1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
11 PP/02 Harmonised TLS library version reporting; only show if debugging.
12 Layout now matches that introduced for other libraries in 4.74 PP/03.
14 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
16 PP/04 New "dns_use_edns0" global option.
22 NM/01 Workround for PCRE version dependency in version reporting
25 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
26 This fixes portability to compilers other than gcc, notably
27 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
29 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
30 makefiles for portability to HP-UX and POSIX correctness.
32 PP/01 Permit LOOKUP_foo enabling on the make command-line.
33 Also via indented variable definition in the Makefile.
34 (Debugging by Oliver Heesakkers).
36 PP/02 Restore caching of spamd results with expanded spamd_address.
37 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
39 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
40 Improves build reliability. Fix from: Frank Elsner
42 NM/02 Fix wide character breakage in the rfc2047 coding
43 Fixes bug 1064. Patch from Andrey N. Oktyabrski
45 NM/03 Allow underscore in dnslist lookups
46 Fixes bug 1026. Patch from Graeme Fowler
48 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
49 Code patches from Adam Ciarcinski of NetBSD.
51 NM/04 Fixed exiqgrep to cope with mailq missing size issue
54 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
55 is logged, to avoid truncation. Patch from John Horne.
57 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
58 Patch from Jakob Hirsch.
60 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
61 SQL string expansion failure details.
62 Patch from Andrey Oktyabrski.
64 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
65 Patch from Simon Arlott.
67 PP/09 New lookups functionality failed to compile on old gcc which rejects
68 extern declarations in function scope.
69 Patch from Oliver Fleischmann
71 PP/10 Use sig_atomic_t for flags set from signal handlers.
72 Check getgroups() return and improve debugging.
73 Fixed developed for diagnosis in bug 927 (which turned out to be
76 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
77 Patch from Mark Zealey.
79 PP/12 Bugzilla 1056: Improved spamd server selection.
80 Patch from Mark Zealey.
82 PP/13 Bugzilla 1086: Deal with maildir quota file races.
83 Based on patch from Heiko Schlittermann.
85 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
86 Patch from Uwe Doering, sign-off by Michael Haardt.
88 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
89 variable declaration deep within a block. Bug and patch from
92 PP/15 lookups-Makefile IRIX compatibilty coercion.
94 PP/16 Make DISABLE_DKIM build knob functional.
96 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
99 TF/03 Fix valgrind.h portability to C89 compilers that do not support
100 variable argument macros. Our copy now differs from upstream.
106 TF/01 Failure to get a lock on a hints database can have serious
107 consequences so log it to the panic log.
109 TF/02 Log LMTP confirmation messages in the same way as SMTP,
110 controlled using the smtp_confirmation log selector.
112 TF/03 Include the error message when we fail to unlink a spool file.
114 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
115 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
116 for maintaining out-of-tree patches for some time.
118 PP/01 Bugzilla 139: Documentation and portability issues.
119 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
120 Handle per-OS dynamic-module compilation flags.
122 PP/02 Let /dev/null have normal permissions.
123 The 4.73 fixes were a little too stringent and complained about the
124 permissions on /dev/null. Exempt it from some checks.
125 Reported by Andreas M. Kirchwitz.
127 PP/03 Report version information for many libraries, including
128 Exim version information for dynamically loaded libraries. Created
129 version.h, now support a version extension string for distributors
130 who patch heavily. Dynamic module ABI change.
132 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
133 privilege escalation vulnerability whereby the Exim run-time user
134 can cause root to append content of the attacker's choosing to
137 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
140 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
141 If dropping privileges for untrusted macros, we disabled normal logging
142 on the basis that it would fail; for the Exim run-time user, this is not
143 the case, and it resulted in successful deliveries going unlogged.
144 Fixed. Reported by Andreas Metzler.
150 PP/01 Date: & Message-Id: revert to normally being appended to a message,
151 only prepend for the Resent-* case. Fixes regression introduced in
152 Exim 4.70 by NM/22 for Bugzilla 607.
154 PP/02 Include check_rfc2047_length in configure.default because we're seeing
155 increasing numbers of administrators be bitten by this.
157 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
159 PP/03 Bugzilla 994: added openssl_options main configuration option.
161 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
163 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
165 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
167 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
168 without a peer certificate, leading to a segfault because of an
169 assumption that peers always have certificates. Be a little more
170 paranoid. Problem reported by Martin Tscholak.
172 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
173 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
174 NB: ClamAV planning to remove STREAM in "middle of 2010".
175 CL also introduces -bmalware, various -d+acl logging additions and
176 more caution in buffer sizes.
178 PP/09 Implemented reverse_ip expansion operator.
180 PP/10 Bugzilla 937: provide a "debug" ACL control.
182 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
184 PP/12 Bugzilla 973: Implement --version.
186 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
188 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
190 PP/15 Bugzilla 816: support multiple condition rules on Routers.
192 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
193 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
194 ignore trailing whitespace.
196 JJ/02 prevent non-panic DKIM error from being sent to paniclog
198 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
201 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
202 Notification from Dr Andrew Aitchison.
204 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
205 ExtendedDetectionInfo response format.
206 Notification from John Horne.
208 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
211 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
212 XSL and documented dependency on system catalogs, with examples of how
215 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
218 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
219 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
220 configuration file which is writeable by the Exim user or group.
222 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
223 of configuration files to cover files specified with the -C option if
224 they are going to be used with root privileges, not just the default
227 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
228 option (effectively making it always true).
230 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
231 files to be used while preserving root privileges.
233 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
234 that rogue child processes cannot use them.
236 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
237 run-time user, instead of root.
239 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
240 Exim run-time user without dropping privileges.
242 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
243 result string, instead of calling string_vformat() twice with the same
246 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
247 for other users. Others should always drop root privileges if they use
248 -C on the command line, even for a whitelisted configure file.
250 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
252 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
258 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
259 $data_path, and $header_path variables; fixed documentation bugs and
262 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
263 exipick to access non-standard spools, including the "frozen" queue
266 NM/01 Bugzilla 965: Support mysql stored procedures.
267 Patch from Alain Williams
269 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
271 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
272 Patch from Andreas Metzler
274 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
275 Patch from Kirill Miazine
277 NM/05 Bugzilla 671: Added umask to procmail example.
279 JJ/03 installed exipick 20100323.0, fixing doc bug
281 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
282 directory. Notification and patch from Dan Rosenberg.
284 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
286 TK/02 Improve log output when DKIM signing operation fails.
288 MH/01 Treat the transport option dkim_domain as a colon separated
289 list, not as a single string, and sign the message with each element,
290 omitting multiple occurences of the same signer.
292 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
293 Bugzilla 985, 986. Patch by Simon Arlott
295 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
296 Patch by Simon Arlott
298 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
299 MBX locking. Notification from Dan Rosenberg.
305 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
307 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
309 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
311 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
313 NM/04 Bugzilla 915: Flag broken perl installation during build.
319 TK/01 Added patch by Johannes Berg that expands the main option
320 "spamd_address" if it starts with a dollar sign.
322 TK/02 Write list of recipients to X-Envelope-Sender header when building
323 the mbox-format spool file for content scanning (suggested by Jakob
326 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
327 (http://www.dcc-servers.net/) support via dccifd. Activated by
328 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
330 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
331 by Mark Daniel Reidel <mr@df.eu>.
333 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
334 When building exim an external PCRE library is now needed -
335 PCRE is a system library on the majority of modern systems.
336 See entry on PCRE_LIBS in EDITME file.
338 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
339 conversation. Added nologin parameter to request.
340 Patch contributed by Kirill Miazine.
342 TF/01 Do not log submission mode rewrites if they do not change the address.
344 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
346 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
347 log files in place. Contributed by Roberto Lima.
349 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
351 TF/03 Bugzilla 615: When checking the local_parts router precondition
352 after a local_part_suffix or local_part_prefix option, Exim now
353 does not use the address's named list lookup cache, since this
354 contains cached lookups for the whole local part.
356 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
357 Robert Millan. Documentation is in experimental-spec.txt.
359 TF/04 Bugzilla 668: Fix parallel build (make -j).
361 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
363 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
364 Patch provided by Jan Srzednicki.
366 TF/05 Leading white space used to be stripped from $spam_report which
367 wrecked the formatting. Now it is preserved.
369 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
370 that they are available at delivery time.
372 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
374 TF/08 TLS error reporting now respects the incoming_interface and
375 incoming_port log selectors.
377 TF/09 Produce a more useful error message if an SMTP transport's hosts
378 setting expands to an empty string.
380 NM/06 Bugzilla 744: EXPN did not work under TLS.
381 Patch provided by Phil Pennock.
383 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
384 Patch provided by Richard Godbee.
386 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
387 acl_smtp_notquit, added index entry.
389 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
390 Patch provided by Eugene Bujak.
392 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
393 accept(). Patch provided by Maxim Dounin.
395 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
396 Patch provided by Phil Pennock.
398 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
400 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
401 Patch provided by Brad "anomie" Jorsch.
403 NM/14 Bugzilla 622: Added timeout setting to transport filter.
404 Patch provided by Dean Brooks.
406 TK/05 Add native DKIM support (does not depend on external libraries).
408 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
409 Patch provided by Graeme Fowler.
411 NM/16 Bugzilla 851: Documentation example syntax fix.
413 NM/17 Changed NOTICE file to remove references to embedded PCRE.
415 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
418 NM/19 Bugzilla 745: TLS version reporting.
419 Patch provided by Phil Pennock.
421 NM/20 Bugzilla 167: bool: condition support.
422 Patch provided by Phil Pennock.
424 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
425 clients. Patch provided by Phil Pennock.
427 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
428 Patch provided by Brad "anomie" Jorsch.
430 NM/23 Bugzilla 687: Fix misparses in eximstats.
431 Patch provided by Heiko Schlittermann.
433 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
434 Patch provided by Heiko Schlittermann.
436 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
437 plus update to original patch.
