables in eximstats unless
1191 there is data to show.
1192 Added average volumes into the eximstats top table text output.
1194 SC/03 Collect data on the number of addresses (recipients) as well
1195 as the number of messages in eximstats.
1197 TF/01 Correct an error in the documentation for the redirect router. Exim
1198 does (usually) call initgroups() when daemonizing.
1200 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1201 with consistent privilege compared to when running as a daemon.
1203 TF/03 Note in the spec that $authenticated_id is not set for local
1204 submissions from trusted users.
1206 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1207 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1209 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1210 by adding some example configuration directives to the default
1211 configuration file. A little bit of work is required to uncomment the
1212 directives and define how usernames and passwords are checked, but
1213 there is now a framework to start from.
1215 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1216 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1217 without this. I don't know how relevant this is to other LDAP libraries.
1219 PH/02 Add the verb name to the "unknown ACL verb" error.
1221 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1223 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1225 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1226 directory not expanded when it should be if an expanded home directory
1227 was set for the address (which is overridden by the transport).
1229 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1232 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1233 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1234 because it is too late at that time, and has no effect.
1236 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1237 security issue with \' (bugzilla #107). I could not use the
1238 PQescapeStringConn() function, because it needs a PGconn value as one of
1241 PH/08 When testing addresses using -bt, indicate those final addresses that
1242 are duplicates that would not cause an additional delivery. At least one
1243 person was confused, thinking that -bt output corresponded to deliveries.
1244 (Suppressing duplicates isn't a good idea as you lose the information
1245 about possibly different redirections that led to the duplicates.)
1247 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1248 systems where poll() doesn't work, in particular OS X.
1250 PH/10 Added more information to debugging output for retry time not reached.
1252 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1253 operations in malware.c.
1255 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1258 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1259 syslog_duplication set false, log lines that would normally be written
1260 both the the main log and to the reject log were not written to syslog at
1263 PH/14 In the default configuration, change the use of "message" in ACL warn
1264 statements to "add_header".
1266 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1267 not followed by a command (e.g. "seen endif").
1269 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1270 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1273 PH/17 Added extra conditions to the default value of delay_warning_condition
1277 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1278 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1279 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1282 The Auto-Submitted: and various List- headers are standardised, whereas I
1283 don't think Precedence: ever was.
1285 PH/18 Refactored debugging code in route_finduser() to show more information,
1286 in particular, the error code if getpwnam() issues one.
1288 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1289 This is apparently needed in addition to the PH/07 change above to avoid
1290 any possible encoding problems.
1292 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1293 but not after initializing Perl.
1295 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1296 output them only if debugging. By default they are written stderr,
1297 apparently, which is not desirable.
1299 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1302 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1305 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1307 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1308 authenticated or an ident call has been made. Suppress the default
1309 values for $authenticated_id and $authenticated_sender (but permit -oMai
1310 and -oMas) when testing with -bh.
1312 PH/24 Re-jigged the order of the tests in the default configuration so that the
1313 tests for valid domains and recipients precede the DNS black list and CSA
1314 tests, on the grounds that those ones are more expensive.
1316 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1317 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1318 This bug exists in every version of Exim that I still have, right back to
1321 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1322 However, an attempt to turn on case-sensitivity in a regex key by
1323 including (?-i) didn't work because the subject string was already
1324 lowercased, and the effects were non-intuitive. It turns out that a
1325 one-line patch can be used to allow (?-i) to work as expected.
1331 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1332 other effects) broke the use of negated acl sub-conditions.
1334 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1337 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1338 "Deny" causes Exim to reject the incoming connection with a 554 error.
1339 Unfortunately, if there is a major crisis, such as a disk failure,
1340 tcp-wrappers gives "deny", whereas what one would like would be some
1341 kind of temporary error. A kludge has been added to help with this.
1342 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1343 554 error is used if errno is still zero or contains ENOENT (which occurs
1344 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1347 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1349 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1350 errors. Otherwise a message that provokes a temporary error (when other
1351 messages do not) can cause a whole host to time out.
1353 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1354 addresses were routed directly to files or pipes from a redirect router.
1355 File deliveries just didn't batch; pipe deliveries might have suffered
1358 PH/06 A failure to get a lock for a hints database would erroneously always say
1359 "Failed to get write lock", even when it was really a read lock.
1361 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1362 of 0600. This has been changed to use the value of the lockfile_mode
1363 option (which defaults to 0600).
1365 PH/08 Applied small patch from the Sieve maintainer.
1367 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1368 folder from quota calculations, a direct delivery into this folder messed
1369 up the contents of the maildirsize file. This was because the regex was
1370 used only to exclude .Trash (or whatever) when the size of the mailbox
1371 was calculated. There was no check that a delivery was happening into an
1372 excluded directory. This bug has been fixed by ignoring all quota
1373 processing for deliveries into excluded directories.
1375 PH/10 Added the maildirfolder_create_regex option to appendfile.
1381 PH/01 The code for finding all the local interface addresses on a FreeBSD
1382 system running IPv6 was broken. This may well have applied to all BSD
1383 systems, as well as to others that have similar system calls. The broken
1384 code found IPv4 interfaces correctly, but gave incorrect values for the
1385 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1386 that it would not match correctly against @[] and not recognize the IPv6
1389 PH/02 The ipliteral router was not recognizing addresses of the form user@
1390 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1392 PH/03 Added disable_ipv6.
1394 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1395 decoded form, because it is most often used to construct To: headers
1396 lines in autoreplies, and the decoded form may well be syntactically
1397 invalid. However, $reply_address has leading white space removed, and all
1398 newlines turned into spaces so that the autoreply transport does not
1401 PH/05 If group was specified without a user on a router, and no group or user
1402 was specified on a transport, the group from the router was ignored.
1404 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1405 for visible compile-time settings that can be used to change these
1406 numbers, for those that want even more. Backwards compatibility with old
1407 spool files has been maintained. However, going back to a previous Exim
1408 release will lost any variables that are in spool files.
1410 PH/07 Two small changes when running in the test harness: increase delay when
1411 passing a TCP/IP connection to a new process, in case the original
1412 process has to generate a bounce, and remove special handling of
1413 127.0.0.2 (sic), which is no longer necessary.
1415 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1416 be the same on different OS.
1418 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1421 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1422 whether --show-vars was specified or not
1424 JJ/02 exipick: Added support for new ACL variable spool format introduced
1427 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1428 syntactically invalid From: or Reply-to: line, and a filter used this to
1429 generate an autoreply, and therefore failed to obtain an address for the
1430 autoreply, Exim could try to deliver to a non-existent relative file
1431 name, causing unrelated and misleading errors. What now happens is that
1432 it logs this as a hard delivery error, but does not attempt to create a
1435 PH/11 The exinext utility has a -C option for testing purposes, but although
1436 the given file was scanned by exinext itself; it wasn't being passed on
1437 when Exim was called.
1439 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1440 an end-of-file indication when reading a command response.
1442 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1443 compiled. In many other places in Exim, IPv6 addresses are always
1444 recognized, so I have changed this. It also means that IPv4 domain
1445 literals of the form [IPV4:n.n.n.n] are now always recognized.
1447 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1448 used if the router is not running as root, for example, when verifying at
1449 ACL time, or when using -bh. The debugging output from this situation was
1450 non-existent - all you got was a failure to exec. I have made two
1453 (a) Failures to set uid/gid, the current directory, or a process leader
1454 in a subprocess such as that created by queryprogram now generate
1455 suitable debugging ouput when -d is set.
1457 (b) The queryprogram router detects when it is not running as root,
1458 outputs suitable debugging information if -d is set, and then runs
1459 the subprocess without attempting to change uid/gid.
1461 PH/15 Minor change to Makefile for building test_host (undocumented testing
1464 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1465 additional section of a DNS packet that returns MX or SRV records.
1466 Instead, it always explicitly searches for A/AAAA records. This avoids
1467 major problems that occur when a DNS server includes only records of one
1468 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1469 fixed another bug: if SRV records were looked up and the corresponding
1470 address records were *not* found in the additional section, the port
1471 values from the SRV records were lost.
1473 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1474 using the correct key (the original address) when searching the retry
1475 rules in order to find which one to use for generating the retry hint.
1477 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1478 adding the default one. Similarly, if it contains a Reply-To: header, the
1479 errors_reply_to option, if set, is not used.
1481 PH/19 When calculating a retry time, Exim used to measure the "time since
1482 failure" by looking at the "first failed" field in the retry record. Now
1483 it does not use this if it is later than than the arrival time of the
1484 message. Instead it uses the arrival time. This makes for better
1485 behaviour in cases where some deliveries succeed, thus re-setting the
1486 "first failed" field. An example is a quota failure for a huge message
1487 when small messages continue to be delivered. Without this change, the
1488 "time since failure" will always be short, possible causing more frequent
1489 delivery attempts for the huge message than are intended.
1490 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1492 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1493 $1, $2, $3) because the numerical variables can be reset during some
1494 expansion items (e.g. "match"), thereby losing the authentication data.
1496 PH/21 Make -bV show the size of off_t variables so that the test suite can
1497 decide whether to run tests for quotas > 2G.
1499 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1500 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1501 filecount value is greater than 2G or if a quota value is greater than 2G
1502 on a system where the size of off_t is not greater than 4, a panic error
1505 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1506 never match. The debug and -bh output now contains an explicit error
1507 message indicating a malformed IPv4 address or mask.
1509 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1510 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1511 PH/23 above applies.
1513 PH/25 Do not write to syslog when running in the test harness. The only
1514 occasion when this arises is a failure to open the main or panic logs
1515 (for which there is an explicit test).
1517 PH/26 Added the /no_tell option to "control=freeze".
1519 PH/27 If a host name lookup failed very early in a connection, for example, if
1520 the IP address matched host_lookup and the reverse lookup yielded a name
1521 that did not have a forward lookup, an error message of the form "no IP
1522 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1523 could be logged. Now it outputs the IP address instead of "NULL".
1525 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1526 allows the sender and the authenticated sender to be set when
1527 submitting a message from within Exim. Since child_open_exim() is
1528 documented for local_scan(), the new function should be too.
1530 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1531 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1532 results in an empty string is now treated as unset.
1534 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1536 PH/31 Added sender_verify_fail logging option.
1538 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1539 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1540 tidied the source and removed it altogether.
1542 PH/33 When a queue run was abandoned because the load average was too high, a
1543 log line was always written; now it is written only if the queue_run log
1544 selector is set. In addition, the log line for abandonment now contains
1545 information about the queue run such as the pid. This is always present
1546 in "start" and "stop" lines but was omitted from the "abandon" line.
1548 PH/34 Omit spaces between a header name and the colon in the error message that
1549 is given when verify = headers_syntax fails (if there are lots of them,
1550 the message gets confusing).
1552 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1553 names, as there are now some PTR records that contain slashes. This check
1554 is only to protect against broken name servers that fall over on strange
1555 characters, so the fact that it applies to all lookups doesn't matter.
1557 PH/36 Now that the new test suite is complete, we can remove some of the
1558 special code in Exim that was needed for the old test suite. For example,
1559 sorting DNS records because real resolvers return them in an arbitrary
1560 order. The new test suite's fake resolver always returns records in the
1563 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1564 bounces) except when queue_only is set, to avoid logging races between
1565 the different processes.
1567 PH/38 Panic-die if .include specifies a non-absolute path.
1569 PH/39 A tweak to the "H" retry rule from its user.
1571 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1572 a label. They prevented compilation on older perls.
1574 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1575 a warning to be raised on newish perls.
1577 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1578 on queue. Changes to match documented behaviour of showing count of
1579 messages matching specified criteria.
1581 PH/40 Changed the default ident timeout from 30s to 5s.
1583 PH/41 Added support for the use of login_cap features, on those BSD systems
1584 that have them, for controlling the resources used by pipe deliveries.
1586 PH/42 The content-scanning code uses fopen() to create files in which to put
1587 message data. Previously it was not paying any attention to the mode of
1588 the files. Exim runs with umask(0) because the rest of the code creates
1589 files with open(), and sets the required mode explicitly. Thus, these
1590 files were ending up world-writeable. This was not a big issue, because,
1591 being within the spool directory, they were not world-accessible. I have
1592 created a function called modefopen, which takes an additional mode
1593 argument. It sets umask(777), creates the file, chmods it to the required
1594 mode, then resets the umask. All the relevant calls to fopen() in the
1595 content scanning code have been changed to use this function.
1597 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1598 to 24 hours. This avoids potential overflow problems when processing G
1599 and H retry rules. I suspect nobody ever tinkers with this value.
1601 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1603 PH/45 When the plaintext authenticator is running as a client, the server's
1604 challenges are checked to ensure they are valid base64 strings. By
1605 default, the authentication attempt is cancelled if an invalid string is
1606 received. Setting client_ignore_invalid_base64 true ignores these errors.
1607 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1608 they are received. Thus, the responses can be made to depend on the
1609 challenges. If an invalid string is ignored, an empty string is placed in
1612 PH/46 Messages that are created by the autoreply transport now contains a
1613 References: header, in accordance with RFCs 2822 and 3834.
1615 PH/47 Added authenticated_sender_force to the smtp transport.
1617 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1619 PH/49 Installed latest patch from the Sieve maintainer.
1621 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1622 was also set, the appendfile transport was unnecessarily scanning a
1623 directory of message files (e.g. for maildir delivery) to find the count
1624 of files (along with the size), even though it did not need this
1625 information. It now does the scan only if it needs to find either the
1626 size of the count of files.
1628 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1630 PH/52 Two bugs concerned with error handling when the smtp transport is
1633 (i) Exim was not creating retry information for temporary errors given
1634 for individual recipients after the DATA command when the smtp transport
1635 was used in LMTP mode. This meant that they could be retried too
1636 frequently, and not timed out correctly.
1638 (ii) Exim was setting the flag that allows error details to be returned
1639 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1640 recipients that were returned after the DATA command.
1642 PH/53 This is related to PH/52, but is more general: for any failing address,
1643 when detailed error information was permitted to be returned to the
1644 sender, but the error was temporary, then after the final timeout, only
1645 "retry timeout exceeded" was returned. Now it returns the full error as
1646 well as "retry timeout exceeded".
1648 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1649 do this, and (what is worse) MTAs that accept it.
1651 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1652 will now be deprecated.
1654 PH/56 New os.c-cygwin from the Cygwin maintainer.
1656 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1657 formats (previously only available in exim formats via -bpr, -bpru,
1658 and -bpra. Now also available in native and exiqgrep formats)
1660 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1661 with very large, slow to parse queues
1663 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1665 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1667 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1668 responses to authentication challenges, though it was showing the
1669 challenges; (ii) I've removed the CR characters from the debug output for
1672 PH/58 Allow for the insertion of a newline as well as a space when a string
1673 is turned into more than one encoded-word during RFC 2047 encoding. The
1674 Sieve code now uses this.
1676 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1677 data_4xx, lost_connection, tls_required.
1679 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1680 message was being sent as an SMTP response.
1682 PH/61 Add -l and -k options to exicyclog.
1684 PH/62 When verifying, if an address was redirected to one new address, so that
1685 verification continued, and the new address failed or deferred after
1686 having set something in $address_data, the value of $address_data was not
1687 passed back to the ACL. This was different to the case when no
1688 redirection occurred. The value is now passed back in both cases.
1690 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1691 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1692 use login_cap.h, so on its own it isn't the distinguishing feature. The
1693 new name refers directly to the setclassresources() function.
1695 PH/65 Added configuration files for NetBSD3.
1697 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1699 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1700 is preferred over IPv4.
1702 PH/68 The bounce_return_message and bounce_return_body options were not being
1703 honoured for bounces generated during the reception of non-SMTP messages.
1704 In particular, this applied to messages rejected by the ACL. This bug has
1705 been fixed. However, if bounce_return_message is true and bounce_return_
1706 body is false, the headers that are returned for a non-SMTP message
1707 include only those that have been read before the error was detected.
1708 (In the case of an ACL rejection, they have all been read.)
1710 PH/69 The HTML version of the specification is now built in a directory called
1711 spec_html instead of spec.html, because the latter looks like a path with
1712 a MIME-type, and this confuses some software.
1714 PH/70 Catch two compiler warnings in sieve.c.
1716 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1717 function verify_get_ident() calls ip_connect() to connect a socket, but
1718 if the "connect()" function timed out, ip_connect() used to close the
1719 socket. However, verify_get_ident() also closes the socket later, and in
1720 between Exim writes to the log, which may get opened at this point. When
1721 the socket was closed in ip_connect(), the log could get the same file
1722 descriptor number as the socket. This naturally causes chaos. The fix is
1723 not to close the socket in ip_connect(); the socket should be closed by
1724 the function that creates it. There was only one place in the code where
1725 this was missing, in the iplookup router, which I don't think anybody now
1726 uses, but I've fixed it anyway.
1728 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1729 well as to direct DNS lookups. Otherwise the handling of names in host
1730 lists is inconsistent and therefore confusing.