439 NM/26 Bugzilla 799: Documentation correction for ratelimit.
441 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
442 Patch provided by David Brownlee.
444 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
446 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
448 NM/30 Bugzilla 888: TLS documentation bugfixes.
450 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
452 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
453 Unlike the original bugzilla I have changed all shell scripts in src tree.
455 NM/33 Bugzilla 898: Transport filter timeout fix.
456 Patch by Todd Rinaldo.
458 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
459 Patch by Serge Demonchaux.
461 NM/35 Bugzilla 39: Base64 decode bug fixes.
462 Patch by Jakob Hirsch.
464 NM/36 Bugzilla 909: Correct connect() call in dcc code.
466 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
468 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
470 NM/39 Bugzilla 911: Fixed MakeLinks build script.
476 TK/01 Add preliminary DKIM support. Currently requires a forked version of
477 ALT-N's libdkim that I have put here:
478 http://duncanthrax.net/exim-experimental/
480 Note to Michael Haardt: I had to rename some vars in sieve.c. They
481 were called 'true' and it seems that C99 defines that as a reserved
482 keyword to be used with 'bool' variable types. That means you could
483 not include C99-style headers which use bools without triggering
484 build errors in sieve.c.
486 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
487 as mailq or other aliases. Changed the --help handling significantly
488 to do whats expected. exim_usage() emits usage/help information.
490 SC/01 Added the -bylocaldomain option to eximstats.
492 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
494 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
496 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
502 PH/01 Another patch from the Sieve maintainer.
504 PH/02 When an IPv6 address is converted to a string for single-key lookup
505 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
506 dots are used instead of colons so that keys in lsearch files need not
507 contain colons. This was done some time before quoting was made available
508 in lsearch files. However, iplsearch files do require colons in IPv6 keys
509 (notated using the quote facility) so as to distinguish them from IPv4
510 keys. This meant that lookups for IP addresses in host lists did not work
511 for iplsearch lookups.
513 This has been fixed by arranging for IPv6 addresses to be expressed with
514 colons if the lookup type is iplsearch. This is not incompatible, because
515 previously such lookups could never work.
517 The situation is now rather anomolous, since one *can* have colons in
518 ordinary lsearch keys. However, making the change in all cases is
519 incompatible and would probably break a number of configurations.
521 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
524 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
525 conversion specification without a maximum field width, thereby enabling
526 a rogue spamd server to cause a buffer overflow. While nobody in their
527 right mind would setup Exim to query an untrusted spamd server, an
528 attacker that gains access to a server running spamd could potentially
529 exploit this vulnerability to run arbitrary code as the Exim user.
531 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
532 $primary_hostname instead of what libspf2 thinks the hosts name is.
534 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
535 a directory entry by the name of the lookup key. Previously, if a
536 symlink pointed to a non-existing file or a file in a directory that
537 Exim lacked permissions to read, a lookup for a key matching that
538 symlink would fail. Now it is enough that a matching directory entry
539 exists, symlink or not. (Bugzilla 503.)
541 PH/03 The body_linecount and body_zerocount variables are now exported in the
544 PH/04 Added the $dnslist_matched variable.
546 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
547 This means they are set thereafter only if the connection becomes
550 PH/06 Added the client_condition to authenticators so that some can be skipped
551 by clients under certain conditions.
553 PH/07 The error message for a badly-placed control=no_multiline_responses left
554 "_responses" off the end of the name.
556 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
558 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
559 (without spaces) instead of just copying the configuration text.
561 PH/10 Added the /noupdate option to the ratelimit ACL condition.
563 PH/11 Added $max_received_linelength.
565 PH/12 Added +ignore_defer and +include_defer to host lists.
567 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
568 way in which PCRE > 7.0 is built.
570 PH/14 Implemented queue_only_load_latch.
572 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
573 MAIL command. The effect was to mangle the value on 64-bit systems.
575 PH/16 Another patch from the Sieve maintainer.
577 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
579 PH/18 If a system quota error occurred while trying to create the file for
580 a maildir delivery, the message "Mailbox is full" was not appended to the
581 bounce if the delivery eventually timed out. Change 4.67/27 below applied
582 only to a quota excession during the actual writing of the file.
584 PH/19 It seems that peer DN values may contain newlines (and other non-printing
585 characters?) which causes problems in log lines. The DN values are now
586 passed through string_printing() before being added to log lines.
588 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
589 and InterBase are left for another time.)
591 PH/21 Added message_body_newlines option.
593 PH/22 Guard against possible overflow in moan_check_errorcopy().
595 PH/23 POSIX allows open() to be a macro; guard against that.
597 PH/24 If the recipient of an error message contained an @ in the local part
598 (suitably quoted, of course), incorrect values were put in $domain and
599 $local_part during the evaluation of errors_copy.
605 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
606 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
609 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
610 issue a MAIL command.
612 PH/02 In an ACL statement such as
614 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
616 if a client was not listed at all, or was listed with a value other than
617 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
618 the condition was not true (as it should be), so access was not denied.
619 The bug was that the ! inversion was incorrectly passed on to the second
620 item. This has been fixed.
622 PH/03 Added additional dnslists conditions == and =& which are different from
623 = and & when the dns lookup returns more than one IP address.
625 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
626 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
628 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
629 FSYNC, which compiles an option called disable_fsync that allows for
630 bypassing fsync(). The documentation is heavily laced with warnings.
632 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
634 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
635 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
636 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
637 including adding "make clean"; (3) Added -fPIC when compiling the test
638 dynamically loaded module, to get rid of a warning.
640 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
641 message fails, move_frozen_messages = true and ignore_bounce_errors_after
642 = 0s. The bug is otherwise harmless.
644 PH/07 There was a bug in the dovecot authenticator such that the value of
645 $auth1 could be overwritten, and so not correctly preserved, after a
646 successful authentication. This usually meant that the value preserved by
647 the server_setid option was incorrect.
649 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
651 PH/09 Installed PCRE release 7.0.
653 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
654 run for batched SMTP input. It is now run at the start of every message
655 in the batch. While fixing this I discovered that the process information
656 (output by running exiwhat) was not always getting set for -bs and -bS
657 input. This is fixed, and it now also says "batched" for BSMTP.
659 PH/11 Added control=no_pipelining.
661 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
662 patch, slightly modified), and move the expansion of helo_data till after
663 the connection is made in the smtp transport (so it can use these
666 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
668 PH/14 Added log_selector = +pid.
670 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
672 PH/16 Add ${if forany and ${if forall.
674 PH/17 Added dsn_from option to vary the From: line in DSNs.
676 PH/18 Flush SMTP output before performing a callout, unless control =
677 no_callout_flush is set.
679 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
680 was true (the default) a successful delivery failed to delete the retry
681 item, thus causing premature timeout of the address. The bug is now
684 PH/20 Added hosts_avoid_pipelining to the smtp transport.
686 PH/21 Long custom messages for fakedefer and fakereject are now split up
687 into multiline reponses in the same way that messages for "deny" and
688 other ACL rejections are.
690 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
691 with slight modification.
693 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
694 draft, changing the syntax and factoring some duplicate code".
696 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
697 for deliveries of the second and subsequent messages over the same SMTP
700 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
701 ${reduce, with only minor "tidies".
703 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
705 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
706 expansion side effects.
708 PH/27 When a message times out after an over-quota error from an Exim-imposed
709 quota, the bounce message says "mailbox is full". This message was not
710 being given when it was a system quota that was exceeded. It now should
713 MH/03 Made $recipients available in local_scan(). local_scan() already has
714 better access to the recipient list through recipients_list[], but
715 $recipients can be useful in postmaster-provided expansion strings.
717 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
718 in the case of a MAIL command with additional options following the
719 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
720 were accidentally chopped off.
722 PH/29 SMTP synchronization checks are implemented when a command is read -
723 there is a check that no more input is waiting when there shouldn't be
724 any. However, for some commands, a delay in an ACL can mean that it is
725 some time before the response is written. In this time, more input might
726 arrive, invalidly. So now there are extra checks after an ACL has run for
727 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
728 pipelining has not been advertised.
730 PH/30 MH's patch to allow iscntrl() characters to be list separators.
732 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
733 returned in the SMTP response when smtp_return_error_details was false.
736 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
737 instead of the C I/O that was originally supplied, because problems were
740 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
741 Exim which did not show up earlier: it was assuming that a call to
742 SSL_CTX_set_info_callback() might give an error value. In fact, there is
743 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
744 was a macro that became an assignment, so it seemed to work. This has
745 changed to a proper function call with a void return, hence the compile
746 error. Exim's code has been fixed.
748 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
751 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
753 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
755 PH/37 If a message is not accepted after it has had an id assigned (e.g.
756 because it turns out to be too big or there is a timeout) there is no
757 "Completed" line in the log. When some messages of this type were
758 selected by exigrep, they were listed as "not completed". Others were
759 picked up by some special patterns. I have improved the selection
760 criteria to be more general.
762 PH/38 The host_find_failed option in the manualroute router can now be set
763 to "ignore", to completely ignore a host whose IP address cannot be
764 found. If all hosts are ignored, the behaviour is controlled by the new
765 host_all_ignored option.
767 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
768 homing or because of multiple MX records with /mx) generated more than
769 one IP address, and the following item turned out to be the local host,
770 all the secondary addresses of the first item were incorrectly removed
771 from the list, along with the local host and any following hosts (which
772 is what is supposed to happen).
774 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
775 whoever called Exim into the -H file. In the case of the daemon it was
776 behaving confusingly. When first started, it used values for whoever
777 started the daemon, but after a SIGHUP it used the Exim user (because it
778 calls itself on a restart). I have changed the code so that it now always
781 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
782 message are rejected with the same error (e.g. no authentication or bad
783 sender address), and a DATA command is nevertheless sent (as can happen
784 with PIPELINING or a stupid MUA), the error message that was given to the
785 RCPT commands is included in the rejection of the DATA command. This is
786 intended to be helpful for MUAs that show only the final error to their
789 PH/42 Another patch from the Sieve maintainer.
791 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
792 Eximstats - Fixed some broken HTML links and added missing column headers
794 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
795 columns for Rejects, Temp Rejects, Ham, and Spam rows.