1736 PH/01 Two changes to the default runtime configuration:
1738 (1) Move the checks for relay_from_hosts and authenticated clients from
1739 after to before the (commented out) DNS black list checks.
1741 (2) Add control=submission to the relay_from_hosts and authenticated
1742 clients checks, on the grounds that messages accepted by these
1743 statements are most likely to be submissions.
1745 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1747 (1) Generate an error if the third argument for the ${prvs expansion is
1750 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1753 (3) Reset the variables that are obtained from the first argument of
1754 ${prvscheck and used in the second argument before leaving the code,
1755 because their memory is reclaimed, so using them afterwards may do
1758 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1759 one (it's much easier than Tom thought :-).
1761 (5) Because of (4), we can now allow for the use of $prvscheck_result
1762 inside the third argument.
1764 PH/03 For some reason, the default setting of PATH when running a command from
1765 a pipe transport was just "/usr/bin". I have changed it to
1768 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1769 anything to be listed in the output from -bV.
1771 PH/05 When a filter generated an autoreply, the entire To: header line was
1772 quoted in the delivery log line, like this:
1774 => >A.N.Other <ano@some.domain> <original@ddress> ...
1776 This has been changed so that it extracts the operative address. There
1777 may be more than one such address. If so, they are comma-separated, like
1780 => >ano@some.domain,ona@other.domain <original@ddress> ...
1782 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1783 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1784 not being looked up in the rDNS to get a host name, Exim was showing the
1785 IP address twice in Received: lines, even though the IP addresses were
1786 identical. For example:
1788 Received: from [1.2.3.4] (helo=[1.2.3.4])
1790 However, if the real host name was known, it was omitting the HELO data
1791 if it matched the actual IP address. This has been tidied up so that it
1792 doesn't show the same IP address twice.
1794 PH/07 When both +timestamp and +memory debugging was on, the value given by
1795 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1796 called by the debug printing, thereby overwriting the timestamp buffer.
1797 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1799 PH/08 When the original message was included in an autoreply transport, it
1800 always said "this is a copy of the message, including all the headers",
1801 even if body_only or headers_only was set. It now gives an appropriate
1804 PH/09 Applied a patch from the Sieve maintainer which:
1806 o fixes some comments
1807 o adds the (disabled) notify extension core
1808 o adds some debug output for the result of if/elsif tests
1809 o points to the current vacation draft in the documentation
1810 and documents the missing references header update
1814 o fixes a bug in processing the envelope test (when testing
1815 multiple envelope elements, the last element determinted the
1818 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1819 Electronic Mail") by including:
1821 Auto-submitted: auto-generated
1823 in the messages that it generates (bounce messages and others, such as
1824 warnings). In the case of bounce messages for non-SMTP mesages, there was
1825 also a typo: it was using "Auto_submitted" (underscore instead of
1826 hyphen). Since every message generated by Exim is necessarily in response
1827 to another message, thes have all been changed to:
1829 Auto-Submitted: auto-replied
1831 in accordance with these statements in the RFC:
1833 The auto-replied keyword:
1835 - SHOULD be used on messages sent in direct response to another
1836 message by an automatic process,
1838 - MUST NOT be used on manually-generated messages,
1840 - MAY be used on Delivery Status Notifications (DSNs) and Message
1841 Disposition Notifications (MDNs),
1843 - MUST NOT be used on messages generated by automatic or periodic
1844 processes, except for messages which are automatic responses to
1847 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1848 to the default Received: header definition.
1850 PH/12 Added log selector acl_warn_skipped (default on).
1852 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1853 variables because (a) they are in the wrong storage pool and (b) even if
1854 they were copied, it wouldn't work properly because of the caching.
1856 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1857 checking when decoding. Apparently there are clients that generate
1858 overlong encoded strings. Why am I not surprised?
1860 PH/15 If the first argument of "${if match_address" was not empty, but did not
1861 contain an "@" character, Exim crashed. Now it writes a panic log message
1862 and treats the condition as false.
1864 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1866 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1867 extension "envelope-auth". The code is finished and in agreement with
1868 other implementations, but there is no documentation so far and in fact,
1869 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1870 not changing the active code.
1872 Print executed "if" and "elsif" statements when debugging is used. This
1873 helps a great deal to understand what a filter does.
1875 Document more things not specified clearly in RFC3028. I had all this
1876 sorted out, when out of a sudden new issues came to my mind. Oops."
1878 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1881 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1882 canonical form (as documented). However, after a host name lookup from
1883 the IP address, check_host() was doing a simple string comparison with
1884 addresses acquired from the DNS when checking that the found name did
1885 have the original IP as one of its addresses. Since any found IPv6
1886 addresses are likely to be in abbreviated form, the comparison could
1887 fail. Luckily, there already exists a function for doing the comparison
1888 by converting both addresses to binary, so now that is used instead of
1889 the text comparison.
1891 PH/20 There was another similar case to PH/19, when a complete host name was
1892 given in a host list; looking up its IP address could give an abbreviated
1893 form, whereas the current host's name might or might not be abbreviated.
1894 The same fix has been applied.
1900 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1901 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1904 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1905 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1907 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1909 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1910 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1911 Also, for "Auto-Submitted", treat anything other than "no" as
1912 non-personal, in accordance with RFC 3834. (Previously it treated
1913 anything starting "auto-" as non-personal.)
1915 TF/01 The control=submission/name=... option had a problem with syntax
1916 errors if the name included a slash character. The /name= option
1917 now slurps the rest of the string, so it can include any characters
1918 but it must come last in the list of options (after /sender_retain
1921 PH/05 Some modifications to the interface to the fake nameserver for the new
1929 TK/01 Added the "success_on_redirect" address verification option. See
1930 NewStuff for rationale and an example.
1932 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1934 PH/02 Patch to exigrep to allow it to work on syslog lines.
1936 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1937 fread() to skip over the body file's header line, because in Cygwin the
1938 header line is locked and is inaccessible.
1940 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1941 co-exist for some time) to make it clear that it is the Exim ID that is
1942 referenced, not the Message-ID: header line.
1944 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1945 string_format() function, because snprintf() does not exist on all
1948 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1949 ${stat: expansion item.
1951 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1952 protocol synchronization error", to keep the pedants happy.
1954 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1955 well as for IRIX systems, when gcc is being used. See the host.c source
1958 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1960 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1963 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1964 transport and to the smtp transport in LMTP mode.
1966 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1968 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1969 run a filter in a subprocess. This could lead to confusion in subsequent
1970 lookups in the parent process. There should also be a search_tidyup() at
1971 the end of the subprocess.
1973 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1974 only if the host matched helo_try_verify_hosts, which caused the
1975 verification to occur when the EHLO/HELO command was issued. The ACL just
1976 tested the remembered result. Now, if a previous verification attempt has
1977 not happened, "verify = helo" does it there and then.
1979 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1981 TK/03 Fix log output including CR from clamd.
1983 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1984 exist provoked a memory error which could cause a segfault.
1986 PH/15 Installed PCRE 6.2
1988 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1990 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1991 of the problem. Specifically, suggested +O2 rather than +O1 for the
1994 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1996 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1997 record, the port was not correctly logged when the outgoing_port log
1998 selector was set (it logged the transort's default port).
2000 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2001 fallback_hosts, and "hosts" in the smtp transport.
2003 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2004 host errors such as "Connection refused".
2006 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2007 authentication with radiusclient 0.4.9:
2009 - Error returned from rc_read_config was caught wrongly
2010 - Username/password not passed on to radius server due to wrong length.
2012 The presumption is that some radiusclient API changes for 4.51/PH/17
2013 were not taken care of correctly. The code is still untested by me (my
2014 Linux distribution still has 0.3.2 of radiusclient), but it was
2015 contributed by a Radius user.
2017 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2018 expanding the "port" option of the smtp transport.
2020 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2021 while reading a MIME header. Thanks to Tom Hughes for a patch.
2023 PH/24 Include config.h inside local_scan.h so that configuration settings are
2026 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2027 that in an ACL for RCPT (for example), you can examine exactly what was
2030 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2031 commands, but it was not correctly comparing the address with the actual
2032 client host address. Thus, it would show the EHLO address in Received:
2033 header lines when this was not necessary.
2035 PH/27 Added the % operator to ${eval:}.
2037 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2038 it should be ignoring failures (because with -C, for example, it has lost
2039 privilege). It wasn't ignoring creation failures other than "already
2042 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2045 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2046 because an input line was too long, either on its own, or by virtue of
2047 too many continuations, the temporary file was not being removed, and the
2048 return code was incorrect.
2050 PH/31 Missing "BOOL" in function definition in filtertest.c.
2052 PH/32 Applied Sieve patches from the maintainer.
2054 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2056 PH/33 Added "verify = not_blind".
2058 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2059 Local/Makefile (with some defaults set). These are used in built scripts
2060 such as exicyclog, but they have never been used in the exim_install
2061 script (though there are many overriding facilities there). I have
2062 arranged that the exim_install script now takes note of these two
2065 PH/35 Installed configuration files for Dragonfly.
2067 PH/36 When a locally submitted message by a trusted user did not contain a
2068 From: header, and the sender address was obtained from -f or from an SMTP
2069 MAIL command, and the trusted user did not use -F to supply a sender
2070 name, $originator_name was incorrectly used when constructing a From:
2071 header. Furthermore, $originator_name was used for submission mode
2072 messages from external hosts without From: headers in a similar way,
2073 which is clearly wrong.
2075 PH/37 Added control=suppress_local_fixups.
2077 PH/38 When log_selector = +received_sender was set, and the addition of the
2078 sender made the log line's construction buffer exactly full, or one byte
2079 less than full, an overflow happened when the terminating "\n" was
2082 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2083 when the result of a list match is failure because a DNS lookup failed.
2085 PH/40 RM_COMMAND is now used in the building process.
2087 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2088 the "build-* directories that it finds.
2090 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2091 address in a domain literal was a prefix of an interface address.
2093 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2094 when verifying a sender address, unless rewrite_headers is false.
2096 PH/44 Wrote a long comment about why errors_to addresses are verified as
2097 recipients, not senders.
2099 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2100 the ratelimit ACL was added.
2102 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2104 PH/46 Added extra information about PostgreSQL errors to the error string.
2106 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2107 avoiding the need to install special zones in a real server. This is
2108 backwards compatible; if it can't find the fake resolver, it drops back.
2109 Thus, both old and new test suites can be run.
2111 TF/02 Added util/ratelimit.pl
2113 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2114 clock is set back in time.
2116 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2117 Candler <B.Candler@pobox.com>.
2119 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2120 OK for addresses that are the result of redirections.
2122 PH/48 A number of further additions for the benefit of the new test suite,
2123 including a fake gethostbyname() that interfaces to the fake DNS resolver
2126 TF/06 The fix for widen_domains has also been applied to qualify_single and
2127 search_parents which are the other dnslookup options that can cause
2130 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2133 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2135 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2136 service" if it encountered an MX record with an empty target hostname.
2137 The message is now "an MX or SRV record indicated no SMTP service".
2139 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2140 if the DNS of the sending site is misconfigured. This is quite a
2141 common situation. This change restores the behaviour of treating a
2142 helo verification defer as a failure.
2144 PH/51 If self=fail was set on a router, the bounce message did not include the
2145 actual error message.
2151 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2153 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2154 command itself ended in error, the underlying message about the transport
2155 filter timeout was being overwritten with the pipe command error. Now the
2156 underlying error message should be appended to the second error message.
2158 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2160 PH/02 Reduce the amount of output that "make" produces by default. Full output
2161 can still be requested.
2163 PH/03 The warning log line about a condition test deferring for a "warn" verb
2164 was being output only once per connection, rather than after each
2165 occurrence (because it was using the same function as for successful
2166 "warn" verbs). This seems wrong, so I have changed it.
2168 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2169 it should not have, which might have caused a crash in the right
2170 circumstances, but probably never did.
2172 PH/04 Installed a modified version of Tony Finch's patch to make submission
2173 mode fix the return path as well as the Sender: header line, and to
2174 add a /name= option so that you can make the user's friendly name appear
2177 TF/03 Added the control = fakedefer ACL modifier.
2179 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2180 Mark Lowes for thorough testing.
2182 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2184 TK/03 Merged latest SRS patch from Miles Wilton.
2186 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2187 with the definition in sysexits.h (which is #included earlier).
2188 Fortunately, Exim does not actually use EX_OK. The code used to try to
2189 preserve the sysexits.h value, by assumimg that macro definitions were
2190 scanned for macro replacements. I have been disabused of this notion,
2191 so now the code just undefines EX_OK before #including unistd.h.
2193 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2194 in the smtp transport. When a block could not be written in a single
2195 write() function, the timeout was being re-applied to each part-write.
2196 This seems wrong - if the receiver was accepting one byte at a time it
2197 would take for ever. The timeout is now adjusted when this happens. It
2198 doesn't have to be particularly precise.
2200 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2201 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2203 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2204 without a domain if the check to <postmaster@domain> fails.
2206 SC/01 Eximstats: added -xls and the ability to specify output files
2207 (patch written by Frank Heydlauf).
2209 SC/02 Eximstats: use FileHandles for outputing results.
2211 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2213 SC/04 Eximstats: fixed display of large numbers with -nvr option
2215 SC/05 Eximstats: fixed merging of reports with empty tables.
2217 SC/06 Eximstats: added the -include_original_destination flag
2219 SC/07 Eximstats: removed tabs and trailing whitespace.
2221 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2223 TK/06 MBOX spool code: Add real "From " MBOX separator line
2224 so the .eml file is really in mbox format (even though
2225 most programs do not really care). Patch from Alex Miller.
2227 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2228 The latter is generated from $received_to and is only set if the
2229 message has one envelope recipient. SA can use these headers,
2230 obviously out-of-the-box. Patch from Alex Miller.
2232 PH/08 The ${def test on a variable was returning false if the variable's
2233 value was "0", contrary to what the specification has always said!
2234 The result should be true unless the variable is empty.
2236 PH/09 The syntax error of a character other than { following "${if
2237 def:variable_name" (after optional whitespace) was not being diagnosed.
2238 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2239 accidental colon was present, for example, could give incorrect results.
2241 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2242 result is used (not including appendfile, where other changes are about
2245 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2246 This involved changing a lot of size variables from int to off_t. It
2247 should work with maildirs and everything.
2249 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2250 spamd dying while we are connected to it.
2252 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2255 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2256 function for BDB 4.3.
2258 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2260 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2261 the sources for PCRE have been re-arranged and more files are now
2264 PH/15 The code I had for printing potentially long long variables in PH/11
2265 above was not the best (it lost precision). The length of off_t variables
2266 is now inspected at build time, and an appropriate printing format (%ld
2267 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2268 to be "long long int" or "long int". This is needed for the internal
2269 formatting function string_vformat().
2271 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2272 the configuration file to be ":syslog", then the script "guesses" where
2273 the logs files are, rather than using the compiled in default. In our
2274 case the guess is not the same as the compiled default, so the script
2275 suddenly stopped working when I started to use syslog. The patch checks
2276 to see if log_file_path is "". If so, it attempts to read it from exim
2277 with no configuration file to get the compiled in version, before it
2278 falls back to the previous guessing code."
2280 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2281 implementing BATV in an Exim configuration. See NewStuff for the gory
2284 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2285 Makefile that are specific to HP-UX.
2287 PH/18 If the "use_postmaster" option was set for a recipient callout together
2288 with the "random" option, the postmaster address was used as the MAIL
2289 FROM address for the random test, but not for the subsequent recipient
2290 test. It is now used for both.
2292 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2293 patch removes a few documentation additions to RFC 3028, because the
2294 latest draft now contains them. It adds the new en;ascii-case comparator
2295 and a new error check for 8bit text in MIME parts. Comparator and
2296 require names are now matched exactly. I enabled the subaddress
2297 extension, but it is not well tested yet (read: it works for me)."
2299 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2300 rework some of the code of TK/09 above to avoid the hardwired use of
2301 "%lld" and "long long". Replaced the call to snprintf() with a call to
2304 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2305 records point to non-existent hosts", "retry timeout exceeded", and
2306 "retry time not reached for any host after a long failure period".
2308 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2309 experimental DomainKeys support:
2311 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2312 (2) On an error such as an illegally used "control", the wrong name for
2313 the control was given.
2315 These problems did NOT occur unless DomainKeys support was compiled.
2317 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2319 PH/24 Added ${if match_ip condition.
2321 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2322 databases so that it will be absolutely obvious if a crash occurs in the
2323 DB library. This is a regular occurrence (often caused by mis-matched
2326 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2327 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2328 picked up on a user's system that detects such things. There doesn't seem
2329 to be a gcc warning option for this - only an attribute that has to be
2330 put on the function's prototype. It seems that in Fedora Core 4 they have
2331 set this on a number of new functions. No doubt there will be more in due
2334 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2335 specify a transport. However, if an address that was verified by such a
2336 router was the subject of a callout, Exim crashed because it tried to
2337 read the rcpt_include_affixes from the non-existent transport. Now it
2338 just assumes that the setting of that option is false. This bug was
2339 introduced by 4.51/PH/31.