797 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
799 PH/43 Yet another patch from the Sieve maintainer.
801 PH/44 I found a way to check for a TCP/IP connection going away before sending
802 the response to the final '.' that terminates a message, but only in the
803 case where the client has not sent further data following the '.'
804 (unfortunately, this is allowed). However, in many cases there won't be
805 any further data because there won't be any more messages to send. A call
806 to select() can be used: if it shows that the input is "ready", there is
807 either input waiting, or the socket has been closed. An attempt to read
808 the next input character can distinguish the two cases. Previously, Exim
809 would have sent an OK response which the client would never have see.
810 This could lead to message repetition. This fix should cure that, at
811 least in a lot of common cases.
813 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
814 advertised in response to EHLO.
820 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
821 fixed by 4.65/MH/01 (is this a record?) are fixed:
823 (i) An empty string was always treated as zero by the numeric comparison
824 operators. This behaviour has been restored.
826 (ii) It is documented that the numeric comparison operators always treat
827 their arguments as decimal numbers. This was broken in that numbers
828 starting with 0 were being interpreted as octal.
830 While fixing these problems I realized that there was another issue that
831 hadn't been noticed. Values of message_size_limit (both the global option
832 and the transport option) were treated as octal if they started with 0.
833 The documentation was vague. These values are now always treated as
834 decimal, and I will make that clear in the documentation.
840 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
841 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
844 MH/01 Don't check that the operands of numeric comparison operators are
845 integers when their expansion is in "skipping" mode (fixes bug
846 introduced by 4.64-PH/07).
848 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
849 child addresses, Exim now panics and dies. Previously, because the count
850 is held in a short int, deliveries were likely to be lost. As such a
851 large number of recipients for a single message is ridiculous
852 (performance will be very, very poor), I have chosen to impose a limit
853 rather than extend the field.
859 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
860 leftover -K file (the existence of which was triggered by #402).
861 While we were at it, introduced process PID as part of the -K
862 filename. This should rule out race conditions when creating
865 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
866 processing considerably. Previous code took too long for large mails,
867 triggering a timeout which in turn triggers #401.
869 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
870 in the DK code in transports.c. sendfile() is not really portable,
871 hence the _LINUX specificness.
873 TF/01 In the add_headers option to the mail command in an Exim filter,
874 there was a bug that Exim would claim a syntax error in any
875 header after the first one which had an odd number of characters
878 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
879 callout verification, Exim cached a "reject" for the entire domain. This
880 is correct for most verifications, but it is not correct for a recipient
881 verification with use_sender or use_postmaster set, because in that case
882 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
883 case of MAIL FROM:<> rejection from other early rejections (e.g.
884 rejection of HELO). When verifying a recipient using a non-null MAIL
885 address, the cache is ignored if it shows MAIL FROM:<> rejection.
886 Whatever the result of the callout, the value of the domain cache is
887 left unchanged (for any other kind of callout, getting as far as trying
888 RCPT means that the domain itself is ok).
890 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
893 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
894 manifest itself as EPIPE rather than ECONNECT. When tidying away a
895 session, the daemon ignores ECONNECT errors and logs others; it now
896 ignores EPIPE as well.
898 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
899 (quoted-printable decoding).
901 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
902 later the small subsequent patch to fix an introduced bug.
904 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
906 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
908 PH/08 An error is now given if message_size_limit is specified negative.
910 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
911 to be given (somewhat) arbitrary names.
913 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
916 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
917 miscellaneous code fixes
919 PH/10 Added the log_reject_target ACL modifier to specify where to log
922 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
923 hostname. This is wrong, because it relates to the incoming message (and
924 probably the interface on which it is arriving) and not to the outgoing
925 callout (which could be using a different interface). This has been
926 changed to use the value of the helo_data option from the smtp transport
927 instead - this is what is used when a message is actually being sent. If
928 there is no remote transport (possible with a router that sets up host
929 addresses), $smtp_active_hostname is used.
931 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
932 tweaks were necessary in order to get it to work (see also 21 below):
933 (a) The code assumed that strncpy() returns a negative number on buffer
934 overflow, which isn't the case. Replaced with Exim's string_format()
936 (b) There were several signed/unsigned issues. I just did the minimum
937 hacking in of casts. There is scope for a larger refactoring.
938 (c) The code used strcasecmp() which is not a standard C function.
939 Replaced with Exim's strcmpic() function.
940 (d) The code set only $1; it now sets $auth1 as well.
941 (e) A simple test gave the error "authentication client didn't specify
942 service in request". It would seem that Dovecot has changed its
943 interface. Fortunately there's a specification; I followed it and
944 changed what the client sends and it appears to be working now.
946 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
949 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
950 address A is aliased to B and C, where B exists and C does not. Without
951 -v the output is "A verified" because verification stops after a
952 successful redirection if more than one address is generated. However,
953 with -v the child addresses are also verified. Exim was outputting "A
954 failed to verify" and then showing the successful verification for C,
955 with its parentage. It now outputs "B failed to verify", showing B's
956 parentage before showing the successful verification of C.
958 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
959 look up a TXT record in a specific list after matching in a combined
962 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
963 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
964 they consult the DNS. I had assumed they would set it the way they
965 wanted; and indeed my experiments on Linux seem to show that in some
966 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
967 To be on the safe side, however, I have now made the interface to
968 host_find_byname() similar to host_find_bydns(), with an argument
969 containing the DNS resolver options. The host_find_byname() function now
970 sets these options at its start, just as host_find_bydns() does. The smtp
971 transport options dns_qualify_single and dns_search_parents are passed to
972 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
973 of host_find_byname() use the default settings of RES_DEFNAMES
974 (qualify_single) but not RES_DNSRCH (search_parents).
976 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
977 spool_read_header() do less string testing, by means of a preliminary
978 switch on the second character of optional "-foo" lines. (This is
979 overdue, caused by the large number of possibilities that now exist.
980 Originally there were few.) While I was there, I also converted the
981 str(n)cmp tests so they don't re-test the leading "-" and the first
982 character, in the hope this might squeeze out yet more improvement.
984 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
985 flag allowing group syntax was set by the header_syntax check but not
986 turned off, possible causing trouble later; (2) The flag was not being
987 set at all for the header_verify test, causing "group"-style headers to
988 be rejected. I have now set it in this case, and also caused header_
989 verify to ignore an empty address taken from a group. While doing this, I
990 came across some other cases where the code for allowing group syntax
991 while scanning a header line wasn't quite right (mostly, not resetting
992 the flag correctly in the right place). These bugs could have caused
993 trouble for malformed header lines. I hope it is now all correct.
995 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
996 with the "reply" argument non-NULL. The code, however (which originally
997 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
998 but it didn't always do it. This confused somebody who was copying the
999 code for some other use. I have removed all the tests.
1001 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1002 feature that was used to support insecure browsers during the U.S. crypto
1003 embargo. It requires special client support, and Exim is probably the
1004 only MTA that supported it -- and would never use it because real RSA is
1005 always available. This code has been removed, because it had the bad
1006 effect of slowing Exim down by computing (never used) parameters for the
1007 RSA_EXPORT functionality.
1009 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1010 authenticator to fail if there's a tab character in the incoming data
1011 (there should never be unless someone is messing about, as it's supposed
1012 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1013 if the connection is using TLS or if the remote IP is the same as the
1014 local IP, and the "valid-client-cert option" if a client certificate has
1017 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1018 authenticators. This can be used for authorization after authentication
1019 succeeds. (In the case of plaintext, it servers for both authentication
1022 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1023 if any retry times were supplied.
1025 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1026 connection, where there is no client IP address to check. In this
1027 situation, the verify now always succeeds.
1029 PH/25 Applied John Jetmore's -Mset patch.
1031 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1033 PH/27 In a string expansion for a processed (not raw) header when multiple
1034 headers of the same name were present, leading whitespace was being
1035 removed from all of them, but trailing whitespace was being removed only
1036 from the last one. Now trailing whitespace is removed from each header
1037 before concatenation. Completely empty headers in a concatenation (as
1038 before) are ignored.
1040 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1041 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1043 PH/29 [Removed. This was a change that I later backed out, and forgot to
1044 correct the ChangeLog entry (that I had efficiently created) before
1045 committing the later change.]
1047 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1048 address errors (4xx response to RCPT) over the same connection as other
1049 messages routed to the same hosts. Such deliveries are always "forced",
1050 so retry times are not inspected. This resulted in far too many retries
1051 for the affected addresses. The effect occurred only when there were more
1052 hosts than the hosts_max_try setting in the smtp transport when it had
1053 the 4xx errors. Those hosts that it had tried were not added to the list
1054 of hosts for which the message was waiting, so if all were tried, there
1055 was no problem. Two fixes have been applied:
1057 (i) If there are any address or message errors in an SMTP delivery, none
1058 of the hosts (tried or untried) are now added to the list of hosts
1059 for which the message is waiting, so the message should not be a
1060 candidate for sending over the same connection that was used for a
1061 successful delivery of some other message. This seems entirely
1062 reasonable: after all the message is NOT "waiting for some host".
1063 This is so "obvious" that I'm not sure why it wasn't done
1064 previously. Hope I haven't missed anything, but it can't do any
1065 harm, as the worst effect is to miss an optimization.
1067 (ii) If, despite (i), such a delivery is accidentally attempted, the
1068 routing retry time is respected, so at least it doesn't keep
1069 hammering the server.
1071 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1072 in ${readsocket because some servers need this prod.
1074 PH/32 Added some extra debug output when updating a wait-xxx database.
1076 PH/33 The hint "could be header name not terminated by colon", which has been
1077 given for certain expansion errors for a long time, was not being given
1078 for the ${if def:h_colon_omitted{... case.
1080 PH/34 The spec says: "With one important exception, whenever a domain list is
1081 being scanned, $domain contains the subject domain." There was at least
1082 one case where this was not true.
1084 PH/35 The error "getsockname() failed: connection reset by peer" was being
1085 written to the panic log as well as the main log, but it isn't really
1086 panic-worthy as it just means the connection died rather early on. I have
1087 removed the panic log writing for the ECONNRESET error when getsockname()
1090 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1091 runs only) independently of the message's sender address. This meant
1092 that, if the 4xx error was in fact related to the sender, a different
1093 message to the same recipient with a different sender could confuse
1094 things. In particualar, this can happen when sending to a greylisting
1095 server, but other circumstances could also provoke similar problems.
1096 I have changed the default so that the retry time for these errors is now
1097 based a combination of the sender and recipient addresses. This change
1098 can be overridden by setting address_retry_include_sender=false in the
1101 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1102 remote server are returned as part of bounce messages. This was not
1103 happening for LMTP over a pipe (the lmtp transport), but now it is the
1104 same for both kinds of LMTP.
1106 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1107 in header lines that were in fact CNAMEs. This is no longer the case.
1109 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1110 and queue runs started by the daemon processed all messages. This has
1111 been fixed so that -R and -S can now usefully be given with -q<time>.
1113 PH/40 Import PCRE release 6.7 (fixes some bugs).
1115 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1117 PH/42 Give an error if -q is specified more than once.
1119 PH/43 Renamed the variables $interface_address and $interface_port as
1120 $received_ip_address and $received_port, to make it clear that these
1121 values apply to message reception, and not to the outgoing interface when
1122 a message is delivered. (The old names remain recognized, of course.)
1124 PH/44 There was no timeout on the connect() call when using a Unix domain
1125 socket in the ${readsocket expansion. There now is.
1127 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1128 be meaningful with "accept".
1130 SC/01 Eximstats V1.43
1131 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1133 SC/02 Eximstats V1.44
1134 Use a glob alias rather than an array ref in the generated
1135 parser. This improves both readability and performance.
1137 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1138 Collect SpamAssassin and rejection statistics.
1139 Don't display local sender or destination tables unless
1140 there is data to show.
1141 Added average volumes into the top table text output.
1143 SC/04 Eximstats V1.46
1144 Collect data on the number of addresses (recipients)
1145 as well as the number of messages.
1147 SC/05 Eximstats V1.47
1148 Added 'Message too big' to the list of mail rejection
1149 reasons (thanks to Marco Gaiarin).
1151 SC/06 Eximstats V1.48
1152 Mainlog lines which have GMT offsets and are too short to
1153 have a flag are now skipped.
1155 SC/07 Eximstats V1.49 (Alain Williams)
1156 Added the -emptyok flag.
1158 SC/08 Eximstats V1.50
1159 Fixes for obtaining the IP address from reject messages.
1161 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1162 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1163 whitesspace changes from 4.64-PH/27
1165 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1168 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1169 are found, allow negative numbers in numeric criteria)
1171 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1173 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1174 to match changes made in 4.64-PH/43
1176 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1178 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1179 "domain" is actually an IP address, and force a failure. This locks out
1180 those revolvers/nameservers that support "A-for-A" lookups, in
1181 contravention of the specifications.
1183 PH/48 When a host name was looked up from an IP address, and the subsequent
1184 forward lookup of the name timed out, the host name was left in
1185 $sender_host_name, contrary to the specification.
1187 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1188 restricted to single-key lookups, Exim was not diagnosing an error if
1189 * or *@ was used with a query-style lookup.
1191 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1193 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1194 long ago, but noone interested enough thought of it. Let's just say that
1195 the "1.1" means that there are some new functions that weren't there at
1196 some point in the past.
1198 PH/51 Error processing for expansion failure of helo_data from an smtp
1199 transport during callout processing was broken.
1201 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1202 tested/used via the -bh/-bhc/-bs options.
1204 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1205 bug, fixed in subsequent PCRE releases).
1207 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1208 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1210 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1216 SC/01 Use a glob alias rather than an array ref in eximstats generated
1217 parser. This improves both readability and performance.
1219 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1220 Don't display local sender or destination tables in eximstats unless
1221 there is data to show.
1222 Added average volumes into the eximstats top table text output.
1224 SC/03 Collect data on the number of addresses (recipients) as well
1225 as the number of messages in eximstats.
1227 TF/01 Correct an error in the documentation for the redirect router. Exim
1228 does (usually) call initgroups() when daemonizing.
1230 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1231 with consistent privilege compared to when running as a daemon.
1233 TF/03 Note in the spec that $authenticated_id is not set for local
1234 submissions from trusted users.
1236 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1237 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1239 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1240 by adding some example configuration directives to the default
1241 configuration file. A little bit of work is required to uncomment the
1242 directives and define how usernames and passwords are checked, but
1243 there is now a framework to start from.
1245 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1246 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1247 without this. I don't know how relevant this is to other LDAP libraries.
1249 PH/02 Add the verb name to the "unknown ACL verb" error.
1251 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1253 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1255 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1256 directory not expanded when it should be if an expanded home directory
1257 was set for the address (which is overridden by the transport).
1259 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1262 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1263 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1264 because it is too late at that time, and has no effect.
1266 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1267 security issue with \' (bugzilla #107). I could not use the
1268 PQescapeStringConn() function, because it needs a PGconn value as one of
1271 PH/08 When testing addresses using -bt, indicate those final addresses that
1272 are duplicates that would not cause an additional delivery. At least one
1273 person was confused, thinking that -bt output corresponded to deliveries.
1274 (Suppressing duplicates isn't a good idea as you lose the information
1275 about possibly different redirections that led to the duplicates.)
1277 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1278 systems where poll() doesn't work, in particular OS X.
1280 PH/10 Added more information to debugging output for retry time not reached.
1282 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1283 operations in malware.c.
1285 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1288 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1289 syslog_duplication set false, log lines that would normally be written
1290 both the the main log and to the reject log were not written to syslog at
1293 PH/14 In the default configuration, change the use of "message" in ACL warn
1294 statements to "add_header".
1296 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1297 not followed by a command (e.g. "seen endif").
1299 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1300 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1303 PH/17 Added extra conditions to the default value of delay_warning_condition
1307 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1308 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1309 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1312 The Auto-Submitted: and various List- headers are standardised, whereas I
1313 don't think Precedence: ever was.
1315 PH/18 Refactored debugging code in route_finduser() to show more information,
1316 in particular, the error code if getpwnam() issues one.
1318 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1319 This is apparently needed in addition to the PH/07 change above to avoid
1320 any possible encoding problems.
1322 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1323 but not after initializing Perl.
1325 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1326 output them only if debugging. By default they are written stderr,
1327 apparently, which is not desirable.
1329 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1332 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1335 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1337 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1338 authenticated or an ident call has been made. Suppress the default
1339 values for $authenticated_id and $authenticated_sender (but permit -oMai
1340 and -oMas) when testing with -bh.
1342 PH/24 Re-jigged the order of the tests in the default configuration so that the
1343 tests for valid domains and recipients precede the DNS black list and CSA
1344 tests, on the grounds that those ones are more expensive.
1346 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1347 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1348 This bug exists in every version of Exim that I still have, right back to
1351 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1352 However, an attempt to turn on case-sensitivity in a regex key by
1353 including (?-i) didn't work because the subject string was already
1354 lowercased, and the effects were non-intuitive. It turns out that a
1355 one-line patch can be used to allow (?-i) to work as expected.
1361 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1362 other effects) broke the use of negated acl sub-conditions.
1364 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1367 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1368 "Deny" causes Exim to reject the incoming connection with a 554 error.
1369 Unfortunately, if there is a major crisis, such as a disk failure,
1370 tcp-wrappers gives "deny", whereas what one would like would be some
1371 kind of temporary error. A kludge has been added to help with this.
1372 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1373 554 error is used if errno is still zero or contains ENOENT (which occurs
1374 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1377 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1379 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1380 errors. Otherwise a message that provokes a temporary error (when other
1381 messages do not) can cause a whole host to time out.
1383 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1384 addresses were routed directly to files or pipes from a redirect router.
1385 File deliveries just didn't batch; pipe deliveries might have suffered
1388 PH/06 A failure to get a lock for a hints database would erroneously always say
1389 "Failed to get write lock", even when it was really a read lock.
1391 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1392 of 0600. This has been changed to use the value of the lockfile_mode
1393 option (which defaults to 0600).
1395 PH/08 Applied small patch from the Sieve maintainer.
1397 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1398 folder from quota calculations, a direct delivery into this folder messed
1399 up the contents of the maildirsize file. This was because the regex was
1400 used only to exclude .Trash (or whatever) when the size of the mailbox
1401 was calculated. There was no check that a delivery was happening into an
1402 excluded directory. This bug has been fixed by ignoring all quota
1403 processing for deliveries into excluded directories.
1405 PH/10 Added the maildirfolder_create_regex option to appendfile.
1411 PH/01 The code for finding all the local interface addresses on a FreeBSD
1412 system running IPv6 was broken. This may well have applied to all BSD
1413 systems, as well as to others that have similar system calls. The broken
1414 code found IPv4 interfaces correctly, but gave incorrect values for the
1415 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1416 that it would not match correctly against @[] and not recognize the IPv6
1419 PH/02 The ipliteral router was not recognizing addresses of the form user@
1420 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1422 PH/03 Added disable_ipv6.
1424 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1425 decoded form, because it is most often used to construct To: headers
1426 lines in autoreplies, and the decoded form may well be syntactically
1427 invalid. However, $reply_address has leading white space removed, and all
1428 newlines turned into spaces so that the autoreply transport does not
1431 PH/05 If group was specified without a user on a router, and no group or user
1432 was specified on a transport, the group from the router was ignored.
1434 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1435 for visible compile-time settings that can be used to change these
1436 numbers, for those that want even more. Backwards compatibility with old
1437 spool files has been maintained. However, going back to a previous Exim
1438 release will lost any variables that are in spool files.