2341 PH/28 Changed -d+all to exclude +memory, because that information is very
2342 rarely of interest, but it makes the output a lot bigger. People tend to
2343 do -d+all out of habit.
2345 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2346 code in os-type was giving problems when libc.so lives in lib64, like on
2349 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2350 aren't the modern standard, and it seems that some systems' include files
2351 don't always have them. Exim was already checking for some of the newer
2352 ones like T_AAAA, and defining it itself. I've added checks for all the
2353 record types that Exim uses.
2355 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2356 not automatically generating a new one, as it is supposed to. This
2357 prevented TLS from working. If the file did exist, but contained invalid
2358 data, a new version was generated, as expected. It was only the case of a
2359 non-existent file that was broken.
2361 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2362 with a change in libdomainkeys > 0.64.
2364 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2365 from DNS. If the selector record carries the flag, it now has
2366 precedence over the domain-wide flag.
2368 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2370 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2371 the use of a port name that isn't defined in /etc/services) occurred, the
2372 message was deferred as in a normal delivery, and thus remained on the
2373 spool, instead of being failed because of the mua_wrapper setting. This
2374 is now fixed, and I tidied up some of the mua_wrapper messages at the
2377 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2378 the messages in a hash of arrays rather than using individual hashes.
2379 This is a bit cleaner and results in dramatic memory savings, albeit
2380 at a slight CPU cost.
2382 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2383 as requested by Marc Sherman.
2385 SC/10 Eximstats: added histograms for user specified patterns as requested
2388 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2390 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2391 fopen() in the content-scanning modules that did not already have it.
2397 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2398 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2400 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2402 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2404 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2405 to test Sieve filters that use "vacation".
2407 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2408 that changes the way the GnuTLS parameters are stored in the cache file.
2409 The new format can be generated externally. For backward compatibility,
2410 if the data in the cache doesn't make sense, Exim assumes it has read an
2411 old-format file, and it generates new data and writes a new file. This
2412 means that you can't go back to an older release without removing the
2415 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2416 work if there are any delivery delays because "one_time" forces the
2417 parent to be marked "delivered", so its unseen clone is never tried
2418 again. For this reason, Exim now forbids the simultaneous setting of
2421 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2422 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2423 entry if you want to know the details. The fix, however, neglected to
2424 consider the case when local delivery batching is involved. The test for
2425 "previously delivered" was not happening when checking to see if an
2426 address could be batched with a previous (undelivered) one; under
2427 certain circumstances this could lead to multiple deliveries to the same
2430 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2431 in its include files, and this causes problems building Exim.
2433 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2434 header_syntax) but Exim was just ignoring anything given after a slash.
2435 In particular, this caused confusion with an attempt to use "verify =
2436 reverse_host_lookup/defer_ok". An error is now given when options are
2437 supplied for verify items that do not have them. (Maybe reverse_host_
2438 lookup should have a defer_ok option, but that's a different point.)
2440 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2441 defined by RFC 821) to 2048, because there were problems with some AUTH
2442 commands, and RFC 1869 says the size should be increased for extended
2443 SMTP commands that take arguments.
2445 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2448 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2449 "unknown" error; now it says that the functionality isn't in the binary.
2451 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2452 an address' error message when a string expansion fails (syntax or
2453 whatever). Otherwise the password may appear in the log. Following change
2454 PH/42 below, there is no longer a chance of it appearing in a bounce
2457 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2459 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2460 of its addresses was ever tried. (Bugzilla bug #2.)
2462 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2463 the result incorrectly in the debug output. (It correctly added a newline
2464 to what was transported.)
2466 TF/01 Added $received_time.
2468 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2469 commented out examples of how to interface to a virus scanner and to
2470 SpamAssassin. Also added commented examples of av_scanner and
2471 spamd_address settings.
2473 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2474 and controls are allowed in which ACLs. There were a couple of minor
2475 errors. Some of the entries in the conditions table (which is a table of
2476 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2477 negation of where the condition IS allowed.
2479 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2481 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2482 header file does not have a version number, so I've had to invent a new
2483 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2484 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2485 radiusclient), but it was contributed by a Radius user.
2487 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2488 files or directories, for OpenSSL.
2490 PH/19 When an Exim process that is running as root has to create an Exim log
2491 file, it does so in a subprocess that runs as exim:exim so as to get the
2492 ownership right at creation (otherwise, other Exim processes might see
2493 the file with the wrong ownership). There was no test for failure of this
2494 fork() call, which would lead to the process getting stuck as it waited
2495 for a non-existent subprocess. Forks do occasionally fail when resources
2496 run out. I reviewed all the other calls to fork(); they all seem to check
2499 PH/20 When checking for unexpected SMTP input at connect time (before writing
2500 the banner), Exim was not dealing correctly with a non-positive return
2501 from the read() function. If the client had disconnected by this time,
2502 the result was a log entry for a synchronization error with an empty
2503 string after "input=" when read() returned zero. If read() returned -1
2504 (an event I could not check), uninitialized data bytes were printed.
2505 There were reports of junk text (parts of files, etc) appearing after
2508 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2510 PH/22 Added support for macro redefinition, and (re)definition in between
2511 driver and ACL definitions.
2513 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2514 forgetting to use the resulting value; it was using the unexpanded value.
2516 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2517 hadn't been configured. The fix is from Juergen Kreileder, who
2518 understands it better than I do:
2520 "Here's what I see happening with three configured cyrus_sasl
2521 authenticators configured (plain, login, cram-md5):
2523 On startup auth_cyrus_sasl_init() gets called for each of these.
2524 This means three calls to sasl_listmech() without a specified mech_list.
2525 => SASL tests which mechs of all available mechs actually work
2526 => three warnings about OTP not working
2527 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2529 With the patch, sasl_listmech() also gets called three times. But now
2530 SASL's mech_list option is set to the server_mech specified in the the
2531 authenticator. Or in other words, the answer from sasl_listmech()
2532 gets limited to just the mech you're testing for (which is different
2534 => the return list contains just 'plain' or 'login', 'cram-md5' or
2535 nothing depending on the value of ob->server_mech.
2537 I've just tested the patch: Authentication still works fine,
2538 unavailable mechs specified in the exim configuration are still
2539 caught, and the auth.log warnings about OTP are gone."
2541 PH/25 When debugging is enabled, the contents of the command line are added
2542 to the debugging output, even when log_selector=+arguments is not
2545 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2546 answer is "GNU", and only if the return is "GNU/something" is the answer
2549 PH/27 $acl_verify_message is now set immediately after the failure of a
2550 verification in an ACL, and so is available in subsequent modifiers. In
2551 particular, the message can be preserved by coding like this:
2553 warn !verify = sender
2554 set acl_m0 = $acl_verify_message
2556 Previously, $acl_verify_message was set only while expanding "message"
2557 and "log_message" when a very denied access.
2559 PH/28 Modified OS/os.c-Linux with
2561 -#ifndef OS_LOAD_AVERAGE
2562 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2564 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2565 nomenclature these days.)
2567 PH/29 Installed patch from the Sieve maintainer that adds the options
2568 sieve_useraddress and sieve_subaddress to the redirect router.
2570 PH/30 In these circumstances:
2571 . Two addresses routed to the same list of hosts;
2572 . First host does not offer TLS;
2573 . First host accepts first address;
2574 . First host gives temporary error to second address;
2575 . Second host offers TLS and a TLS session is established;
2576 . Second host accepts second address.
2577 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2578 and peerdn, if requested) that were in fact used only for the second
2581 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2582 attention to any local part prefix or suffix that was matched by the
2583 router that accepted the address. It now behaves in the same way as it
2584 does for delivery: the affixes are removed from the local part unless
2585 rcpt_include_affixes is set on the transport.
2587 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2588 timeout during the DATA phase of an incoming message.
2590 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2591 applied a patch sanctioned by the Sieve maintainer.
2593 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2594 the uid or gid is negative. A case of a negative gid caused this to be
2595 noticed. The fix allows for either to be negative.
2597 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2598 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2601 PH/36 The change PH/12 above was broken. Fixed it.
2603 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2604 the grounds that routing the same address twice would always produce the
2605 same answer. This might have been true once, but it is certainly no
2606 longer true now. Routing a child address may depend on the previous
2607 routing that produced that child. Some complicated redirection strategies
2608 went wrong when messages had multiple recipients, and made Exim's
2609 behaviour dependent on the order in which the addresses were given.
2611 I have moved the duplicate checking until after the routing is complete.
2612 Exim scans the addresses that are assigned to local and remote
2613 transports, and removes any duplicates. This means that more work will be
2614 done, as duplicates will always all be routed, but duplicates are
2615 presumably rare, so I don't expect this is of any significance.
2617 For deliveries to pipes, files, and autoreplies, the duplicate checking
2618 still happens during the routing process, since they are not going to be
2621 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2622 It corrects a timeout issue with spamd. This is Ian's comment: "The
2623 background is that sometimes spamd either never reads data from a
2624 connection it has accepted, or it never writes response data. The exiscan
2625 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2626 blindly assumes that writes won't block so it may never time out."
2628 PH/39 Allow G after quota size as well as K and M.
2630 PH/40 The value set for $authenticated_id in an authenticator may not contain
2631 binary zeroes or newlines because the value is written to log lines and
2632 to spool files. There was no check on this. Now the value is run through
2633 the string_printing() function so that such characters are converted to
2634 printable escape sequences.
2636 PH/41 $message_linecount is a new variable that contains the total number of
2637 lines in the message. Compare $body_linecount, which is the count for the
2640 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2641 bounce and delay warning messages, except in certain special cases, which
2644 (a) An SMTP error message from a remote host;
2645 (b) A message specified in a :fail: redirection;
2646 (c) A message specified in a "fail" command in a system filter;
2647 (d) A message specified in a FAIL return from the queryprogram router;
2648 (e) A message specified by the cannot_route_message router option.
2650 In these cases only, Exim does include the error details in bounce and
2651 warning messages. There are also a few cases where bland messages such
2652 as "unrouteable address" or "local delivery error" are given.
2654 PH/43 $value is now also set for the "else" part of a ${run expansion.
2656 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2657 being worked on, but at least Exim now implements the latest version to
2660 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2661 process to complete was treated as a delivery failure, a timeout while
2662 writing the message to the pipe was logged, but erroneously treated as a
2663 successful delivery. Such timeouts include transport filter timeouts. For
2664 consistency with the overall process timeout, these timeouts are now
2665 treated as errors, giving rise to delivery failures by default. However,
2666 there is now a new Boolean option for the pipe transport called
2667 timeout_defer, which, if set TRUE, converts the failures into defers for
2668 both kinds of timeout. A transport filter timeout is now identified in
2671 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2672 systems where "make" and "gmake" are different, calling "gmake" at top
2673 level broke things. I've arranged for the value of $(MAKE) to be passed
2674 from the Makefile to this script so that it can call the same version of
2678 A note about Exim versions 4.44 and 4.50
2679 ----------------------------------------
2681 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2682 changes of various kinds. As a consequence, a big documentation update was
2683 needed. This delayed the release for rather longer than seemed good, especially
2684 in the light of a couple of (minor) security issues. Therefore, the changes
2685 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2686 release. So 4.44 and 4.50 are in effect two different branches that both start
2689 I have left the 4.50 change log unchanged; it contains all the changes since
2690 4.43. The change log for 4.44 is below; many of its items are identical to
2691 those for 4.50. This seems to be the most sensible way to preserve the
2692 historical information.
2698 1. Minor wording change to the doc/README.SIEVE file.
2700 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2701 computation of the current number of files was incorrect.
2703 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2704 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2707 4. Give more explanation in the error message when the command for a transport
2708 filter fails to execute.
2710 5. There are several places where Exim runs a non-Exim command in a
2711 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2712 was being done only for the command run by the queryprogram router. It is
2713 now done for all such subprocesses. The other cases are: ${run, transport
2714 filters, and the commands run by the lmtp and pipe transports.
2716 6. Added CONFIGURE_GROUP build-time option.
2718 7. Some older OS have a limit of 256 on the maximum number of file
2719 descriptors. Exim was using setrlimit() to set 1000 as a large value
2720 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2721 systems. I've change it so that if it can't get 1000, it tries for 256.
2723 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2724 was an oversight, and furthermore, ever since the addition of extra
2725 controls (e.g. 4.43/32), the checks on when to allow different forms of
2726 "control" were broken. There should now be diagnostics for all cases when a
2727 control that does not make sense is encountered.
2729 9. Added the /retain_sender option to "control=submission".
2731 10. $recipients is now available in the predata ACL (oversight).
2733 11. Tidy the search cache before the fork to do a delivery from a message
2734 received from the command line. Otherwise the child will trigger a lookup
2735 failure and thereby defer the delivery if it tries to use (for example) a
2736 cached ldap connection that the parent has called unbind on.
2738 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2739 of $address_data from the recipient verification was clobbered by the
2740 sender verification.
2742 13. The value of address_data from a sender verification is now available in
2743 $sender_address_data in subsequent conditions in the ACL statement.
2745 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2747 15. Added a new option "connect=<time>" to callout options, to set a different
2750 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2751 was its contents. (It was OK if the option was not defined at all.)
2753 17. A "Completed" log line is now written for messages that are removed from
2754 the spool by the -Mrm option.
2756 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2757 information about exactly what failed.
2759 19. Added -dd to debug only the daemon process.
2761 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2762 handles timeouts, both on the server side and network timeouts. Renamed the
2763 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2765 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2766 It is now set to "smtps".
2768 22. $host_address is now set to the target address during the checking of
2769 ignore_target_hosts.
2771 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2772 being passed; this would have caused $sender_host_name to have been used if
2773 matching the list had actually called for a host name (not very likely,
2774 since this list is usually IP addresses). A host name is now passed as
2777 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2778 code that specifies a non-restarting handler (typically sigaction() in
2779 modern systems) in an attempt to fix a rare and obscure crash bug.
2781 25. Narrowed the window for a race in the daemon that could cause it to ignore
2782 SIGCHLD signals. This is not a major problem, because they are used only to
2783 wake it up if nothing else does.
2785 26. A malformed maildirsize file could cause Exim to calculate negative values
2786 for the mailbox size or file count. Odd effects could occur as a result.
2787 The maildirsize information is now recalculated if the size or filecount
2790 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2791 support for a long time. Removed HAVE_SYS_VFS_H.
2793 28. Installed the latest version of exipick from John Jetmore.
2795 29. In an address list, if the pattern was not a regular expression, an empty
2796 subject address (from a bounce message) matched only if the pattern was an
2797 empty string. Non-empty patterns were not even tested. This was the wrong
2798 because it is perfectly reasonable to use an empty address as part of a
2799 database query. An empty address is now tested by patterns that are
2800 lookups. However, all the other forms of pattern expect the subject to
2801 contain a local part and a domain, and therefore, for them, an empty
2802 address still always fails if the pattern is not itself empty.
2804 30. Exim went into a mad DNS loop when attempting to do a callout where the
2805 host was specified on an smtp transport, and looking it up yielded more
2806 than one IP address.
2808 31. Re-factored the code for checking spool and log partition space into a
2809 function that finds that data and another that does the check. The former
2810 is then used to implement four new variables: $spool_space, $log_space,
2811 $spool_inodes, and $log_inodes.
2813 32. The RFC2047 encoding function was originally intended for short strings
2814 such as real names; it was not keeping to the 75-character limit for
2815 encoded words that the RFC imposes. It now respects the limit, and
2816 generates multiple encoded words if necessary. To be on the safe side, I
2817 have increased the buffer size for the ${rfc2047: expansion operator from
2820 33. It is now permitted to omit both strings after an "if" condition; if the
2821 condition is true, the result is "true". As before, when the second string
2822 is omitted, a false condition yields an empty string. This makes it less
2823 cumbersome to write custom ACL and router conditions.
2825 34. Failure to deliver a bounce message always caused it to be frozen, even if
2826 there was an errors_to setting on the router. The errors_to setting is now
2829 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2830 canonical form (fully expanded) before being placed in
2831 $sender_host_address.
2833 36. The table in the code that translates DNS record types into text (T_A to
2834 "A" for instance) was missing entries for NS and CNAME. It is just possible
2835 that this could have caused confusion if both these types were looked up
2836 for the same domain, because the text type is used as part of Exim's
2837 per-process caching. But the chance of anyone hitting this buglet seems
2840 37. The dnsdb lookup has been extended in a number of ways.
2842 (1) There is a new type, "zns", which walks up the domain tree until it
2843 finds some nameserver records. It should be used with care.
2845 (2) There is a new type, "mxh", which is like "mx" except that it returns
2846 just the host names, not the priorities.
2848 (3) It is now possible to give a list of domains (or IP addresses) to be
2849 looked up. The behaviour when one of the lookups defers can be
2850 controlled by a keyword.