1440 PH/07 Two small changes when running in the test harness: increase delay when
1441 passing a TCP/IP connection to a new process, in case the original
1442 process has to generate a bounce, and remove special handling of
1443 127.0.0.2 (sic), which is no longer necessary.
1445 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1446 be the same on different OS.
1448 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1451 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1452 whether --show-vars was specified or not
1454 JJ/02 exipick: Added support for new ACL variable spool format introduced
1457 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1458 syntactically invalid From: or Reply-to: line, and a filter used this to
1459 generate an autoreply, and therefore failed to obtain an address for the
1460 autoreply, Exim could try to deliver to a non-existent relative file
1461 name, causing unrelated and misleading errors. What now happens is that
1462 it logs this as a hard delivery error, but does not attempt to create a
1465 PH/11 The exinext utility has a -C option for testing purposes, but although
1466 the given file was scanned by exinext itself; it wasn't being passed on
1467 when Exim was called.
1469 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1470 an end-of-file indication when reading a command response.
1472 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1473 compiled. In many other places in Exim, IPv6 addresses are always
1474 recognized, so I have changed this. It also means that IPv4 domain
1475 literals of the form [IPV4:n.n.n.n] are now always recognized.
1477 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1478 used if the router is not running as root, for example, when verifying at
1479 ACL time, or when using -bh. The debugging output from this situation was
1480 non-existent - all you got was a failure to exec. I have made two
1483 (a) Failures to set uid/gid, the current directory, or a process leader
1484 in a subprocess such as that created by queryprogram now generate
1485 suitable debugging ouput when -d is set.
1487 (b) The queryprogram router detects when it is not running as root,
1488 outputs suitable debugging information if -d is set, and then runs
1489 the subprocess without attempting to change uid/gid.
1491 PH/15 Minor change to Makefile for building test_host (undocumented testing
1494 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1495 additional section of a DNS packet that returns MX or SRV records.
1496 Instead, it always explicitly searches for A/AAAA records. This avoids
1497 major problems that occur when a DNS server includes only records of one
1498 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1499 fixed another bug: if SRV records were looked up and the corresponding
1500 address records were *not* found in the additional section, the port
1501 values from the SRV records were lost.
1503 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1504 using the correct key (the original address) when searching the retry
1505 rules in order to find which one to use for generating the retry hint.
1507 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1508 adding the default one. Similarly, if it contains a Reply-To: header, the
1509 errors_reply_to option, if set, is not used.
1511 PH/19 When calculating a retry time, Exim used to measure the "time since
1512 failure" by looking at the "first failed" field in the retry record. Now
1513 it does not use this if it is later than than the arrival time of the
1514 message. Instead it uses the arrival time. This makes for better
1515 behaviour in cases where some deliveries succeed, thus re-setting the
1516 "first failed" field. An example is a quota failure for a huge message
1517 when small messages continue to be delivered. Without this change, the
1518 "time since failure" will always be short, possible causing more frequent
1519 delivery attempts for the huge message than are intended.
1520 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1522 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1523 $1, $2, $3) because the numerical variables can be reset during some
1524 expansion items (e.g. "match"), thereby losing the authentication data.
1526 PH/21 Make -bV show the size of off_t variables so that the test suite can
1527 decide whether to run tests for quotas > 2G.
1529 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1530 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1531 filecount value is greater than 2G or if a quota value is greater than 2G
1532 on a system where the size of off_t is not greater than 4, a panic error
1535 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1536 never match. The debug and -bh output now contains an explicit error
1537 message indicating a malformed IPv4 address or mask.
1539 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1540 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1541 PH/23 above applies.
1543 PH/25 Do not write to syslog when running in the test harness. The only
1544 occasion when this arises is a failure to open the main or panic logs
1545 (for which there is an explicit test).
1547 PH/26 Added the /no_tell option to "control=freeze".
1549 PH/27 If a host name lookup failed very early in a connection, for example, if
1550 the IP address matched host_lookup and the reverse lookup yielded a name
1551 that did not have a forward lookup, an error message of the form "no IP
1552 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1553 could be logged. Now it outputs the IP address instead of "NULL".
1555 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1556 allows the sender and the authenticated sender to be set when
1557 submitting a message from within Exim. Since child_open_exim() is
1558 documented for local_scan(), the new function should be too.
1560 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1561 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1562 results in an empty string is now treated as unset.
1564 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1566 PH/31 Added sender_verify_fail logging option.
1568 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1569 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1570 tidied the source and removed it altogether.
1572 PH/33 When a queue run was abandoned because the load average was too high, a
1573 log line was always written; now it is written only if the queue_run log
1574 selector is set. In addition, the log line for abandonment now contains
1575 information about the queue run such as the pid. This is always present
1576 in "start" and "stop" lines but was omitted from the "abandon" line.
1578 PH/34 Omit spaces between a header name and the colon in the error message that
1579 is given when verify = headers_syntax fails (if there are lots of them,
1580 the message gets confusing).
1582 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1583 names, as there are now some PTR records that contain slashes. This check
1584 is only to protect against broken name servers that fall over on strange
1585 characters, so the fact that it applies to all lookups doesn't matter.
1587 PH/36 Now that the new test suite is complete, we can remove some of the
1588 special code in Exim that was needed for the old test suite. For example,
1589 sorting DNS records because real resolvers return them in an arbitrary
1590 order. The new test suite's fake resolver always returns records in the
1593 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1594 bounces) except when queue_only is set, to avoid logging races between
1595 the different processes.
1597 PH/38 Panic-die if .include specifies a non-absolute path.
1599 PH/39 A tweak to the "H" retry rule from its user.
1601 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1602 a label. They prevented compilation on older perls.
1604 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1605 a warning to be raised on newish perls.
1607 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1608 on queue. Changes to match documented behaviour of showing count of
1609 messages matching specified criteria.
1611 PH/40 Changed the default ident timeout from 30s to 5s.
1613 PH/41 Added support for the use of login_cap features, on those BSD systems
1614 that have them, for controlling the resources used by pipe deliveries.
1616 PH/42 The content-scanning code uses fopen() to create files in which to put
1617 message data. Previously it was not paying any attention to the mode of
1618 the files. Exim runs with umask(0) because the rest of the code creates
1619 files with open(), and sets the required mode explicitly. Thus, these
1620 files were ending up world-writeable. This was not a big issue, because,
1621 being within the spool directory, they were not world-accessible. I have
1622 created a function called modefopen, which takes an additional mode
1623 argument. It sets umask(777), creates the file, chmods it to the required
1624 mode, then resets the umask. All the relevant calls to fopen() in the
1625 content scanning code have been changed to use this function.
1627 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1628 to 24 hours. This avoids potential overflow problems when processing G
1629 and H retry rules. I suspect nobody ever tinkers with this value.
1631 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1633 PH/45 When the plaintext authenticator is running as a client, the server's
1634 challenges are checked to ensure they are valid base64 strings. By
1635 default, the authentication attempt is cancelled if an invalid string is
1636 received. Setting client_ignore_invalid_base64 true ignores these errors.
1637 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1638 they are received. Thus, the responses can be made to depend on the
1639 challenges. If an invalid string is ignored, an empty string is placed in
1642 PH/46 Messages that are created by the autoreply transport now contains a
1643 References: header, in accordance with RFCs 2822 and 3834.
1645 PH/47 Added authenticated_sender_force to the smtp transport.
1647 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1649 PH/49 Installed latest patch from the Sieve maintainer.
1651 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1652 was also set, the appendfile transport was unnecessarily scanning a
1653 directory of message files (e.g. for maildir delivery) to find the count
1654 of files (along with the size), even though it did not need this
1655 information. It now does the scan only if it needs to find either the
1656 size of the count of files.
1658 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1660 PH/52 Two bugs concerned with error handling when the smtp transport is
1663 (i) Exim was not creating retry information for temporary errors given
1664 for individual recipients after the DATA command when the smtp transport
1665 was used in LMTP mode. This meant that they could be retried too
1666 frequently, and not timed out correctly.
1668 (ii) Exim was setting the flag that allows error details to be returned
1669 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1670 recipients that were returned after the DATA command.
1672 PH/53 This is related to PH/52, but is more general: for any failing address,
1673 when detailed error information was permitted to be returned to the
1674 sender, but the error was temporary, then after the final timeout, only
1675 "retry timeout exceeded" was returned. Now it returns the full error as
1676 well as "retry timeout exceeded".
1678 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1679 do this, and (what is worse) MTAs that accept it.
1681 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1682 will now be deprecated.
1684 PH/56 New os.c-cygwin from the Cygwin maintainer.
1686 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1687 formats (previously only available in exim formats via -bpr, -bpru,
1688 and -bpra. Now also available in native and exiqgrep formats)
1690 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1691 with very large, slow to parse queues
1693 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1695 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1697 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1698 responses to authentication challenges, though it was showing the
1699 challenges; (ii) I've removed the CR characters from the debug output for
1702 PH/58 Allow for the insertion of a newline as well as a space when a string
1703 is turned into more than one encoded-word during RFC 2047 encoding. The
1704 Sieve code now uses this.
1706 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1707 data_4xx, lost_connection, tls_required.
1709 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1710 message was being sent as an SMTP response.
1712 PH/61 Add -l and -k options to exicyclog.
1714 PH/62 When verifying, if an address was redirected to one new address, so that
1715 verification continued, and the new address failed or deferred after
1716 having set something in $address_data, the value of $address_data was not
1717 passed back to the ACL. This was different to the case when no
1718 redirection occurred. The value is now passed back in both cases.
1720 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1721 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1722 use login_cap.h, so on its own it isn't the distinguishing feature. The
1723 new name refers directly to the setclassresources() function.
1725 PH/65 Added configuration files for NetBSD3.
1727 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1729 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1730 is preferred over IPv4.
1732 PH/68 The bounce_return_message and bounce_return_body options were not being
1733 honoured for bounces generated during the reception of non-SMTP messages.
1734 In particular, this applied to messages rejected by the ACL. This bug has
1735 been fixed. However, if bounce_return_message is true and bounce_return_
1736 body is false, the headers that are returned for a non-SMTP message
1737 include only those that have been read before the error was detected.
1738 (In the case of an ACL rejection, they have all been read.)
1740 PH/69 The HTML version of the specification is now built in a directory called
1741 spec_html instead of spec.html, because the latter looks like a path with
1742 a MIME-type, and this confuses some software.
1744 PH/70 Catch two compiler warnings in sieve.c.
1746 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1747 function verify_get_ident() calls ip_connect() to connect a socket, but
1748 if the "connect()" function timed out, ip_connect() used to close the
1749 socket. However, verify_get_ident() also closes the socket later, and in
1750 between Exim writes to the log, which may get opened at this point. When
1751 the socket was closed in ip_connect(), the log could get the same file
1752 descriptor number as the socket. This naturally causes chaos. The fix is
1753 not to close the socket in ip_connect(); the socket should be closed by
1754 the function that creates it. There was only one place in the code where
1755 this was missing, in the iplookup router, which I don't think anybody now
1756 uses, but I've fixed it anyway.
1758 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1759 well as to direct DNS lookups. Otherwise the handling of names in host
1760 lists is inconsistent and therefore confusing.
1766 PH/01 Two changes to the default runtime configuration:
1768 (1) Move the checks for relay_from_hosts and authenticated clients from
1769 after to before the (commented out) DNS black list checks.
1771 (2) Add control=submission to the relay_from_hosts and authenticated
1772 clients checks, on the grounds that messages accepted by these
1773 statements are most likely to be submissions.
1775 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1777 (1) Generate an error if the third argument for the ${prvs expansion is
1780 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1783 (3) Reset the variables that are obtained from the first argument of
1784 ${prvscheck and used in the second argument before leaving the code,
1785 because their memory is reclaimed, so using them afterwards may do
1788 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1789 one (it's much easier than Tom thought :-).
1791 (5) Because of (4), we can now allow for the use of $prvscheck_result
1792 inside the third argument.
1794 PH/03 For some reason, the default setting of PATH when running a command from
1795 a pipe transport was just "/usr/bin". I have changed it to
1798 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1799 anything to be listed in the output from -bV.
1801 PH/05 When a filter generated an autoreply, the entire To: header line was
1802 quoted in the delivery log line, like this:
1804 => >A.N.Other <ano@some.domain> <original@ddress> ...
1806 This has been changed so that it extracts the operative address. There
1807 may be more than one such address. If so, they are comma-separated, like
1810 => >ano@some.domain,ona@other.domain <original@ddress> ...
1812 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1813 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1814 not being looked up in the rDNS to get a host name, Exim was showing the
1815 IP address twice in Received: lines, even though the IP addresses were
1816 identical. For example:
1818 Received: from [1.2.3.4] (helo=[1.2.3.4])
1820 However, if the real host name was known, it was omitting the HELO data
1821 if it matched the actual IP address. This has been tidied up so that it
1822 doesn't show the same IP address twice.
1824 PH/07 When both +timestamp and +memory debugging was on, the value given by
1825 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1826 called by the debug printing, thereby overwriting the timestamp buffer.
1827 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1829 PH/08 When the original message was included in an autoreply transport, it
1830 always said "this is a copy of the message, including all the headers",
1831 even if body_only or headers_only was set. It now gives an appropriate
1834 PH/09 Applied a patch from the Sieve maintainer which:
1836 o fixes some comments
1837 o adds the (disabled) notify extension core
1838 o adds some debug output for the result of if/elsif tests
1839 o points to the current vacation draft in the documentation
1840 and documents the missing references header update
1844 o fixes a bug in processing the envelope test (when testing
1845 multiple envelope elements, the last element determinted the
1848 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1849 Electronic Mail") by including:
1851 Auto-submitted: auto-generated
1853 in the messages that it generates (bounce messages and others, such as
1854 warnings). In the case of bounce messages for non-SMTP mesages, there was
1855 also a typo: it was using "Auto_submitted" (underscore instead of
1856 hyphen). Since every message generated by Exim is necessarily in response
1857 to another message, thes have all been changed to:
1859 Auto-Submitted: auto-replied
1861 in accordance with these statements in the RFC:
1863 The auto-replied keyword:
1865 - SHOULD be used on messages sent in direct response to another
1866 message by an automatic process,
1868 - MUST NOT be used on manually-generated messages,
1870 - MAY be used on Delivery Status Notifications (DSNs) and Message
1871 Disposition Notifications (MDNs),
1873 - MUST NOT be used on messages generated by automatic or periodic
1874 processes, except for messages which are automatic responses to
1877 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1878 to the default Received: header definition.
1880 PH/12 Added log selector acl_warn_skipped (default on).
1882 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1883 variables because (a) they are in the wrong storage pool and (b) even if
1884 they were copied, it wouldn't work properly because of the caching.
1886 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1887 checking when decoding. Apparently there are clients that generate
1888 overlong encoded strings. Why am I not surprised?
1890 PH/15 If the first argument of "${if match_address" was not empty, but did not
1891 contain an "@" character, Exim crashed. Now it writes a panic log message
1892 and treats the condition as false.
1894 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1896 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1897 extension "envelope-auth". The code is finished and in agreement with
1898 other implementations, but there is no documentation so far and in fact,
1899 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1900 not changing the active code.
1902 Print executed "if" and "elsif" statements when debugging is used. This
1903 helps a great deal to understand what a filter does.
1905 Document more things not specified clearly in RFC3028. I had all this
1906 sorted out, when out of a sudden new issues came to my mind. Oops."
1908 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1911 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1912 canonical form (as documented). However, after a host name lookup from
1913 the IP address, check_host() was doing a simple string comparison with
1914 addresses acquired from the DNS when checking that the found name did
1915 have the original IP as one of its addresses. Since any found IPv6
1916 addresses are likely to be in abbreviated form, the comparison could
1917 fail. Luckily, there already exists a function for doing the comparison
1918 by converting both addresses to binary, so now that is used instead of
1919 the text comparison.
1921 PH/20 There was another similar case to PH/19, when a complete host name was
1922 given in a host list; looking up its IP address could give an abbreviated
1923 form, whereas the current host's name might or might not be abbreviated.
1924 The same fix has been applied.
1930 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1931 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1934 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1935 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1937 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1939 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1940 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1941 Also, for "Auto-Submitted", treat anything other than "no" as
1942 non-personal, in accordance with RFC 3834. (Previously it treated
1943 anything starting "auto-" as non-personal.)
1945 TF/01 The control=submission/name=... option had a problem with syntax
1946 errors if the name included a slash character. The /name= option
1947 now slurps the rest of the string, so it can include any characters
1948 but it must come last in the list of options (after /sender_retain
1951 PH/05 Some modifications to the interface to the fake nameserver for the new
1959 TK/01 Added the "success_on_redirect" address verification option. See
1960 NewStuff for rationale and an example.
1962 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1964 PH/02 Patch to exigrep to allow it to work on syslog lines.
1966 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1967 fread() to skip over the body file's header line, because in Cygwin the
1968 header line is locked and is inaccessible.
1970 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1971 co-exist for some time) to make it clear that it is the Exim ID that is
1972 referenced, not the Message-ID: header line.
1974 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1975 string_format() function, because snprintf() does not exist on all
1978 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1979 ${stat: expansion item.
1981 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1982 protocol synchronization error", to keep the pedants happy.
1984 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1985 well as for IRIX systems, when gcc is being used. See the host.c source
1988 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1990 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1993 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1994 transport and to the smtp transport in LMTP mode.
1996 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1998 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1999 run a filter in a subprocess. This could lead to confusion in subsequent
2000 lookups in the parent process. There should also be a search_tidyup() at
2001 the end of the subprocess.
2003 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2004 only if the host matched helo_try_verify_hosts, which caused the
2005 verification to occur when the EHLO/HELO command was issued. The ACL just
2006 tested the remembered result. Now, if a previous verification attempt has
2007 not happened, "verify = helo" does it there and then.
2009 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2011 TK/03 Fix log output including CR from clamd.
2013 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2014 exist provoked a memory error which could cause a segfault.
2016 PH/15 Installed PCRE 6.2
2018 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2020 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2021 of the problem. Specifically, suggested +O2 rather than +O1 for the
2024 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2026 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2027 record, the port was not correctly logged when the outgoing_port log
2028 selector was set (it logged the transort's default port).
2030 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2031 fallback_hosts, and "hosts" in the smtp transport.
2033 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2034 host errors such as "Connection refused".
2036 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2037 authentication with radiusclient 0.4.9:
2039 - Error returned from rc_read_config was caught wrongly
2040 - Username/password not passed on to radius server due to wrong length.
2042 The presumption is that some radiusclient API changes for 4.51/PH/17
2043 were not taken care of correctly. The code is still untested by me (my
2044 Linux distribution still has 0.3.2 of radiusclient), but it was
2045 contributed by a Radius user.
2047 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2048 expanding the "port" option of the smtp transport.
2050 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2051 while reading a MIME header. Thanks to Tom Hughes for a patch.
2053 PH/24 Include config.h inside local_scan.h so that configuration settings are
2056 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2057 that in an ACL for RCPT (for example), you can examine exactly what was
2060 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2061 commands, but it was not correctly comparing the address with the actual
2062 client host address. Thus, it would show the EHLO address in Received:
2063 header lines when this was not necessary.
2065 PH/27 Added the % operator to ${eval:}.
2067 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2068 it should be ignoring failures (because with -C, for example, it has lost
2069 privilege). It wasn't ignoring creation failures other than "already
2072 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2075 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2076 because an input line was too long, either on its own, or by virtue of
2077 too many continuations, the temporary file was not being removed, and the
2078 return code was incorrect.