2852 (4) It is now possible to specify the separator character for use when
2853 multiple records are returned.
2855 38. The dnslists ACL condition has been extended: it is now possible to supply
2856 a list of IP addresses and/or domains to be looked up in a particular DNS
2859 39. Added log_selector=+queue_time_overall.
2861 40. When running the queue in the test harness, wait just a tad after forking a
2862 delivery process, to get repeatability of debugging output.
2864 41. Include certificate and key file names in error message when GnuTLS fails
2865 to set them up, because the GnuTLS error message doesn't include the name
2866 of the failing file when there is a problem reading it.
2868 42. Allow both -bf and -bF in the same test run.
2870 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2872 44. The "Exiscan patch" is now merged into the mainline Exim source.
2874 45. Sometimes the final signoff response after QUIT could fail to get
2875 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2876 < 0 before doing a fflush(). This bug looks as though it goes back to the
2877 introduction of TLS in release 3.20, but "sometimes" must have been rare
2878 because the tests only now provoked it.
2880 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2881 (this can affect the format of dates).
2883 47. exim_tidydb, when checking for the continued existence of a message for
2884 which it has found a message-specific retry record, was not finding
2885 messages that were in split spool directories. Consequently, it was
2886 deleting retry records that should have stayed in existence.
2888 48. Steve fixed some bugs in eximstats.
2890 49. The SPA authentication driver was not abandoning authentication and moving
2891 on to the next authenticator when an expansion was forced to fail,
2892 contradicting the general specification for all authenticators. Instead it
2893 was generating a temporary error. It now behaves as specified.
2895 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2896 (the order specifies the preference for clients). The order is now AES256,
2897 AES128, 3DES, ARCFOUR128.
2899 51. Small patch to Sieve code - explicitly set From: when generating an
2902 52. Exim crashed if a remote delivery caused a very long error message to be
2903 recorded - for instance if somebody sent an entire SpamAssassin report back
2904 as a large number of 550 error lines. This bug was coincidentally fixed by
2905 increasing the size of one of Exim's internal buffers (big_buffer) that
2906 happened as part of the Exiscan merge. However, to be on the safe side, I
2907 have made the code more robust (and fixed the comments that describe what
2910 53. Now that there can be additional text after "Completed" in log lines (if
2911 the queue_time_overall log selector is set), a one-byte patch to exigrep
2912 was needed to allow it to recognize "Completed" as not the last thing in
2915 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2916 patch that reportedly fixes this has been added. I am not expert enough to
2917 create a test for it. This is what the patch creator wrote:
2919 "I found a little strange behaviour of ldap code when working with
2920 Windows 2003 AD Domain, where users was placed in more than one
2921 Organization Units. When I tried to give exim partial DN, the exit code
2922 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2923 But simultaneously result of request was absolutely normal ldap result,
2924 so I produce this patch..."
2926 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2927 REFERENCE, so I have modified the code to exclude the patch when that macro
2930 55. Some experimental protocols are using DNS PTR records for new purposes. The
2931 keys for these records are domain names, not reversed IP addresses. The
2932 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2933 leaves it alone. Component reversal etc. now happens only for IP addresses.
2936 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2938 57. Double the size of the debug message buffer (to 2048) so that more of very
2939 long debug lines gets shown.
2941 58. The exicyclog utility now does better if the number of log files to keep
2942 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2944 59. Two changes related to the smtp_active_hostname option:
2946 (1) $smtp_active_hostname is now available as a variable.
2947 (2) The default for smtp_banner uses $smtp_active_hostname instead
2948 of $primary_hostname.
2950 60. The host_aton() function is supposed to be passed a string that is known
2951 to be a valid IP address. However, in the case of IPv6 addresses, it was
2952 not checking this. This is a hostage to fortune. Exim now panics and dies
2953 if the condition is not met. A case was found where this could be provoked
2954 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2955 components; fortuitously, this particular loophole had already been fixed
2956 by change 4.50/55 above.
2958 If there are any other similar loopholes, the new check in host_aton()
2959 itself should stop them being exploited. The report I received stated that
2960 data on the command line could provoke the exploit when Exim was running as
2961 exim, but did not say which command line option was involved. All I could
2962 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2963 running as the user.
2966 61. There was a buffer overflow vulnerability in the SPA authentication code
2967 (which came originally from the Samba project). I have added a test to the
2968 spa_base64_to_bits() function which I hope fixes it.
2971 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2972 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2974 63. The daemon start-up calls getloadavg() while still root for those OS that
2975 need the first call to be done as root, but it missed one case: when
2976 deliver_queue_load_max is set with deliver_drop_privilege. This is
2977 necessary for the benefit of the queue runner, because there is no re-exec
2978 when deliver_drop_privilege is set.
2980 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2981 This has been fixed.
2983 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2984 was in use, was not putting the data itself into the right store pool;
2985 consequently, it could be overwritten for a subsequent message in the same
2986 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2989 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2991 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2992 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2993 the function were treating the return as a boolean value, which happened to
2994 work because 0=false and not-0=true, but is not correct code.
2996 68. The host_aton() function was not handling scoped IPv6 addresses (those
2997 with, for example, "%eth0" on the end) correctly.
2999 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3000 negated items (that is, ~something) in unsigned ints. Some compilers
3001 apparently mutter when there is no cast.
3003 70. If an address verification called from an ACL failed, and did not produce a
3004 user-specific message (i.e. there was only a "system" message), nothing was
3005 put in $acl_verify_message. In this situation, it now puts the system
3008 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3009 session; change 4.31/43 added the unwanted input to the log line - except
3010 that it did not do this in the start of session case. It now does.
3012 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3013 This is wrong and can cause the other end to generate a synchronization
3014 error if it is another Exim or anything else that does the synchronization
3015 check. A QUIT command is no longer sent after a timeout.
3017 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3018 during host lookups.
3020 74. The defer_ok option of callout verification was not working if it was used
3021 when verifying addresses in header lines, that is, for this case:
3023 verify = header_sender/callout=defer_ok
3025 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3026 those file descriptors could be used for SMTP connections. If anything
3027 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3028 could be sent to the SMTP client, causing chaos. The daemon now opens
3029 stdin, stdout, and stderr to /dev/null when it puts itself into the
3032 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3033 log by default. The user can override this with suitable Perl magic.
3035 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3036 the log message when discard triggers, was not working for the DATA ACL or
3037 for the non-SMTP ACL.
3039 78. Error message wording change in sieve.c.
3041 79. If smtp_accept_max_per_host was set, the number of connections could be
3042 restricted to fewer than expected, because the daemon was trying to set up
3043 a new connection before checking whether the processes handling previous
3044 connections had finished. The check for completed processes is now done
3045 earlier. On busy systems, this bug wouldn't be noticed because something
3046 else would have woken the daemon, and it would have reaped the completed
3049 80. If a message was submitted locally by a user whose login name contained one
3050 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3051 It caused a spool format error. I have fixed the spool reading code. A
3052 related problem was that the "from" clause in the Received: line became
3053 illegal because of the space(s). It is now covered by ${quote_local_part.
3055 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3058 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3060 83. Patch from Sieve maintainer to fix unterminated string problem in
3061 "vacation" handling.
3063 84. Some minor changes to the Linux configuration files to help with other
3064 OS variants using glibc.
3066 85. One more patch for Sieve to update vacation handling to latest spec.
3069 ----------------------------------------------------
3070 See the note above about the 4.44 and 4.50 releases.
3071 ----------------------------------------------------
3077 1. Change 4.43/35 introduced a bug that caused file counts to be
3078 incorrectly computed when quota_filecount was set in an appendfile
3081 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3082 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3085 3. Give more explanation in the error message when the command for a transport
3086 filter fails to execute.
3088 4. There are several places where Exim runs a non-Exim command in a
3089 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3090 was being done only for the command run by the queryprogram router. It is
3091 now done for all such subprocesses. The other cases are: ${run, transport
3092 filters, and the commands run by the lmtp and pipe transports.
3094 5. Some older OS have a limit of 256 on the maximum number of file
3095 descriptors. Exim was using setrlimit() to set 1000 as a large value
3096 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3097 systems. I've change it so that if it can't get 1000, it tries for 256.
3099 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3100 was an oversight, and furthermore, ever since the addition of extra
3101 controls (e.g. 4.43/32), the checks on when to allow different forms of
3102 "control" were broken. There should now be diagnostics for all cases when a
3103 control that does not make sense is encountered.
3105 7. $recipients is now available in the predata ACL (oversight).
3107 8. Tidy the search cache before the fork to do a delivery from a message
3108 received from the command line. Otherwise the child will trigger a lookup
3109 failure and thereby defer the delivery if it tries to use (for example) a
3110 cached ldap connection that the parent has called unbind on.
3112 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3113 of $address_data from the recipient verification was clobbered by the
3114 sender verification.
3116 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3117 was its contents. (It was OK if the option was not defined at all.)
3119 11. A "Completed" log line is now written for messages that are removed from
3120 the spool by the -Mrm option.
3122 12. $host_address is now set to the target address during the checking of
3123 ignore_target_hosts.
3125 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3126 being passed; this would have caused $sender_host_name to have been used if
3127 matching the list had actually called for a host name (not very likely,
3128 since this list is usually IP addresses). A host name is now passed as
3131 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3132 code that specifies a non-restarting handler (typically sigaction() in
3133 modern systems) in an attempt to fix a rare and obscure crash bug.
3135 15. Narrowed the window for a race in the daemon that could cause it to ignore
3136 SIGCHLD signals. This is not a major problem, because they are used only to
3137 wake it up if nothing else does.
3139 16. A malformed maildirsize file could cause Exim to calculate negative values
3140 for the mailbox size or file count. Odd effects could occur as a result.
3141 The maildirsize information is now recalculated if the size or filecount
3144 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3145 support for a long time. Removed HAVE_SYS_VFS_H.
3147 18. Updated exipick to current release from John Jetmore.
3149 19. Allow an empty sender to be matched against a lookup in an address list.
3150 Previously the only cases considered were a regular expression, or an
3153 20. Exim went into a mad DNS lookup loop when doing a callout where the
3154 host was specified on the transport, if the DNS lookup yielded more than
3157 21. The RFC2047 encoding function was originally intended for short strings
3158 such as real names; it was not keeping to the 75-character limit for
3159 encoded words that the RFC imposes. It now respects the limit, and
3160 generates multiple encoded words if necessary. To be on the safe side, I
3161 have increased the buffer size for the ${rfc2047: expansion operator from
3164 22. Failure to deliver a bounce message always caused it to be frozen, even if
3165 there was an errors_to setting on the router. The errors_to setting is now
3168 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3169 canonical form (fully expanded) before being placed in
3170 $sender_host_address.
3172 24. Updated eximstats to version 1.33
3174 25. Include certificate and key file names in error message when GnuTLS fails
3175 to set them up, because the GnuTLS error message doesn't include the name
3176 of the failing file when there is a problem reading it.
3178 26. Expand error message when OpenSSL has problems setting up cert/key files.
3181 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3182 (this can affect the format of dates).
3184 28. exim_tidydb, when checking for the continued existence of a message for
3185 which it has found a message-specific retry record, was not finding
3186 messages that were in split spool directories. Consequently, it was
3187 deleting retry records that should have stayed in existence.
3189 29. eximstats updated to version 1.35
3190 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3191 1.35 - bugfix such that pie charts by volume are generated correctly
3193 30. The SPA authentication driver was not abandoning authentication and moving
3194 on to the next authenticator when an expansion was forced to fail,
3195 contradicting the general specification for all authenticators. Instead it
3196 was generating a temporary error. It now behaves as specified.
3198 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3199 (the order specifies the preference for clients). The order is now AES256,
3200 AES128, 3DES, ARCFOUR128.
3202 31. Small patch to Sieve code - explicitly set From: when generating an
3205 32. Exim crashed if a remote delivery caused a very long error message to be
3206 recorded - for instance if somebody sent an entire SpamAssassin report back
3207 as a large number of 550 error lines. This bug was coincidentally fixed by
3208 increasing the size of one of Exim's internal buffers (big_buffer) that
3209 happened as part of the Exiscan merge. However, to be on the safe side, I
3210 have made the code more robust (and fixed the comments that describe what
3213 33. Some experimental protocols are using DNS PTR records for new purposes. The
3214 keys for these records are domain names, not reversed IP addresses. The
3215 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3216 leaves it alone. Component reversal etc. now happens only for IP addresses.
3219 34. The host_aton() function is supposed to be passed a string that is known
3220 to be a valid IP address. However, in the case of IPv6 addresses, it was
3221 not checking this. This is a hostage to fortune. Exim now panics and dies
3222 if the condition is not met. A case was found where this could be provoked
3223 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3224 components; fortuitously, this particular loophole had already been fixed
3225 by change 4.50/55 or 4.44/33 above.
3227 If there are any other similar loopholes, the new check in host_aton()
3228 itself should stop them being exploited. The report I received stated that
3229 data on the command line could provoke the exploit when Exim was running as
3230 exim, but did not say which command line option was involved. All I could
3231 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3232 running as the user.
3235 35. There was a buffer overflow vulnerability in the SPA authentication code
3236 (which came originally from the Samba project). I have added a test to the
3237 spa_base64_to_bits() function which I hope fixes it.
3240 36. The daemon start-up calls getloadavg() while still root for those OS that
3241 need the first call to be done as root, but it missed one case: when
3242 deliver_queue_load_max is set with deliver_drop_privilege. This is
3243 necessary for the benefit of the queue runner, because there is no re-exec
3244 when deliver_drop_privilege is set.
3246 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3247 was in use, was not putting the data itself into the right store pool;
3248 consequently, it could be overwritten for a subsequent message in the same
3249 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3252 38. Sometimes the final signoff response after QUIT could fail to get
3253 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3254 < 0 before doing a fflush(). This bug looks as though it goes back to the
3255 introduction of TLS in release 3.20, but "sometimes" must have been rare
3256 because the tests only now provoked it.
3262 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3263 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3264 Now it returns an integer. A number of calls were still expecting a T/F
3265 return. Fortuitously, in all cases, the tests worked in OK situations,
3266 which is the norm. However, things would have gone wrong on any write
3267 failures on the smtp file descriptor. This function is used when sending
3268 messages over SMTP and also when doing verify callouts.
3270 2. When Exim is called to do synchronous delivery of a locally submitted
3271 message (the -odf or -odi options), it no longer closes stderr before doing
3274 3. Implemented the mua_wrapper option.
3276 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3278 5. Implemented the functions header_remove(), header_testname(),
3279 header_add_at_position(), and receive_remove_recipient(), and exported them
3282 6. If an ACL "warn" statement specified the addition of headers, Exim already
3283 inserted X-ACL-Warn: at the start if there was no header name. However, it
3284 was not making this test for the second and subsequent header lines if
3285 there were newlines in the string. This meant that an invalid header could
3286 be inserted if Exim was badly configured.
3288 7. Allow an ACL "warn" statement to add header lines at the start or after all
3289 the Received: headers, as well as at the end.
3291 8. Added the rcpt_4xx retry error code.
3293 9. Added postmaster_mailfrom=xxx to callout verification option.
3295 10. Added mailfrom=xxxx to the callout verification option, for verify=
3298 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3299 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3301 12. Inserted some casts to stop certain compilers warning when using pointer
3302 differences as field lengths or precisions in printf-type calls (mostly
3303 affecting debugging statements).
3305 13. Added optional readline() support for -be (dynamically loaded).
3307 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3308 same clock tick as a message's arrival, so that its received time was the
3309 same as the "first fail" time on the retry record, and that message
3310 remained on the queue past the ultimate address timeout, every queue runner
3311 would try a delivery (because it was past the ultimate address timeout) but
3312 after another failure, the ultimate address timeout, which should have then
3313 bounced the address, did not kick in. This was a "< instead of <=" error;
3314 in most cases the first failure would have been in the next clock tick
3315 after the received time, and all would be well.
3317 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3318 being recognized when the domain list was tested by the match_domain
3319 condition in an expansion string.
3321 16. Added the ${str2b64: operator.
3323 17. Exim was always calling setrlimit() to set a large limit for the number of
3324 processes, without checking whether the existing limit was already
3325 adequate. (It did check for the limit on file descriptors.) Furthermore,
3326 errors from getrlimit() and setrlimit() were being ignored. Now they are
3327 logged to the main and panic logs, but Exim does carry on, to try to do its
3328 job under whatever limits there are.
3330 18. Imported PCRE 5.0.
3332 19. Trivial typo in log message " temporarily refused connection" (the leading
3335 20. If the log selector return_path_on_delivery was set and an address was
3336 redirected to /dev/null, the delivery process crashed because it assumed
3337 that a return path would always be set for a "successful" delivery. In this
3338 case, the whole delivery is bypassed as an optimization, and therefore no
3341 21. Internal re-arrangement: the function for sending a challenge and reading
3342 a response while authentication was assuming a zero-terminated challenge
3343 string. It's now changed to take a pointer and a length, to allow for
3344 binary data in such strings.
3346 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3348 23. Exim was not respecting finduser_retries when seeking the login of the
3349 uid under which it was called; it was always trying 10 times. (The default
3350 setting of finduser_retries is zero.) Also, it was sleeping after the final
3351 failure, which is pointless.
3353 24. Implemented tls_on_connect_ports.
3355 25. Implemented acl_smtp_predata.
3357 26. If the domain in control=submission is set empty, Exim assumes that the
3358 authenticated id is a complete email address when it generates From: or
3359 Sender: header lines.
3361 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3362 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3363 chown and chgrp in /bin and hostname in /usr/bin.
3365 28. Exim was keeping the "process log" file open after each use, just as it
3366 does for the main log. This opens the possibility of it remaining open for
3367 long periods when the USR1 signal hits a daemon. Occasional processlog
3368 errors were reported, that could have been caused by this. Anyway, it seems
3369 much more sensible not to leave this file open at all, so that is what now
3372 29. The long-running daemon process does not normally write to the log once it
3373 has entered its main loop, and it closes the log before doing so. This is
3374 so that log files can straightforwardly be renamed and moved. However,
3375 there are a couple of unusual error situations where the daemon does write
3376 log entries, and I had neglected to close the log afterwards.
3378 30. The text of an SMTP error response that was received during a remote
3379 delivery was being truncated at 512 bytes. This is too short for some of
3380 the long messages that one sometimes sees. I've increased the limit to
3383 31. It is now possible to make retry rules that apply only when a message has a
3384 specific sender, in particular, an empty sender.
3386 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3387 it possible to be selective about when SMTP synchronization is enforced.
3389 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3391 32. Implemented hosts_connection_nolog.
3393 33. Added an ACL for QUIT.
3395 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3398 35. Added mailbox_size and mailbox_filecount to appendfile.
3400 36. Added control = no_multiline_responses to ACLs.
3402 37. There was a bug in the logic of the code that waits for the clock to tick
3403 in the case where the clock went backwards by a substantial amount such
3404 that the microsecond fraction of "now" was more than the microsecond
3405 fraction of "then" (but the whole seconds number was less).
3407 38. Added support for the libradius Radius client library this is found on
3408 FreeBSD (previously only the radiusclient library was supported).
3414 1. When certain lookups returned multiple values in the form name=value, the
3415 quoting of the values was not always being done properly. Specifically:
3416 (a) If the value started with a double quote, but contained no whitespace,
3418 (b) If the value contained whitespace other than a space character (i.e.
3419 tabs or newlines or carriage returns) it was not quoted.
3420 This fix has been applied to the mysql and pgsql lookups by writing a
3421 separate quoting function and calling it from the lookup code. The fix
3422 should probably also be applied to nisplus, ibase and oracle lookups, but
3423 since I cannot test any of those, I have not disturbed their existing code.
3425 2. A hit in the callout cache for a specific address caused a log line with no
3426 reason for rejecting RCPT. Now it says "Previous (cached) callout
3427 verification failure".
3429 3. There was an off-by-one bug in the queryprogram router. An over-long
3430 return line was truncated at 256 instead of 255 characters, thereby
3431 overflowing its buffer with the terminating zero. As well as fixing this, I
3432 have increased the buffer size to 1024 (and made a note to document this).
3434 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3435 when Exim is waiting for an SMTP response from a remote server, Exim
3436 restarts its select() call on the socket, thereby resetting its timeout.
3437 This is not a problem when such interrupts are rare. Somebody set up a cron
3438 job to run exiwhat every 2 minutes, which is less than the normal select()
3439 timeout (5 or 10 minutes). This meant that the select() timeout never
3440 kicked in because it was always reset. I have fixed this by comparing the
3441 time when an interrupt arrives with the time at the start of the first call
3442 to select(). If more time than the timeout has elapsed, the interrupt is
3443 treated as a timeout.
3445 5. Some internal re-factoring in preparation for the addition of Sieve
3446 extensions (by MH). In particular, the "personal" test is moved to a
3447 separate function, and given an option for scanning Cc: and Bcc: (which is
3448 not set for Exim filters).
3450 6. When Exim created an email address using the login of the caller as the
3451 local part (e.g. when creating a From: or Sender: header line), it was not
3452 quoting the local part when it contained special characters such as @.
3454 7. Installed new OpenBSD configuration files.
3456 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3457 try to make them clearer.
3459 9. Callout options, other than the timeout value, were being ignored when
3460 verifying sender addresses in header lines. For example, when using
3462 verify = header_sender/callout=no_cache
3464 the cache was (incorrectly) being used.
3466 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3467 only the Cygwin environment.
3469 11. When return_path_on_delivery was set as a log selector, if different remote
3470 addresses in the same message used different return paths and parallel
3471 remote delivery occurred, the wrong values would sometimes be logged.
3472 (Whenever a remote delivery process finished, the return path value from
3473 the most recently started remote delivery process was logged.)
3475 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3476 lines when AUTH and/or TLS are in use. This is the "received protocol"
3477 field. Exim used to use "asmtp" for authenticated SMTP, without any
3478 indication (in the protocol name) for TLS use. Now it follows the RFC and
3479 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3480 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3481 names appear in log lines as well as in Received: header lines.
3483 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3484 extensions, and comparison tests, and to fix some bugs.
3486 14. Changes to the "personal" filter test:
3488 (1) The test was buggy in that it was just doing the equivalent of
3489 "contains" tests on header lines. For example, if a user's address was
3490 anne@some.where, the "personal" test would incorrectly be true for
3492 To: susanne@some.where
3494 This test is now done by extracting each address from the header in turn,
3495 and checking the entire address. Other tests that are part of "personal"
3496 are now done using regular expressions (for example, to check local parts
3497 of addresses in From: header lines).
3499 (2) The list of non-personal local parts in From: addresses has been
3500 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3501 taken from the Sieve specification recommendations.
3503 (3) If the message contains any header line starting with "List-" it is
3504 treated as non-personal.
3506 (4) The test for "circular" in the Subject: header line has been removed
3507 because it now seems ill-conceived.
3509 15. Minor typos in src/EDITME comments corrected.
3511 16. Installed latest exipick from John Jetmore.
3513 17. If headers_add on a router specified a text string that was too long for
3514 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3515 of string_sprintf() is now avoided.
3517 18. $message_body_size was not set (it was always zero) when running the DATA
3518 ACL and the local_scan() function.
3520 19. For the "mail" command in an Exim filter, no default was being set for
3521 the once_repeat time, causing a random time value to be used if "once" was
3522 specified. (If the value happened to be <= 0, no repeat happened.) The
3523 default is now 0s, meaning "never repeat". The "vacation" command was OK
3524 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3525 (I found it when inspecting the code).
3527 20. There is now an overall timeout for performing a callout verification. It
3528 defaults to 4 times the callout timeout, which applies to individual SMTP
3529 commands during the callout. The overall timeout applies when there is more
3530 than one host that can be tried. The timeout is checked before trying the
3531 next host. This prevents very long delays if there are a large number of
3532 hosts and all are timing out (e.g. when the network connections are timing
3533 out). The value of the overall timeout can be changed by specifying an
3534 additional sub-option for "callout", called "maxwait". For example:
3536 verify = sender/callout=5s,maxwait=20s
3538 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3539 to the end before writing, but this should make it even safer).
3541 22. Exim was forgetting that it had advertised PIPELINING for the second and
3542 subsequent messages on an SMTP connection. It was also not resetting its
3543 memory on STARTTLS and an internal HELO.
3545 23. When Exim logs an SMTP synchronization error within a session, it now
3546 records whether PIPELINING has been advertised or not.
3548 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3549 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3550 rather than long int.
3552 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3554 26. Added the never_mail option to autoreply.
3560 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3561 crash if the getsockname() call failed; this can happen if a connection is
3562 closed very soon after it is established. The problem was simply in the
3563 order in which certain operations were done, causing Exim to try to write
3564 to the SMTP stream before it had set up the file descriptor. The bug has
3565 been fixed by making things happen in the correct order.
3571 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3572 before the connection was closed, thus losing the rejection response.
3574 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3575 some early Solaris releases, but causes trouble in current releases where
3576 socklen_t is defined.
3578 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3581 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3584 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3585 that releasing the top of it at the end releases what was used for sub-
3586 expansions (unless the block got too big). However, discard this block if
3587 the first thing is a variable or header, so that we can use its block when
3588 it is dynamic (useful for very large $message_headers, for example).
3590 6. Lookups now cache *every* query, not just the most recent. A new, separate
3591 store pool is used for this. It can be recovered when all lookup caches are
3592 flushed. Lookups now release memory at the end of their result strings.
3593 This has involved some general refactoring of the lookup sources.
3595 7. Some code has been added to the store_xxx() functions to reduce the amount
3596 of flapping under certain conditions.
3598 8. log_incoming_interface used to affect only the <= reception log lines. Now
3599 it causes the local interface and port to be added to several more SMTP log
3600 lines, for example "SMTP connection from", and rejection lines.
3602 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3604 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3606 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3607 could be overwritten at the end of the current message (or the start of a
3608 new message if it was set in a HELO ACL). The value is now preserved for
3609 the duration of the SMTP connection.
3611 12. If a transport had a headers_rewrite setting, and a matching header line
3612 contained an unqualified address, that address was qualified, even if it
3613 did not match any rewriting rules. The underlying bug was that the values
3614 of the flags that permit the existence of unqualified sender and recipient
3615 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3616 non-local messages, and by -bnq for local messages) were not being
3617 preserved with the message after it was received.
3619 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3620 "next input=" as part of the text comprising the host identity instead of
3621 the correct text. The code was using the same buffer for two different
3622 strings. However, depending on which order the printing function evaluated
3623 its arguments, the bug did not always show up. Under Linux, for example, my
3624 test suite worked just fine.
3626 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3627 doesn't work with some older versions of Perl. It has been changed to "my",
3628 which in any case is probably the better facility to use.
3630 15. A really picky compiler found some instances of statements for creating
3631 error messages that either had too many or two few arguments for the format
3634 16. The size of the buffer for calls to the DNS resolver has been increased
3635 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3636 for addresses that have a lot of PTR records. This alleviates a problem; it
3637 does not fully solve it.
3639 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3640 buffer now truncates the list and logs the incident, which is the same
3641 action as happens when Exim is looking up a host name and its aliases.
3642 Previously in this situation something unpredictable would happen;
3643 sometimes it was "internal error: store_reset failed".
3645 18. If a server dropped the connection unexpectedly when an Exim client was
3646 using GnuTLS and trying to read a response, the client delivery process
3647 crashed while trying to generate an error log message.
3649 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3650 string, for example:
3652 warn message = H1: something\nH2: something
3654 the text was added as a single header line from Exim's point of view
3655 though it ended up OK in the delivered message. However, searching for the
3656 second and subsequent header lines using $h_h2: did not work. This has been
3657 fixed. Similarly, if a system filter added multiple headers in this way,
3658 the routers could not see them.
3660 20. Expanded the error message when iplsearch is called with an invalid key to
3661 suggest using net-iplsearch in a host list.
3663 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3664 ACLs are no longer actually imposed (and a message to that effect is
3667 22. If a "gecos" field in a passwd entry contained escaped characters, in
3668 particular, if it contained a \" sequence, Exim got it wrong when building
3669 a From: or a Sender: header from that name. A second bug also caused
3670 incorrect handling when an unquoted " was present following a character
3671 that needed quoting.
3673 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3674 was not being matched caselessly.
3676 24. Arranged for all hyphens in the exim.8 source to be escaped with
3679 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3680 properly. Recipient callout cache records were still being keyed to include
3681 the sender, even when use_sender was set false. This led to far more
3682 callouts that were necessary. The sender is no longer included in the key
3683 when use_sender is false.
3685 26. Added "control = submission" modifier to ACLs.
3687 27. Added the ${base62d: operator to decode base 62 numbers.
3689 28. dnsdb lookups can now access SRV records.
3691 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3692 the configuration file.
3694 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3695 (-v) mode. This makes the output for a verbose queue run more intelligible.
3697 31. Added a use_postmaster feature to recipient callouts.
3699 32. Added the $body_zerocount variable, containing the number of binary zero
3700 bytes in the message body.
3702 33. The time of last modification of the "new" subdirectory is now used as the
3703 "mailbox time last read" when there is a quota error for a maildir
3706 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3708 35. Added +ignore_unknown as a special item in host lists.
3710 36. Code for decoding IPv6 addresses in host lists is now included, even if
3711 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3712 address was recognized as an IP address, but was then not correctly decoded
3713 into binary, causing unexpected and incorrect effects when compared with
3720 1. Very minor rewording of debugging text in manualroute to say "list of
3721 hosts" instead of "hostlist".
3723 2. If verify=header_syntax was set, and a header line with an unqualified
3724 address (no domain) and a large number of spaces between the end of the
3725 name and the colon was received, the reception process suffered a buffer
3726 overflow, and (when I tested it) crashed. This was caused by some obsolete
3727 code that should have been removed. The fix is to remove it!
3729 3. When running in the test harness, delay a bit after writing a bounce
3730 message to get a bit more predictability in the log output.
3732 4. Added a call to search_tidyup() just before forking a reception process. In
3733 theory, someone could use a lookup in the expansion of smtp_accept_max_
3734 per_host which, without the tidyup, could leave open a database connection.
3736 5. Added the variables $recipient_data and $sender_data which get set from a
3737 lookup success in an ACL "recipients" or "senders" condition, or a router
3738 "senders" option, similar to $domain_data and $local_part_data.
3740 6. Moved the writing of debug_print from before to after the "senders" test
3743 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3744 problems for message scanning, either using a data ACL, or using
3745 local_scan() because the Received: header was not generated till after they
3746 were called (in order to set the time as the time of reception completion).
3747 I have revised the way this works. The header is now generated after the
3748 body is received, but before the ACL or local_scan() are called. After they
3749 are run, the timestamp in the header is updated.
3755 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3756 before starting a queue runner without re-exec. This happened only when
3757 deliver_drop_privilege was set or when the Exim user was set to root. The
3758 effect of the bug was that timeouts during subsequent deliveries caused
3759 crashes instead of being properly handled. The handler is now left at its
3760 default (and expected) setting.
3762 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3763 message, again when deliver_drop_privilege is set or Exim is run as root.
3764 The bug described in (1) was not present in this case, but the tidying up
3765 of the other signals was missing. I have made the two cases consistent.
3767 3. The ignore_target_hosts setting on a manualroute router was being ignored
3768 for hosts that were looked up using the /MX notation.
3770 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3773 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3774 operated on the sender address. After changing the $sender_address to <>
3775 for the sender address verify, Exim was re-instated it as the original
3776 (before rewriting) address, but remembering that it had rewritten it, so it
3777 wasn't rewriting it again. This bug also had the effect of breaking the
3778 sender address verification caching when the sender address was rewritten.
3780 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3781 This has been changed so that if the ip literal address matches
3782 ignore_target_hosts, the router declines.
3784 7. Added expansion conditions match_domain, match_address, and match_local_
3785 part (NOT match_host).
3787 8. The placeholder for the Received: header didn't have a length field set.
3789 9. Added code to Exim itself and to exim_lock to test for a specific race
3790 condition that could lead to file corruption when using MBX delivery. The
3791 issue is with the lockfile that is created in /tmp. If this file is removed
3792 after a process has opened it but before that process has acquired a lock,
3793 there is the potential for a second process to recreate the file and also
3794 acquire a lock. This could lead to two Exim processes writing to the file
3795 at the same time. The added code performs the same test as UW imapd; it
3796 checks after acquiring the lock that its file descriptor still refers to
3797 the same named file.
3799 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3800 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3803 11. Added the smtp_active_hostname option. If used, this will typically be made
3804 to depend on the incoming interface address. Because $interface_address is
3805 not set up until the daemon has forked a reception process, error responses
3806 that can happen earlier (such as "too many connections") no longer contain
3809 12. If an expansion in a condition on a "warn" statement fails because a lookup
3810 defers, the "warn" statement is abandoned, and the next ACL statement is
3811 processed. Previously this caused the whole ACL to be aborted.
3813 13. Added the iplsearch lookup type.
3815 14. Added ident_timeout as a log selector.
3817 15. Added tls_certificate_verified as a log selector.
3819 16. Added a global option tls_require_ciphers (compare the smtp transport
3820 option of the same name). This controls incoming TLS connections.
3822 17. I finally figured out how to make tls_require_ciphers do a similar thing
3823 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3824 before starting the TLS session.
3826 18. Tabs are now shown as \t in -bP output.
3828 19. If the log selector return_path_on_delivery was set, Exim crashed when
3829 bouncing a message because it had too many Received: header lines.
3831 20. If two routers both had headers_remove settings, and the first one included
3832 a superfluous trailing colon, the final name in the first list and the
3833 first name in the second list were incorrectly joined into one item (with a
3834 colon in the middle).