2080 PH/31 Missing "BOOL" in function definition in filtertest.c.
2082 PH/32 Applied Sieve patches from the maintainer.
2084 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2086 PH/33 Added "verify = not_blind".
2088 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2089 Local/Makefile (with some defaults set). These are used in built scripts
2090 such as exicyclog, but they have never been used in the exim_install
2091 script (though there are many overriding facilities there). I have
2092 arranged that the exim_install script now takes note of these two
2095 PH/35 Installed configuration files for Dragonfly.
2097 PH/36 When a locally submitted message by a trusted user did not contain a
2098 From: header, and the sender address was obtained from -f or from an SMTP
2099 MAIL command, and the trusted user did not use -F to supply a sender
2100 name, $originator_name was incorrectly used when constructing a From:
2101 header. Furthermore, $originator_name was used for submission mode
2102 messages from external hosts without From: headers in a similar way,
2103 which is clearly wrong.
2105 PH/37 Added control=suppress_local_fixups.
2107 PH/38 When log_selector = +received_sender was set, and the addition of the
2108 sender made the log line's construction buffer exactly full, or one byte
2109 less than full, an overflow happened when the terminating "\n" was
2112 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2113 when the result of a list match is failure because a DNS lookup failed.
2115 PH/40 RM_COMMAND is now used in the building process.
2117 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2118 the "build-* directories that it finds.
2120 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2121 address in a domain literal was a prefix of an interface address.
2123 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2124 when verifying a sender address, unless rewrite_headers is false.
2126 PH/44 Wrote a long comment about why errors_to addresses are verified as
2127 recipients, not senders.
2129 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2130 the ratelimit ACL was added.
2132 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2134 PH/46 Added extra information about PostgreSQL errors to the error string.
2136 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2137 avoiding the need to install special zones in a real server. This is
2138 backwards compatible; if it can't find the fake resolver, it drops back.
2139 Thus, both old and new test suites can be run.
2141 TF/02 Added util/ratelimit.pl
2143 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2144 clock is set back in time.
2146 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2147 Candler <B.Candler@pobox.com>.
2149 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2150 OK for addresses that are the result of redirections.
2152 PH/48 A number of further additions for the benefit of the new test suite,
2153 including a fake gethostbyname() that interfaces to the fake DNS resolver
2156 TF/06 The fix for widen_domains has also been applied to qualify_single and
2157 search_parents which are the other dnslookup options that can cause
2160 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2163 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2165 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2166 service" if it encountered an MX record with an empty target hostname.
2167 The message is now "an MX or SRV record indicated no SMTP service".
2169 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2170 if the DNS of the sending site is misconfigured. This is quite a
2171 common situation. This change restores the behaviour of treating a
2172 helo verification defer as a failure.
2174 PH/51 If self=fail was set on a router, the bounce message did not include the
2175 actual error message.
2181 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2183 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2184 command itself ended in error, the underlying message about the transport
2185 filter timeout was being overwritten with the pipe command error. Now the
2186 underlying error message should be appended to the second error message.
2188 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2190 PH/02 Reduce the amount of output that "make" produces by default. Full output
2191 can still be requested.
2193 PH/03 The warning log line about a condition test deferring for a "warn" verb
2194 was being output only once per connection, rather than after each
2195 occurrence (because it was using the same function as for successful
2196 "warn" verbs). This seems wrong, so I have changed it.
2198 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2199 it should not have, which might have caused a crash in the right
2200 circumstances, but probably never did.
2202 PH/04 Installed a modified version of Tony Finch's patch to make submission
2203 mode fix the return path as well as the Sender: header line, and to
2204 add a /name= option so that you can make the user's friendly name appear
2207 TF/03 Added the control = fakedefer ACL modifier.
2209 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2210 Mark Lowes for thorough testing.
2212 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2214 TK/03 Merged latest SRS patch from Miles Wilton.
2216 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2217 with the definition in sysexits.h (which is #included earlier).
2218 Fortunately, Exim does not actually use EX_OK. The code used to try to
2219 preserve the sysexits.h value, by assumimg that macro definitions were
2220 scanned for macro replacements. I have been disabused of this notion,
2221 so now the code just undefines EX_OK before #including unistd.h.
2223 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2224 in the smtp transport. When a block could not be written in a single
2225 write() function, the timeout was being re-applied to each part-write.
2226 This seems wrong - if the receiver was accepting one byte at a time it
2227 would take for ever. The timeout is now adjusted when this happens. It
2228 doesn't have to be particularly precise.
2230 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2231 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2233 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2234 without a domain if the check to <postmaster@domain> fails.
2236 SC/01 Eximstats: added -xls and the ability to specify output files
2237 (patch written by Frank Heydlauf).
2239 SC/02 Eximstats: use FileHandles for outputing results.
2241 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2243 SC/04 Eximstats: fixed display of large numbers with -nvr option
2245 SC/05 Eximstats: fixed merging of reports with empty tables.
2247 SC/06 Eximstats: added the -include_original_destination flag
2249 SC/07 Eximstats: removed tabs and trailing whitespace.
2251 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2253 TK/06 MBOX spool code: Add real "From " MBOX separator line
2254 so the .eml file is really in mbox format (even though
2255 most programs do not really care). Patch from Alex Miller.
2257 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2258 The latter is generated from $received_to and is only set if the
2259 message has one envelope recipient. SA can use these headers,
2260 obviously out-of-the-box. Patch from Alex Miller.
2262 PH/08 The ${def test on a variable was returning false if the variable's
2263 value was "0", contrary to what the specification has always said!
2264 The result should be true unless the variable is empty.
2266 PH/09 The syntax error of a character other than { following "${if
2267 def:variable_name" (after optional whitespace) was not being diagnosed.
2268 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2269 accidental colon was present, for example, could give incorrect results.
2271 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2272 result is used (not including appendfile, where other changes are about
2275 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2276 This involved changing a lot of size variables from int to off_t. It
2277 should work with maildirs and everything.
2279 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2280 spamd dying while we are connected to it.
2282 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2285 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2286 function for BDB 4.3.
2288 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2290 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2291 the sources for PCRE have been re-arranged and more files are now
2294 PH/15 The code I had for printing potentially long long variables in PH/11
2295 above was not the best (it lost precision). The length of off_t variables
2296 is now inspected at build time, and an appropriate printing format (%ld
2297 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2298 to be "long long int" or "long int". This is needed for the internal
2299 formatting function string_vformat().
2301 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2302 the configuration file to be ":syslog", then the script "guesses" where
2303 the logs files are, rather than using the compiled in default. In our
2304 case the guess is not the same as the compiled default, so the script
2305 suddenly stopped working when I started to use syslog. The patch checks
2306 to see if log_file_path is "". If so, it attempts to read it from exim
2307 with no configuration file to get the compiled in version, before it
2308 falls back to the previous guessing code."
2310 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2311 implementing BATV in an Exim configuration. See NewStuff for the gory
2314 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2315 Makefile that are specific to HP-UX.
2317 PH/18 If the "use_postmaster" option was set for a recipient callout together
2318 with the "random" option, the postmaster address was used as the MAIL
2319 FROM address for the random test, but not for the subsequent recipient
2320 test. It is now used for both.
2322 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2323 patch removes a few documentation additions to RFC 3028, because the
2324 latest draft now contains them. It adds the new en;ascii-case comparator
2325 and a new error check for 8bit text in MIME parts. Comparator and
2326 require names are now matched exactly. I enabled the subaddress
2327 extension, but it is not well tested yet (read: it works for me)."
2329 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2330 rework some of the code of TK/09 above to avoid the hardwired use of
2331 "%lld" and "long long". Replaced the call to snprintf() with a call to
2334 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2335 records point to non-existent hosts", "retry timeout exceeded", and
2336 "retry time not reached for any host after a long failure period".
2338 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2339 experimental DomainKeys support:
2341 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2342 (2) On an error such as an illegally used "control", the wrong name for
2343 the control was given.
2345 These problems did NOT occur unless DomainKeys support was compiled.
2347 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2349 PH/24 Added ${if match_ip condition.
2351 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2352 databases so that it will be absolutely obvious if a crash occurs in the
2353 DB library. This is a regular occurrence (often caused by mis-matched
2356 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2357 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2358 picked up on a user's system that detects such things. There doesn't seem
2359 to be a gcc warning option for this - only an attribute that has to be
2360 put on the function's prototype. It seems that in Fedora Core 4 they have
2361 set this on a number of new functions. No doubt there will be more in due
2364 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2365 specify a transport. However, if an address that was verified by such a
2366 router was the subject of a callout, Exim crashed because it tried to
2367 read the rcpt_include_affixes from the non-existent transport. Now it
2368 just assumes that the setting of that option is false. This bug was
2369 introduced by 4.51/PH/31.
2371 PH/28 Changed -d+all to exclude +memory, because that information is very
2372 rarely of interest, but it makes the output a lot bigger. People tend to
2373 do -d+all out of habit.
2375 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2376 code in os-type was giving problems when libc.so lives in lib64, like on
2379 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2380 aren't the modern standard, and it seems that some systems' include files
2381 don't always have them. Exim was already checking for some of the newer
2382 ones like T_AAAA, and defining it itself. I've added checks for all the
2383 record types that Exim uses.
2385 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2386 not automatically generating a new one, as it is supposed to. This
2387 prevented TLS from working. If the file did exist, but contained invalid
2388 data, a new version was generated, as expected. It was only the case of a
2389 non-existent file that was broken.
2391 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2392 with a change in libdomainkeys > 0.64.
2394 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2395 from DNS. If the selector record carries the flag, it now has
2396 precedence over the domain-wide flag.
2398 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2400 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2401 the use of a port name that isn't defined in /etc/services) occurred, the
2402 message was deferred as in a normal delivery, and thus remained on the
2403 spool, instead of being failed because of the mua_wrapper setting. This
2404 is now fixed, and I tidied up some of the mua_wrapper messages at the
2407 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2408 the messages in a hash of arrays rather than using individual hashes.