3840 1. Added -C and -D options to the exinext utility, mainly to make it easier
3841 to include in the automated testing, but these could be helpful when
3842 multiple configurations are in use.
3844 2. The exinext utility was not formatting the output nicely when there was
3845 an alternate port involved in the retry record key, nor when there was a
3846 message id as well (for retries that were specific to a specific message
3847 and a specific host). It was also confused by IPv6 addresses, because of
3848 the additional colons they contain. I have fixed the IPv4 problem, and
3849 patched it up to do a reasonable job for IPv6.
3851 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3852 delivery, the log line now contains "pipelined" if PIPELINING was used.
3854 4. An SMTP transport process used to panic and die if the bind() call to set
3855 an explicit outgoing interface failed. This has been changed; it is now
3856 treated in the same way as a connect() failure.
3858 5. A reference to $sender_host_name in the part of a conditional expansion
3859 that was being skipped was still causing a DNS lookup. This no longer
3862 6. The def: expansion condition was not recognizing references to header lines
3863 that used bh_ and bheader_.
3865 7. Added the _cache feature to named lists.
3867 8. The code for checking quota_filecount in the appendfile transport was
3868 allowing one more file than it should have been.
3870 9. For compatibility with Sendmail, the command line option
3878 and sets the incoming protocol and host name (for trusted callers). The
3879 host name and its colon can be omitted when only the protocol is to be set.
3880 Note the Exim already has two private options, -pd and -ps, that refer to
3881 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3882 "s", but I don't think that's a major issue.
3884 10. A number of refactoring changes to the code, none of which should affect
3887 (a) The number of logging options was getting close to filling up the
3888 32-bit word that was used as a bit map. I have split them into two classes:
3889 those that are passed in the argument to log_write(), and those that are
3890 only ever tested independently outside of that function. These are now in
3891 separate 32-bit words, so there is plenty of room for expansion again.
3892 There is no change in the user interface or the logging behaviour.
3894 (b) When building, for example, log lines, the code previously used a
3895 macro that called string_cat() twice, in order to add two strings. This is
3896 not really sufficiently general. Furthermore, there was one instance where
3897 it was actually wrong because one of the argument was used twice, and in
3898 one call a function was used. (As it happened, calling the function twice
3899 did not affect the overall behaviour.) The macro has been replaced by a
3900 function that can join an arbitrary number of extra strings onto a growing
3903 (c) The code for expansion conditions now uses a table and a binary chop
3904 instead of a serial search (which was left over from when there were very
3905 few conditions). Also, it now recognizes conditions like "pam" even when
3906 the relevant support is not compiled in: a suitably worded error message is
3907 given if an attempt is made to use such a condition.
3909 11. Added ${time_interval:xxxxx}.
3911 12. A bug was causing one of the ddress fields not to be passed back correctly
3912 from remote delivery subprocesses. The field in question was not being
3913 subsequently used, so this caused to problems in practice.
3915 13. Added new log selectors queue_time and deliver_time.
3917 14. Might have fixed a bug in maildirsizefile handling that threw up
3918 "unexpected character" debug warnings, and recalculated the data
3919 unnecessarily. In any case, I expanded the warning message to give more
3922 15. Added the message "Restricted characters in address" to the statements in
3923 the default ACL that block characters like @ and % in local parts.
3925 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3926 Three changes have been made:
3928 (a) There was a serious bug; a negative response to MAIL caused the whole
3929 recipient domain to be cached as invalid, thereby blocking all messages
3930 to all local parts at the same domain, from all senders. This bug has
3931 been fixed. The domain is no longer cached after a negative response to
3932 MAIL if the sender used is not empty.
3934 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3937 (c) A new callout option, "use_sender" has been added for people who want
3938 the modified behaviour.
3944 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3947 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3948 indeed breaks things for older releases.
3950 3. Added additional logging to the case where there is a problem reading data
3951 from a filter that is running in a subprocess using a pipe, in order to
3952 try to track down a specific problem.
3954 4. Testing facility fudge: when running in the test harness and attempting
3955 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3956 getting "No route to host". Convert this to a timeout.
3958 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3961 6. Some OS don't have socklen_t but use size_t instead. This affects the
3962 fifth argument of getsockopt() amongst other things. This is now
3963 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3964 can be set for individual OS. I have set it for SunOS5, OSF1, and
3965 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3966 some earlier ones do not.
3968 7. Change 4.30/15 was not doing the test caselessly.
3970 8. The standard form for an IPv6 address literal was being rejected by address
3971 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3972 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3973 this, as well as the form without the "IPv6" on the front (but only when
3974 address literals are enabled, of course).
3976 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3978 10. Exim crashed if a message with an empty sender address specified by -f
3979 encountered a router with an errors_to setting. This could be provoked only
3980 by a command such as
3984 where an empty string was supplied; "<>" did not hit this bug.
3986 11. Installed PCRE release 4.5.
3988 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3989 remained set. It is now erased.
3991 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3992 times from message ids (which are base 36 rather than the normal 62).
3994 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3995 were being counted as actual protocol errors, and logged if the log
3996 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3997 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3998 and DATA following a set of rejected RCPTs do not count as protocol errors.
3999 In other words, Exim assumes they were pipelined, though this may not
4000 actually be the case. Of course, in all cases the client gets an
4001 appropriate error code.
4003 15. If a lookup fails in an ACL condition, a message about the failure may
4004 be available; it is used if testing the ACL cannot continue, because most
4005 such messages specify what the cause of the deferral is. However, some
4006 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4007 that caused an old message to be retained and used if a later statement
4008 caused a defer, replacing the real cause of the deferral.
4010 16. If an IP address had so many PTR records that the DNS lookup buffer
4011 was not large enough to hold them, Exim could crash while trying to process
4012 the truncated data. It now detects and logs this case.
4014 17. Further to 4.21/58, another change has been made: if (and only if) the
4015 first line of a message (the first header line) ends with CRLF, a bare LF
4016 in a subsequent header line has a space inserted after it, so as not to
4017 terminate the header.
4019 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4020 unnecessarily, used atoi() instead of strtol(), and didn't check the
4021 termination when getting file sizes from file names by regex.
4023 19. Completely re-implemented the support for maildirsize files, in the light
4024 of a number of problems with the previous contributed implementation
4025 (4.30/29). In particular:
4027 . If the quota is zero, the maildirsize file is maintained, but no quota is
4030 . If the maildir directory does not exist, it is created before any attempt
4031 to write a maildirsize file.
4033 . The quota value in the file is just a cache; if the quota is changed in
4034 the transport, the new value overrides.
4036 . A regular expression is available for excluding directories from the
4039 20. The autoreply transport checks the characters in options that define the
4040 message's headers; it allows continued headers, but it was checking with
4041 isspace() after an embedded newline instead of explicitly looking for a
4044 21. If all the "regular" hosts to which an address was routed had passed their
4045 expiry times, and had not reached their retry times, the address was
4046 bounced, even if fallback hosts were defined. Now Exim should go on to try
4049 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4050 equivalent code in the SMTP transport. Some hosts send humungous responses
4051 to HELO/EHLO, more than 1024 it seems.
4053 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4054 gives compiler warnings in some environments. I've now done it "properly",
4057 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4058 (because of problems with the built-in one) was declared to return uschar *
4059 instead of char *, causing compiler failure.
4061 25. Fixed a file descriptor leak when processing alias/forward files.
4063 26. Fixed a minor format string issue in dbfn.c.
4065 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4067 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4068 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4071 29. When a sender address is verified, it is cached, to save repeating the test
4072 when there is more than one recipient in a message. However, when the
4073 verification involves a callout, it is possible for different callout
4074 options to be set for different recipients. It is too complicated to keep
4075 track of this in the cache, so now Exim always runs a verification when a
4076 callout is required, relying on the callout cache for the optimization.
4077 The overhead is duplication of the address routing, but this should not be
4080 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4081 to be verified with callout=postmaster, and the main callout worked but the
4082 postmaster check failed, the verification correctly failed. However, if a
4083 subsequent RCPT command asked for sender verification *without* the
4084 postmaster check, incorrect caching caused this verification also to fail,
4087 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4088 it was not caching the DNS options (qualify_single, search_parents) that
4089 were used when the lookup failed. A subsequent lookup with different
4090 options therefore always gave the same answer, though there were cases
4091 where it should not have. (Example: a "domains = !$mx_any" option on a
4092 dnslookup router: the "domains" option is always processed without any
4093 widening, but the router might have qualify_single set.) Now Exim uses the
4094 cached value only when the same options are set.
4096 32. Added John Jetmore's "exipick" utility to the distribution.
4098 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4099 than a timeout (e.g. a certificate is required, and is not provided), an
4100 Exim server now closes the connection immediately. Previously it waited for
4101 the client to close - but if the client is SSL, it seems that they each
4102 wait for each other, leading to a delay before one of them times out.
4104 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4105 maintained 0.8.x compatibility because I don't think many are using it, and
4106 it is clearly obsolete.
4108 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4111 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4112 client certificate was expired. A simple patch fixes this, though I don't
4113 understand the full logic of why the verify callback is called multiple
4116 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4117 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4118 which causes problems with some clients (such as the Certicom SSL Plus
4119 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4120 disables the coutermeasure allowing Eudora to connect."
4122 38. Exim was not checking that a write() to a log file succeeded. This could
4123 lead to Bad Things if a log got too big, in particular if it hit a file
4124 size limit. Exim now panics and dies if it cannot write to a log file, just
4125 as it does if it cannot open a log file.
4127 39. Modified OS/Makefile-Linux so that it now contains
4129 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4131 The two -D definitions ensure that Exim is compiled with large file
4132 support, which makes it possible to handle log files that are bigger than
4135 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4136 instance) a domain was checked against a named list that involved a lookup,
4137 causing $domain_data to be set, then another domain was checked against the
4138 same list, then the first domain was re-checked, the value of $domain_data
4139 after the final check could be wrong. In particular, if the second check
4140 failed, it could be set empty. This bug probably also applied to
4143 41. The strip_trailing_dot option was not being applied to the address given
4144 with the -f command-line option.
4146 42. The code for reading a message's header from the spool was incrementing
4147 $received_count, but never initializing it. This meant that the value was
4148 incorrect (doubled) while delivering a message in the same process in which
4149 it was received. In the most common configuration of Exim, this never
4150 happens - a fresh exec is done - but it can happen when
4151 deliver_drop_privilege is set.
4153 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4154 it now includes up to 150 characters of the unexpected data in the log
4157 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4158 and building data strings. The size of both of these buffers was 10 000
4159 bytes - far larger than anybody would *ever* want, thought I. Needless to
4160 say, somebody hit the limit. I have increased the maximum line length to
4161 20 000 and the maximum data length of concatenated lines to 100 000. I have
4162 also fixed two bugs, because there was no checking on these buffers. Tsk,
4163 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4164 buffer is too small.
4166 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4167 lsearch lookups. Now it does.
4169 46. When parsing a route_list item in a manualroute router, a fixed-length
4170 buffer was used for the list of hosts. I made this 1024 bytes long,
4171 thinking that nobody would ever have a list of hosts that long. Wrong.
4172 Somebody had a whole pile of complicated expansion conditions, and the
4173 string was silently truncated, leading to an expansion error. It turns out
4174 that it is easier to change to an unlimited length (owing to other changes
4175 that have happened since this code was originally written) than to build
4176 structure for giving a limitation error. The length of the item that
4177 expands into the list of hosts is now unlimited.
4179 47. The lsearch lookup could not handle data where the length of text line was
4180 more than 4095 characters. Such lines were truncated, leading to shortened
4181 data being returned. It should now handle lines of any length.
4183 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4184 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4187 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4188 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4189 no longer happens in comment lines. A list of replacements is now placed
4190 at the head of all of the source files, except those whose only change is
4191 to replace PERL_COMMAND in the very first #! line.
4193 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4194 messages on the queue, with a bottom-up merge sort, using code contributed
4195 by Michael Haardt. This should make operations like -bp somewhat faster on
4196 large queues. It won't affect queue runners, except when queue_run_in_order
4199 51. Installed eximstats 1.31 in the distribution.
4201 52. Added support for SRV lookups to the dnslookup router.
4203 53. If an ACL referred to $message_body or $message_body_end, the value was not
4204 reset for any messages that followed in the same SMTP session.
4206 54. The store-handling optimization for building very long strings was not
4207 differentiating between the different store pools. I don't think this
4208 actually made any difference in practice, but I've tidied it.
4210 55. While running the routers to verify a sender address, $sender_address
4211 was still set to the sender address. This is wrong, because when routing to
4212 send a bounce to the sender, it would be empty. Therefore, I have changed
4213 it so that, while verifying a sender address, $sender_address is set to <>.
4214 (There is no change to what happens when verifying a recipient address.)
4216 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4217 target A or AAAA records (if not already returned) without resetting the
4218 qualify_single or search_parents options of the DNS resolver. These are
4219 inappropriate in this case because the targets of MX and SRV records must
4220 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4221 target that, when qualified, matched something in the local domain. These
4222 two options are now turned off when doing these lookups.
4224 57. It seems that at least some releases of Reiserfs (which does not have the
4225 concept of a fixed number of inodes) returns zero and not -1 for the
4226 number of available inodes. This interacted badly with check_spool_inodes,
4227 which assumed that -1 was the "no such thing" setting. What I have done is
4228 to check that the total number of inodes is greater than zero before doing
4229 the test of how many are available.
4231 58. When a "warn" ACL statement has a log_message modifier, the message is
4232 remembered, and not repeated. This is to avoid a lot of repetition when a
4233 message has many recipients that cause the same warning to be written.
4234 Howewer, Exim was preserving the list of already written lines for an
4235 entire SMTP session, which doesn't seem right. The memory is now reset if a
4236 new message is started.
4238 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4239 debugging output unless log_rewrite was also set.
4241 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4242 of (int)(handle) when we know that handle contains (void *)(-1).
4244 61. The Exim daemon panic-logs an error return when it closes the incoming
4245 connection. However "connection reset by peer" seems to be common, and
4246 isn't really an error worthy of noting specially, so that particular error
4249 62. When Exim is trying to find all the local interfaces, it used to panic and
4250 die if the ioctl to get the interface flags failed. However, it seems that
4251 on at least one OS (Solaris 9) it is possible to have an interface that is
4252 included in the list of interfaces, but for which you get a failure error
4253 for this call. This happens when the interface is not "plumbed" into a
4254 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4255 failure of the "get flags" call assumes that the interface is down.
4257 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4258 makes life easier for people who are doing arithmetic on fields extracted
4259 from dates, where you often get leading zeros that should not be
4260 interpreted as octal.
4262 64. Added qualify_domain to the redirect router, to override the global
4265 65. If a pathologically long header line contained very many addresses (the
4266 report of this problem mentioned 10 000) and each of them was rewritten,
4267 Exim could use up a very large amount of memory. (It kept on making new
4268 copies of the header line as it rewrote, and never released the old ones.)
4269 At the expense of a bit more processing, the header rewriting function has
4270 been changed so that it no longer eats memory in this way.
4272 66. The generation of the Received: header has been moved from the time that a
4273 message starts to be received, to the time that it finishes. The timestamp
4274 in the Received: header should now be very close to that of the <= log
4275 line. There are two side-effects of this change:
4277 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4278 logged header lines no longer include the local Received: line, because
4279 it has not yet been created. The same applies to a copy of the message
4280 that is returned to a non-SMTP sender when a message is rejected.
4282 (b) When a filter file is tested using -bf, no additional Received: header
4283 is added to the test message. After some thought, I decided that this
4286 This change does not affect the value of $received_for. It is still set
4287 after address rewriting, but before local_scan() is called.
4289 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4291 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4292 gave an unhelpful panic error message, and a defer error. I have managed to
4293 change this behaviour so that it now rejects any supplied certificate,
4294 which seems right, as the list of acceptable certificates is empty.
4296 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4297 gave an unhelpful defer error. I have not managed to make this reject any
4298 supplied certificates, but the error message it gives is "no certificate
4299 supplied", which is not helpful.
4301 70. exigrep's output now also includes lines that are not associated with any
4302 message, but which match the given pattern. Implemented by a patch from
4303 Martin Sluka, which also tidied up the Perl a bit.
4305 71. Recipient callout verification, like sender verification, was using <> in
4306 the MAIL FROM command. This isn't really the right thing, since the actual
4307 sender may affect whether the remote host accepts the recipient or not. I
4308 have changed it to use the actual sender in the callout; this means that
4309 the cache record is now keyed on a recipient/sender pair, not just the
4310 recipient address. There doesn't seem to be a real danger of callout loops,
4311 since a callout by the remote host to check the sender would use <>.
4312 [SEE ABOVE: changed after hitting problems.]
4314 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4315 temporary errors. However, in the case of such a code being given after
4316 the end of a data transmission (i.e. after ".") Exim was failing to write
4317 a retry record for the message. (Yes, there was some broken host that was
4318 actually sending 8xx at this point.)
4320 73. An unknown lookup type in a host list could cause Exim to panic-die when
4321 the list was checked. (An example that provoked this was putting <; in the
4322 middle of a list instead of at the start.) If this happened during a DATA
4323 ACL check, a -D file could be left lying around. This kind of configuration
4324 error no longer causes Exim to die; instead it causes a defer errror. The
4325 incident is still logged to the main and panic logs.
4327 74. Buglet left over from Exim 3 conversion. The message "too many messages
4328 in one connection" was written to the rejectlog but not the mainlog, except
4329 when address rewriting (yes!) was being logged.
4331 75. Added write_rejectlog option.
4333 76. When a system filter was run not as root (that is, when system_filter_user
4334 was set), the values of the $n variables were not being returned to the
4335 main process; thus, they were not subsequently available in the $sn
4338 77. Added +return_path_on_delivery log selector.
4340 78. A connection timeout was being treated differently from recipients deferred
4341 when testing hosts_max_try with a message that was older than the host's
4342 retry timeout. (The host should not be counted, thus allowing all hosts to
4343 be tried at least once before bouncing.) This may have been the cause of an
4344 occasionally reported bug whereby a message would remain on the queue
4345 longer than the retry timeout, but would be bounced if a delivery was
4346 forced. I say "may" because I never totally pinned down the problem;
4347 setting up timeout/retry tests is difficult. See also the next item.
4349 79. The ultimate address timeout was not being applied to errors that involved
4350 a combination of host plus message (for example, a timeout on a MAIL
4351 command). When an address resolved to a number of possible hosts, and they
4352 were not all tried for each delivery (e.g. because of hosts_max_try), a
4353 message could remain on the queue longer than the retry timeout.
4355 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4358 81. Fixed an obscure SMTP outgoing bug which required at least the following
4359 conditions: (a) there was another message waiting for the same server;
4360 (b) the server returned 5xx to all RCPT commands in the first message so
4361 that the message was not completed; (c) the server dropped the connection
4362 or gave a negative response to the RSET that Exim sends to abort the
4363 transaction. The observed case was a dropped connection after DATA that had
4364 been sent in pipelining mode. That is, the server had advertised PIPELINING
4365 but was not implementing it correctly. The effect of the bug was incorrect
4366 behaviour, such as trying another host, and this could lead to a crash.
4372 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4373 and daemon.c were passed as pointers to ints; they should have been
4374 pointers to socklen_t variables (which are typically unsigned ints).
4376 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4379 3. Fixed a really odd bug that affected only the testing scheme; patching a
4380 certain fixed string in the binary changed the value of another string that
4381 happened to be identical to the end of the original first string.
4383 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4384 name", it returns that address as the IP address. On some operating
4385 systems (e.g. Solaris), it also passes back the IP address string as the
4386 "host name". However, on others (e.g. Linux), it passes back an empty
4387 string. Exim wasn't checking for this, and was changing the host name to an
4388 empty string, assuming it had been canonicized.
4390 5. Although rare, it is permitted to have more than one PTR record for a given
4391 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4392 all the names associated with an address, because they do in Solaris.
4393 However, it seems that they do not in Linux for data that comes from the
4394 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4395 I found this out when I moved to a new Linux workstation and tried to run
4396 the Exim test suite.
4398 To get round this problem I have changed the code so that it now does its
4399 own call to the DNS to look up PTR records when searching for a host name.
4400 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4401 addresses that are only in /etc/hosts are still found.
4403 This behaviour is, however, controlled by an option called host_lookup_
4404 order, which defaults to "bydns:byaddr". If people want to use the other
4405 order, or indeed, just use one or the other means of lookup, they can
4406 specify it in this variable.
4408 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4409 some operating systems, this comes back from gethostbyaddr() as an empty
4410 string, and this is what Exim used to test for. However, it seems that in
4411 other systems, "." is yielded. Exim now tests for this case too.
4413 7. The values of check_spool_space and check_log_space are now held internally
4414 as a number of kilobytes instead of an absolute number of bytes. If a
4415 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4416 kilobyte. This means that much larger values can be stored.
4418 8. Exim monitor: an attempt to get the action menu when not actually pointing
4419 at a message produces an empty menu entitled "No message selected". This
4420 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4421 no entries in it ("Shell widget menu has zero width and/or height"). So I
4422 have added a single, blank menu entry in this case.
4424 9. Added ${quote_local_part.
4426 10. MIME decoding is now applied to the contents of Subject: header lines when
4429 11. Now that a reference to $sender_host_address automatically causes a reverse
4430 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4431 host lookup before query-style lookups in lists that might use this
4432 variable. This has therefore been abolished, and the "net-" prefix is no
4433 longer necessary for query-style lookups.
4435 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4436 have been a typo for LFLAGS, so it has been changed.
4438 13. The install script calls Exim with "-C /dev/null" in order to find the
4439 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4440 to be output. Howeve, since Exim outputs its version number before the
4441 error, it didn't break the script. It just looked ugly. I fixed this by
4442 always allowing "-C /dev/null" if the caller is root.
4444 14. Ignore overlarge ACL variable number when reading spool file - insurance
4445 against a later release with more variables having written the file.
4447 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4448 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4449 this, as well as the form without the "IPv6" on the front.
4451 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4452 OS/Makefile-Darwin file.
4454 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4455 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4457 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4458 variable. However, some OS use ints for inodes. I've added cast to long int
4459 to get rid of the compiler warning.
4461 19. I had forgotten to lock out "/../" in configuration file names when
4462 ALT_CONFIG_PREFIX was set.
4464 20. Routers used for verification do not need to specify transports. However,
4465 if such a router generated a host list, and callout was configured, Exim
4466 crashed, because it could not find a port number from the (non-existent)
4467 transport. It now assumes port 25 in this circumstance.
4469 21. Added the -t option to exigrep.
4471 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4472 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4473 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4474 them is set, LOOKUP_LSEARCH is forced.
4476 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4478 24. Added sender and host information to the "rejected by local_scan()" log
4479 line; previously there was no indication of these.
4481 25. Added .include_if_exists.
4483 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4484 Linux. It turns out that not all file systems support this. Apparently some
4485 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4486 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4487 not supported on the file descriptor, is now ignored when Exim is trying to
4488 sync a directory. This applies only to Linux.
4490 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4492 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4493 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4496 29. Added support for maildirsize files from supplied patch (modified a bit).
4498 30. The use of :fail: followed by an empty string could lead Exim to respond to
4499 sender verification failures with (e.g.):
4501 550 Verification failed for <xxx>
4502 550 Sender verify failed
4504 where the first response line was missing the '-' that indicates it is not
4505 the final line of the response.
4507 31. The loop for finding the name of the user that called Exim had a hardwired
4508 limit of 10; it now uses the value of finduser_retries, which is used for
4509 all other user lookups.
4511 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4514 33. Exim was neglecting to zero errno before one call of strtol() when
4515 expanding a string and expecting an integer value. On some systems this
4516 resulted in spurious "integer overflow" errors. Also, it was casting the
4517 result into an int without checking.
4519 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4520 did not work. The code looks as if it has *never* worked, though it appears
4521 to have been documented since at least releast 1.62. I have made it work.
4523 35. The "timeout_DNS" error in retry rules, also documented since at least
4524 1.62, also never worked. As it isn't clear exactly what this means, and
4525 clearly it isn't a major issue, I have abolished the feature by treating it
4526 as "timeout", and writing a warning to the main and panic logs.
4528 36. The display of retry rules for -brt wasn't always showing the error code
4531 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4532 timeout_connect_A, timeout_connect_MX.
4534 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4535 to the empty sender.
4537 39. The daemon was not analysing the content of -oX till after it had closed
4538 stderr and disconnected from the controlling terminal. This meant that any
4539 syntax errors were only noted on the panic log, and the return code from
4540 the command was 0. By re-arranging the code a little, I've made the
4541 decoding happen first, so such errors now appear on stderr, and the return
4542 code is 1. However, the actual setting up of the sockets still happens in
4543 the disconnected process, so errors there are still only recorded on the
4546 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4547 connections (as happens on some IP stacks) was logged at start up time as
4548 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4549 it from "IPv6 and IPv4", which means that two separate sockets are being
4552 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4553 says whether AF_INET or AF_INET6 was passed as an argument.
4555 42. Exiwhat output was messed up when time zones were included in log
4558 43. Exiwhat now gives more information about the daemon's listening ports,
4559 and whether -tls-on-connect was used.
4561 44. The "port" option of the smtp transport is now expanded.
4563 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4564 silently ignored. Now an error message is written to the main and panic
4567 46. There's a new ACL modifier called "logwrite" which writes to a log file
4568 as soon as it is encountered.
4570 47. Added $local_user_uid and $local_user_gid at routing time.
4572 48. Exim crashed when trying to verify a sender address that was being
4575 49. Exim was recognizing only a space character after ".include". It now also
4576 recognizes a tab character.
4578 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4579 extracting the relevant information from the specification. The man page no
4580 longer contains scrambled data for the -d option, and I've added a section
4581 at the front about calling Exim under different names.
4583 51. Added "extra_headers" argument to the "mail" command in filter files.
4585 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4588 53. Installed eximstats 1.29.
4590 54. Added transport_filter_timeout as a generic transport option.
4592 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4593 Cc: header lines. This was required by RFC 822, but it not required by RFC
4596 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4597 incoming messages that did not have them. Now it does so only if the
4598 message originates locally, that is, if there is no associated remote host
4599 address. When Resent- header lines are present, this applies to the Resent-
4600 lines rather than the non-Resent- lines.
4602 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4603 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4605 58. Messages for configuration errors now include the name of the main
4606 configuration file - useful now that there may be more than one file in a
4607 list (.included file names were always shown).
4609 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4610 for those rare installations that do not start the daemon as root or run it
4611 setuid root. I've cut out the call to initgroups() if the daemon is not
4614 60. The Exim user and group can now be bound into the binary as text strings
4615 that are looked up at the start of Exim's processing.
4617 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4619 62. Added $mailstore_basename variable.
4621 63. Installed patch to sieve.c from Michael Haardt.
4623 64. When Exim failed to open the panic log after failing to open the main log,
4624 the original message it was trying to log was written to stderr and debug
4625 output, but if they were not available (the usual case in production), it
4626 was lost. Now it is written to syslog before the two lines that record the
4627 failures to open the logs.
4629 65. Users' Exim filters run in subprocesses under the user's uid. It is
4630 possible for a "deliver" command or an alias in a "personal" command to
4631 provoke an address rewrite. If logging of address rewriting is configured,
4632 this fails because the process is not running as root or exim. There may be
4633 a better way of dealing with this, but for the moment (because 4.30 needs
4634 to be released), I have disabled address rewrite logging when running a
4635 filter in a non-root, non-exim process.
4641 1. The buildconfig auxiliary program wasn't quoting the value set for
4642 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4643 not defined. This bug was masked in 4.22 by the effect that was fixed in
4646 2. Some messages that were rejected after a message id was allocated were
4647 shown as "incomplete" by exigrep. It no longer does this for messages that
4648 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4650 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4651 have allow_domain_literals set, the ID did not get logged in the <= line.
4652 Domain literals are now always recognized in Message-ID: header lines.
4654 4. The first argument for a ${extract expansion item is the key name or field
4655 number. Leading and trailing spaces in this item were not being ignored,
4656 causing some misleading effects.
4658 5. When deliver_drop_privilege was set, single queue runner processes started
4659 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4660 same command in the process it spins off) were not dropping privilege.
4662 6. When the daemon running as "exim" started a queue runner, it always
4663 re-executed Exim in the spun-off process. This is a waste of effort when
4664 deliver_drop_privilege is set. The new process now just calls the
4665 queue-runner function directly.
4671 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4674 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4675 ignored. Though the use of -oP was forcing the writing of a pid file, it
4676 was always written to the default place.
4678 3. If the message "no IP address found for host xxxx" is generated during
4679 incoming verification, it is now followed by identification of the incoming
4680 connection (so you can more easily find what provoked it).
4682 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4684 5. Added some features to "harden" Exim a bit more against certain attacks:
4686 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4687 be put in Local/Makefile. This is like the never_users runtime option,
4688 but it cannot be overridden. The default setting is "root".
4690 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4691 prefix string with which any file named in a -C command line option
4694 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4695 is retained for -C and -D only if the caller of Exim is root. Without
4696 it, the exim user may also use -C and -D and retain privilege.
4698 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4699 command line option is disabled.
4701 6. Macro names set by the -D option must start with an upper case letter, just
4702 like macro names defined in the configuration file.
4704 7. Added "dereference=" facility to LDAP.
4706 8. Two instances of the typo "uknown" in the source files are fixed.
4708 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4709 the Configure-Makefile script screwed up while processing it.
4711 10. Incorporated PCRE 4.4.
4713 11. The SMTP synchronization check was not operating right at the start of an
4714 SMTP session. For example, it could not catch a HELO sent before the client
4715 waited for the greeting. There is now a check for outstanding input at the
4716 point when the greeting is written. Because of the duplex, asynchronous
4717 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4718 way, but not yet received, when the check is performed.
4720 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4721 on TCP/IP sockets, because this apparently causes some broken clients to
4724 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4725 unchanged) from the Cygwin maintainer.
4727 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4728 was supported instead of testing for mbx. Effectively a typo.
4730 15. The spa authenticator server code was not checking that the input it
4731 received was valid base64.
4733 16. The debug output line for the "set" modifier in ACLs was not showing the
4734 name of the variable that was being set.
4736 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4738 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4739 lookup on its own. Something else was needed to trigger the lookup. For
4740 example, a match in host_lookup or the need for a host name in a host list.
4741 Now, if $sender_host_name is referenced and the host name has not yet been
4742 looked up, a lookup is performed. If the lookup fails, the variable remains
4743 empty, and $host_lookup_failed is set to "1".
4745 19. Added "eqi" as a case-independent comparison operator.
4747 20. The saslauthd authentication condition could segfault if neither service
4748 nor realm was specified.
4750 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4751 error message that was logged was misleading, and incoming SMTP
4752 connections were dropped. The message is now more accurate, and temporary
4753 errors are given to SMTP connections.
4755 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4756 (except RSET) with a 421 error, until QUIT is received. However, it was
4757 failing to send a response to QUIT.
4759 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4760 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4761 ACL is now run after the helo_try_verify_hosts code.
4763 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4764 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4765 case-independent, so other case variants are also recognized). Apparently
4766 some systems use these upper case variants.
4768 25. If more than two messages were waiting for the same host, and a transport
4769 filter was specified for the transport, Exim sent two messages over the
4770 same TCP/IP connection, and then failed with "socket operation on non-
4771 socket" when it tried to send the third.
4773 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4775 27. The extern definition of crypt16() in expand.c was not being excluded when
4776 the OS had its own crypt16() function.
4778 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4779 as a preferred synonym for return_size_limit, both as an option and as an
4782 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4784 30. Changed the default configuration ACL to relax the local part checking rule
4785 for addresses that are not in any local domains. For these addresses,
4786 slashes and pipe symbols are allowed within local parts, but the sequence
4787 /../ is explicitly forbidden.
4789 31. SPA server authentication was not clearing the challenge buffer before
4792 32. log_message in a "warn" ACL statement was writing to the reject log as
4793 well as to the main log, which contradicts the documentation and doesn't
4794 seem right (because no rejection is happening). So I have stopped it.
4796 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4797 However, I am unable to do any testing of this.
4799 34. Fixed an infelicity in the appendfile transport. When checking directories
4800 for a mailbox, to see if any needed to be created, it was accidentally
4801 using path names with one or more superfluous leading slashes; tracing
4802 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4804 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4805 contents are added to the log line that is written for every discarded
4806 recipient. (Previously a log_message setting was ignored.)
4808 36. The ${quote: operator now quotes the string if it is empty.
4810 37. The install script runs exim in order to find its version number. If for
4811 some reason other than non-existence or emptiness, which it checks, it
4812 could not run './exim', it was installing it with an empty version number,
4813 i.e. as "exim-". This error state is now caught, and the installation is
4816 38. An argument was missing from the function that creates an error message
4817 when Exim fails to connect to the socket for saslauthd authentication.
4818 This could cause Exim to crash, or give a corrupted message.
4820 39. Added isip, isip4, and isip6 to ${if conditions.
4822 40. The ACL variables $acl_xx are now saved with the message, and can be
4823 accessed later in routers, transports, and filters.
4825 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4826 strings in the file are not string-expanded.
4828 42. If a MAIL command specified a SIZE value that was too large to fit into an
4829 int variable, the check against message_size_limit failed. Such values are
4830 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4831 day this will have to be increased, but I don't think I want to be around
4832 when emails are that large.
4839 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4840 iconv() is not standard in FreeBSD.
4842 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4843 IPv6 enabled. The observed symptom was a segmentation fault on return from
4844 the function os_common_find_running_interfaces() in src/os.c.
4846 3. In the check_special_case() function in daemon.c I had used "errno" as an
4847 argument name, which causes warnings on some systems. This was basically a
4848 typo, since it was named "eno" in the comments!