2409 This is a bit cleaner and results in dramatic memory savings, albeit
2410 at a slight CPU cost.
2412 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2413 as requested by Marc Sherman.
2415 SC/10 Eximstats: added histograms for user specified patterns as requested
2418 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2420 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2421 fopen() in the content-scanning modules that did not already have it.
2427 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2428 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2430 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2432 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2434 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2435 to test Sieve filters that use "vacation".
2437 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2438 that changes the way the GnuTLS parameters are stored in the cache file.
2439 The new format can be generated externally. For backward compatibility,
2440 if the data in the cache doesn't make sense, Exim assumes it has read an
2441 old-format file, and it generates new data and writes a new file. This
2442 means that you can't go back to an older release without removing the
2445 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2446 work if there are any delivery delays because "one_time" forces the
2447 parent to be marked "delivered", so its unseen clone is never tried
2448 again. For this reason, Exim now forbids the simultaneous setting of
2451 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2452 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2453 entry if you want to know the details. The fix, however, neglected to
2454 consider the case when local delivery batching is involved. The test for
2455 "previously delivered" was not happening when checking to see if an
2456 address could be batched with a previous (undelivered) one; under
2457 certain circumstances this could lead to multiple deliveries to the same
2460 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2461 in its include files, and this causes problems building Exim.
2463 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2464 header_syntax) but Exim was just ignoring anything given after a slash.
2465 In particular, this caused confusion with an attempt to use "verify =
2466 reverse_host_lookup/defer_ok". An error is now given when options are
2467 supplied for verify items that do not have them. (Maybe reverse_host_
2468 lookup should have a defer_ok option, but that's a different point.)
2470 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2471 defined by RFC 821) to 2048, because there were problems with some AUTH
2472 commands, and RFC 1869 says the size should be increased for extended
2473 SMTP commands that take arguments.
2475 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2478 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2479 "unknown" error; now it says that the functionality isn't in the binary.
2481 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2482 an address' error message when a string expansion fails (syntax or
2483 whatever). Otherwise the password may appear in the log. Following change
2484 PH/42 below, there is no longer a chance of it appearing in a bounce
2487 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2489 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2490 of its addresses was ever tried. (Bugzilla bug #2.)
2492 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2493 the result incorrectly in the debug output. (It correctly added a newline
2494 to what was transported.)
2496 TF/01 Added $received_time.
2498 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2499 commented out examples of how to interface to a virus scanner and to
2500 SpamAssassin. Also added commented examples of av_scanner and
2501 spamd_address settings.
2503 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2504 and controls are allowed in which ACLs. There were a couple of minor
2505 errors. Some of the entries in the conditions table (which is a table of
2506 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2507 negation of where the condition IS allowed.
2509 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2511 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2512 header file does not have a version number, so I've had to invent a new
2513 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2514 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2515 radiusclient), but it was contributed by a Radius user.
2517 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2518 files or directories, for OpenSSL.
2520 PH/19 When an Exim process that is running as root has to create an Exim log
2521 file, it does so in a subprocess that runs as exim:exim so as to get the
2522 ownership right at creation (otherwise, other Exim processes might see
2523 the file with the wrong ownership). There was no test for failure of this
2524 fork() call, which would lead to the process getting stuck as it waited
2525 for a non-existent subprocess. Forks do occasionally fail when resources
2526 run out. I reviewed all the other calls to fork(); they all seem to check
2529 PH/20 When checking for unexpected SMTP input at connect time (before writing
2530 the banner), Exim was not dealing correctly with a non-positive return
2531 from the read() function. If the client had disconnected by this time,
2532 the result was a log entry for a synchronization error with an empty
2533 string after "input=" when read() returned zero. If read() returned -1
2534 (an event I could not check), uninitialized data bytes were printed.
2535 There were reports of junk text (parts of files, etc) appearing after
2538 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2540 PH/22 Added support for macro redefinition, and (re)definition in between
2541 driver and ACL definitions.
2543 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2544 forgetting to use the resulting value; it was using the unexpanded value.
2546 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2547 hadn't been configured. The fix is from Juergen Kreileder, who
2548 understands it better than I do:
2550 "Here's what I see happening with three configured cyrus_sasl
2551 authenticators configured (plain, login, cram-md5):
2553 On startup auth_cyrus_sasl_init() gets called for each of these.
2554 This means three calls to sasl_listmech() without a specified mech_list.
2555 => SASL tests which mechs of all available mechs actually work
2556 => three warnings about OTP not working
2557 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2559 With the patch, sasl_listmech() also gets called three times. But now
2560 SASL's mech_list option is set to the server_mech specified in the the
2561 authenticator. Or in other words, the answer from sasl_listmech()
2562 gets limited to just the mech you're testing for (which is different
2564 => the return list contains just 'plain' or 'login', 'cram-md5' or
2565 nothing depending on the value of ob->server_mech.
2567 I've just tested the patch: Authentication still works fine,
2568 unavailable mechs specified in the exim configuration are still
2569 caught, and the auth.log warnings about OTP are gone."
2571 PH/25 When debugging is enabled, the contents of the command line are added
2572 to the debugging output, even when log_selector=+arguments is not
2575 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2576 answer is "GNU", and only if the return is "GNU/something" is the answer
2579 PH/27 $acl_verify_message is now set immediately after the failure of a
2580 verification in an ACL, and so is available in subsequent modifiers. In
2581 particular, the message can be preserved by coding like this:
2583 warn !verify = sender
2584 set acl_m0 = $acl_verify_message
2586 Previously, $acl_verify_message was set only while expanding "message"
2587 and "log_message" when a very denied access.
2589 PH/28 Modified OS/os.c-Linux with
2591 -#ifndef OS_LOAD_AVERAGE
2592 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2594 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2595 nomenclature these days.)
2597 PH/29 Installed patch from the Sieve maintainer that adds the options
2598 sieve_useraddress and sieve_subaddress to the redirect router.
2600 PH/30 In these circumstances:
2601 . Two addresses routed to the same list of hosts;
2602 . First host does not offer TLS;
2603 . First host accepts first address;
2604 . First host gives temporary error to second address;
2605 . Second host offers TLS and a TLS session is established;
2606 . Second host accepts second address.
2607 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2608 and peerdn, if requested) that were in fact used only for the second
2611 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2612 attention to any local part prefix or suffix that was matched by the
2613 router that accepted the address. It now behaves in the same way as it
2614 does for delivery: the affixes are removed from the local part unless
2615 rcpt_include_affixes is set on the transport.
2617 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2618 timeout during the DATA phase of an incoming message.
2620 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2621 applied a patch sanctioned by the Sieve maintainer.
2623 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2624 the uid or gid is negative. A case of a negative gid caused this to be
2625 noticed. The fix allows for either to be negative.
2627 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2628 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2631 PH/36 The change PH/12 above was broken. Fixed it.
2633 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2634 the grounds that routing the same address twice would always produce the
2635 same answer. This might have been true once, but it is certainly no
2636 longer true now. Routing a child address may depend on the previous
2637 routing that produced that child. Some complicated redirection strategies
2638 went wrong when messages had multiple recipients, and made Exim's
2639 behaviour dependent on the order in which the addresses were given.
2641 I have moved the duplicate checking until after the routing is complete.
2642 Exim scans the addresses that are assigned to local and remote
2643 transports, and removes any duplicates. This means that more work will be
2644 done, as duplicates will always all be routed, but duplicates are
2645 presumably rare, so I don't expect this is of any significance.
2647 For deliveries to pipes, files, and autoreplies, the duplicate checking
2648 still happens during the routing process, since they are not going to be
2651 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2652 It corrects a timeout issue with spamd. This is Ian's comment: "The
2653 background is that sometimes spamd either never reads data from a
2654 connection it has accepted, or it never writes response data. The exiscan
2655 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2656 blindly assumes that writes won't block so it may never time out."
2658 PH/39 Allow G after quota size as well as K and M.
2660 PH/40 The value set for $authenticated_id in an authenticator may not contain
2661 binary zeroes or newlines because the value is written to log lines and
2662 to spool files. There was no check on this. Now the value is run through
2663 the string_printing() function so that such characters are converted to
2664 printable escape sequences.
2666 PH/41 $message_linecount is a new variable that contains the total number of
2667 lines in the message. Compare $body_linecount, which is the count for the
2670 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2671 bounce and delay warning messages, except in certain special cases, which
2674 (a) An SMTP error message from a remote host;
2675 (b) A message specified in a :fail: redirection;
2676 (c) A message specified in a "fail" command in a system filter;
2677 (d) A message specified in a FAIL return from the queryprogram router;
2678 (e) A message specified by the cannot_route_message router option.
2680 In these cases only, Exim does include the error details in bounce and
2681 warning messages. There are also a few cases where bland messages such
2682 as "unrouteable address" or "local delivery error" are given.
2684 PH/43 $value is now also set for the "else" part of a ${run expansion.
2686 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2687 being worked on, but at least Exim now implements the latest version to
2690 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2691 process to complete was treated as a delivery failure, a timeout while
2692 writing the message to the pipe was logged, but erroneously treated as a
2693 successful delivery. Such timeouts include transport filter timeouts. For
2694 consistency with the overall process timeout, these timeouts are now
2695 treated as errors, giving rise to delivery failures by default. However,
2696 there is now a new Boolean option for the pipe transport called
2697 timeout_defer, which, if set TRUE, converts the failures into defers for
2698 both kinds of timeout. A transport filter timeout is now identified in
2701 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2702 systems where "make" and "gmake" are different, calling "gmake" at top
2703 level broke things. I've arranged for the value of $(MAKE) to be passed
2704 from the Makefile to this script so that it can call the same version of
2708 A note about Exim versions 4.44 and 4.50
2709 ----------------------------------------
2711 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2712 changes of various kinds. As a consequence, a big documentation update was
2713 needed. This delayed the release for rather longer than seemed good, especially
2714 in the light of a couple of (minor) security issues. Therefore, the changes
2715 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2716