4850 4. The code that waits for the clock to tick (at a resolution of some fraction
4851 of a second) so as to ensure message-id uniqueness was always waiting for
4852 at least one whole tick, when it could have waited for less. [This is
4853 almost certainly not relevant at current processor speeds, where it is
4854 unlikely to ever wait at all. But we try to future-proof.]
4856 5. The function that sleeps for a time interval that includes fractions of a
4857 second contained a race. It did not block SIGALRM between setting the
4858 timer, and suspending (a couple of lines later). If the interval was short
4859 and the sigsuspend() was delayed until after it had expired, the suspension
4860 never ended. On busy systems this could lead to processes getting stuck for
4863 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4864 process, before it forks any delivery processes. The open lookup caching
4865 mechanism meant that the open file or database connection was passed into
4866 the delivery process. The problem was that delivery processes always tidy
4867 up cached lookup data. This could cause a problem for the next delivery
4868 process started by the queue runner, because the external queue runner
4869 process does not know about the closure. So the next delivery process
4870 still has data in the lookup cache. In the case of a file lookup, there was
4871 no problem because closing a file descriptor in a subprocess doesn't affect
4872 the parent. However, if the lookup was caching a connection to a database,
4873 the connection was closed, and the second delivery process was likely to
4874 see errors such as "PGSQL: query failed: server closed the connection
4875 unexpectedly". The problem has been fixed by closing all cached lookups
4876 in a queue runner before running a delivery process.
4878 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4879 seem to have the "const" qualifier which it has on other OS. I've
4882 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4883 *of the same type* (client or server) with the same public name that an
4884 error should be diagnosed.
4886 9. When Exim looked up a host name for an IP address, but failed to find the
4887 original IP address when looking up the host name (a safety check), it
4888 output the message "<ip address> does not match any IP for NULL", which was
4889 confusing, to say the least. The bug was that the host name should have
4890 appeared instead of "NULL".
4892 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4893 user that is built into the binary, and the -C or -D options is used, root
4894 privilege is dropped before the configuration file is read. In addition,
4895 logging is switched to stderr instead of the normal log files. If the
4896 configuration then re-defines the Exim user, the unprivileged environment
4897 is probably not what is expected, so Exim logs a panic warning message (but
4900 However, if deliver_drop_privilege is set, the unprivileged state may well
4901 be exactly what is intended, so the warning has been cut out in that case,
4902 and Exim is allowed to try to write to its normal log files.
4908 1. smtp_return_error_details was not giving details for temporary sender
4909 or receiver verification errors.
4911 2. Diagnose a configuration error if two authenticators have the same public
4914 3. Exim used not to create the message log file for a message until the first
4915 delivery attempt. This could be confusing when incoming messages were held
4916 for policy or load reasons. The message log file is now created at the time
4917 the message is received, and an initial "Received" line is written to it.
4919 4. The automatically generated man page for command line options had a minor
4920 bug that caused no ill effects; however, a more serious problem was that
4921 the procedure for building the man page automatically didn't always
4922 operate. Consequently, release 4.20 contains an out-of-date version. This
4923 shouldn't happen again.
4925 5. When building Exim with embedded Perl support, the script that builds the
4926 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4927 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4929 6. The freeze_tell option was not being used for messages that were frozen on
4930 arrival, either by an ACL or by local_scan().
4932 7. Added the smtp_incomplete_transaction log selector.
4934 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4935 was accepting AUTH without a new EHLO.
4937 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4938 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4941 10. Logging of TCP/IP connections (when configured) now happens in the main
4942 daemon process instead of the child process, so that the TCP/IP connection
4943 count is more accurate (but it can never be perfect).
4945 11. The use of "drop" in a nested ACL was not being handled correctly in the
4946 outer ACL. Now, if condition failure induced by the nested "drop" causes
4947 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4948 or "require"), the connection is dropped.
4950 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4951 that yield "discard" can now be used with an "accept" or a "discard" verb,
4952 but an error is generated for any others (because I can't see a useful way
4953 to define what should happen).
4955 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4956 are now processed in the same way as lines in the Exim configuration file.
4957 In particular, continuation lines are supported.
4959 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4961 15. Added -ti meaning -t -i.
4963 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4964 domains, and warn by logging if others are found.
4966 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4967 returned from GIFCONF when Exim is trying to find the list of interfaces on
4968 a host. The code in os.c has been modified to copy each ifreq to an aligned
4969 structure in all cases.
4971 Also, in some cases, the returned ifreq's were being copied to a 'struct
4972 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4973 means the last couple of bytes of an IPv6 address could be chopped if the
4974 ifreq contained only a normal sockaddr (14 bytes storage).
4976 18. Named domain lists were not supported in the hosts_treat_as_local option.
4977 An entry such as +xxxx was not recognized, and was treated as a literal
4980 19. Ensure that header lines added by a DATA ACL are included in the reject log
4981 if the ACL subsequently rejects the message.
4983 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4984 MD5 (which is deprecated).
4986 21. When testing a filter file using -bf, Exim was writing a message when it
4987 took the sender from a "From " line in the message, but it was not doing so
4988 when it took $return_path from a Return-Path: header line. It now does.
4990 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4991 with a valid header line field name (a series of printing characters
4992 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4994 23. Changed "disc" in the source to "disk" to conform to the documentation and
4995 the book and for uniformity.
4997 24. Ignore Sendmail's -Ooption=value command line item.
4999 25. When execve() failed while trying to run a command in a pipe transport,
5000 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5001 could be confused with a return value of 69 from the command itself. This
5002 has been changed to 127, the value the shell returns if it is asked to run
5003 a non-existent command. The wording for the related log line suggests a
5004 non-existent command as the problem.
5006 26. If received_header_text expands to an empty string, do not add a Received:
5007 header line to the message. (Well, it adds a token one on the spool, but
5008 marks it "old" so that it doesn't get used or transmitted.)
5010 27. Installed eximstats 1.28 (addition of -nt option).
5012 28. There was no check for failure on the call to getsockname() in the daemon
5013 code. This can fail if there is a shortage of resources on the system, with
5014 ENOMEM, for example. A temporary error is now given on failure.
5016 29. Contrary to the C standard, it seems that in some environments, the
5017 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5018 program. Exim now does this explicitly; it affects the formatting of
5019 timestamps using strftime().
5021 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5022 complaints. I've now initialized all the variables, to avoid this.
5024 32. Header lines added by a system filter were not being "seen" during
5025 transport-time rewrites.
5027 33. The info_callback() function passed to OpenSSL is set up with type void
5028 (*)(SSL *, int, int), as described somewhere. However, when calling the
5029 function (actually a macro) that sets it up, the type void(*)() is
5030 expected. I've put in a cast to prevent warnings from picky compilers.
5032 34. If a DNS black list lookup found a CNAME record, but there were no A
5033 records associated with the domain it pointed at, Exim crashed.
5035 35. If a DNS black list lookup returned more than one A record, Exim ignored
5036 all but the first. It now scans all returned addresses if a particular IP
5037 value is being sought. In this situation, the contents of the
5038 $dnslist_value variable are a list of all the addresses, separated by a
5041 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5042 to accept a host name and all its aliases if the forward lookup for any of
5043 them yielded the IP address of the incoming connection. Now it accepts only
5044 those names whose forward lookup yields the correct IP address. Any other
5045 names are discarded. This closes a loophole whereby a rogue DNS
5046 administrator could create reverse DNS records to break through a
5047 wildcarded host restriction in an ACL.
5049 37. If a user filter or a system filter that ran in a subprocess used any of
5050 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5051 the wrong values were passed to the pipe command ($thisaddress had the
5052 value of $0, $0 had the value of $1, etc). This bug was introduced by
5053 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5055 38. Improved the line breaking for long SMTP error messages from ACLs.
5056 Previously, if there was no break point between 40 and 75 characters, Exim
5057 left the rest of the message alone. Two changes have been made: (a) I've
5058 reduced the minimum length to 35 characters; (b) if it can't find a break
5059 point between 35 and 75 characters, it looks ahead and uses the first one
5060 that it finds. This may give the occasional overlong line, but at least the
5061 remaining text gets split now.
5063 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5064 file descriptors might be low, and that setting 1000 would always raise it.
5065 It turns out that in some environments, the limit is already over 1000 and
5066 that lowering it causes trouble. So now Exim takes care not to decrease it.
5068 40. When delivering a message, the value of $return_path is set to $sender_
5069 address at the start of routing (routers may change the value). By an
5070 oversight, this default was not being set up when an address was tested by
5071 -bt or -bv, which affected the outcome if any router or filter referred to
5074 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5075 log only when "message" or "log_message" are set. However, if one of the
5076 conditions was an address verification, or a call to a nested ACL, the
5077 messages generated by the underlying test were being passed through. This
5078 no longer happens. The underlying message is available in $acl_verify_
5079 message for both "message" and "log_message" expansions, so it can be
5080 passed through if needed.
5082 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5083 new expansion $bh_ to give the encoded byte string without charset
5084 translation. Translation happens only if iconv() is available; HAVE_ICONV
5085 indicates this at build time. HEADERS_CHARSET gives the charset to
5086 translate to; headers_charset can change it in the configuration, and
5087 "headers charset" can change it in an individual filter file.
5089 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5090 that creates RFC 2047 encoded "words" labels them as that charset instead
5091 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5092 expansion operator; (ii) when Exim creates a From: line for a local
5093 message; (iii) when a header line is rewritten to include a "phrase" part.
5095 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5096 buggy, causing it to skip the first lines of messages whose message ID
5097 ended in 'D'. This would not have bitten before Exim release 4.14, because
5098 message IDs were unlikely to end in 'D' before then. The effect was to have
5099 incorrect size information for certain domains.
5101 45. #include "config.h" was missing at the start of the crypt16.c module. This
5102 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5105 46. If there was a timeout during a "random" callout check, Exim treated it as
5106 a failure of the random address, and carried on sending RSET and the real
5107 address. If the delay was just some slowness somewhere, the response to the
5108 original RCPT would be taken as a response to RSET and so on, causing
5109 mayhem of various kinds.
5111 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5112 when I implemented it. It didn't allow for the fact that some option values
5113 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5114 the right test for positive values.
5116 48. Domain names in DNS records are case-independent. Exim always looks them up
5117 in lower case. Some resolvers return domain names in exactly the case they
5118 appear in the zone file, that is, they may contain uppercase letters. Not
5119 all resolvers do this - some return always lower case. Exim was treating a
5120 change of case by a resolver as a change of domain, similar to a widening
5121 of a domain abbreviation. This triggered its re-routing code and so it was
5122 trying to route what was effectively the same domain again. This normally
5123 caused routing to fail (because the router wouldn't handle the domain
5124 twice). Now Exim checks for this case specially, and just changes the
5125 casing of the domain that it ultimately uses when it transmits the message
5128 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5131 50. If a filter generated a file delivery with a non-absolute name (possible if
5132 no home directory exists for the router), the forbid_file option was not
5135 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5136 the existing equality matching.
5138 52. Exim was using ints instead of ino_t variables in some places where it was
5139 dealing with inode numbers.
5141 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5142 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5143 called TMPDIR, and if it finds it is different, it changes its value.
5145 54. The smtp_printf() function is now made available to local_scan() so
5146 additional output lines can be written before returning. There is also an
5147 smtp_fflush() function to enable the detection of a dropped connection.
5148 The variables smtp_input and smtp_batched_input are exported to
5151 55. Changed the default runtime configuration: the message "Unknown user"
5152 has been removed from the ACL, and instead placed on the localuser router,
5153 using the cannot_route_message feature. This means that any verification
5154 failures that generate their own messages won't get overridden. Similarly,
5155 the "Unrouteable address" message that was in the ACL for unverifiable
5156 relay addresses has also been removed.
5158 56. Added hosts_avoid_esmtp to the smtp transport.
5160 57. The exicyclog script was not checking for the esoteric option
5161 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5162 will work only if exicyclog is run under the appropriate euid.
5164 58. Following a discussion on the list, the rules by which Exim recognises line
5165 endings on incoming messages have been changed. The -dropcr and drop_cr
5166 options are now no-ops, retained only for backwards compatibility. The
5167 following line terminators are recognized: LF CRLF CR. However, special
5168 processing applies to CR:
5170 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5171 nor a local message in the state where . is a terminator.
5173 (ii) If a bare CR is encountered in a header line, an extra space is added
5174 after the line terminator so as not to end the header. The reasoning
5175 behind this is that bare CRs in header lines are most likely either
5176 to be mistakes, or people trying to play silly games.
5178 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5179 was being incorrectly given as 18 bytes larger than it should have been.
5180 This is a VOB (very old bug).
5182 60. This may never have affected anything current, but just in case it has:
5183 When the local host is found other than at the start of a list of hosts,
5184 the local host, those with the same MX, and any that follow, are discarded.
5185 When the list in question was part of a longer list of hosts, the following
5186 hosts (not currently being processed) were also being discarded. This no
5187 longer happens. I'm not sure if this situation could ever has previously
5190 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5193 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5194 header. This is something that is recommended in a new Internet Draft, and
5195 is something that is documented as being done by Sendmail. There are two
5196 possible values. For messages generated by the autoreply transport, Exim
5199 Auto-Submitted: auto-replied
5201 whereas for all other generated messages (e.g. bounces) it adds
5203 Auto-Submitted: auto-generated
5205 63. The "personal" condition in filters now includes a test for the
5206 Auto-Submitted: header. If it contains the string "auto-" the message it
5207 not considered personal.
5209 64. Added rcpt_include_affixes as a generic transport option.
5211 65. Added queue_only_override (default true).
5213 66. Added the syslog_duplication option.
5215 67. If what should have been the first header line of a message consisted of
5216 a space followed by a colon, Exim was mis-interpreting it as a header line.
5217 It isn't of course - it is syntactically invalid and should therefore be
5218 treated as the start of the message body. The misbehaviour could have
5219 caused a number of strange effects, including loss of data in subsequent
5220 header lines, and spool format errors.
5222 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5223 client host had authenticated. This control can now be exercised by an ACL
5224 for more flexibility.
5226 69. By default, callouts do not happen when testing with -bh. There is now a
5227 variant, -bhc, which does actually run the callout code, including
5228 consulting and updating the callout cache.
5230 70. Added support for saslauthd authentication, courtesy of Alexander
5233 71. If statvfs() failed on the spool or log directories while checking their
5234 size for availability, Exim confusingly gave the error "space shortage".
5235 Furthermore, in debugging mode it crashed with a floating point exception.
5236 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5237 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5238 serious problem, Exim now writes to the main and panic logs when this
5239 happens, with details of the failure. It then refuses to accept the
5240 incoming message, giving the message "spool directory problem" or "log
5241 directory problem" with a 421 code for SMTP messages.
5243 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5244 0, 1, and 2 exist, because some OS complain for execs without them (see
5245 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5246 descriptors. However, the code omitted to check that the open succeeded,
5247 causing mysterious errors if for some reason the permissions on /dev/null
5248 got screwed. Now Exim writes a message to the main and panic logs, and
5249 bombs out if it can't open /dev/null.
5251 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5252 interact so that it is all more flexible. It is supposed to remain
5253 backwards compatible. Also added extra_local_interfaces.
5255 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5256 to bomb out with an assertion failure - to the client this appears as a
5257 connection drop. This problem occurs in the part of the code that was taken
5258 from the Samba project. Fortunately, the assertion is in a very simple
5259 function, so I have fixed this by reproducing the function inline in the
5260 one place where it is called, and arranging for authentication to fail
5261 instead of killing the process with assert().
5263 75. The SPA client code was not working when the server requested OEM rather
5264 than Unicode encoding.
5266 76. Added code to make require_files with a specific uid setting more usable in
5267 the case where statting the file as root fails - usually a non-root-mounted
5268 NFS file system. When this happens and the failure is EACCES, Exim now
5269 forks a subprocess and does the per-uid checking as the relevant uid.
5271 77. Added process_log_path.
5273 78. If log_file_path was not explicitly set, a setting of check_log_space or
5274 check_log_inodes was ignored.
5276 79. If a space check for the spool or log partitions fails, the incident is now
5277 logged. Of course, in the latter case the data may get lost...
5279 80. Added the %p formatting code to string_format() so that it can be used to
5280 print addresses in debug_print(). Adjusted all the address printing in the
5281 debugging in store.c to use %p rather than %d.
5283 81. There was a concern that a line of code in smtp_in.c could overflow a
5284 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5285 initially expressed, the concern was not well-founded, because trailing
5286 spaces are removed early. However, if the trailing spaces were followed by
5287 a NULL, they did not get removed, so the overflow was possible. Two fixes
5290 (a) I re-wrote the offending code in a cleaner fashion.
5291 (b) If an incoming SMTP command contains a NULL character, it is rejected
5294 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5295 runs initgroups(), so that if the Exim user is in any additional groups,
5296 they will be used during message reception.
5302 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob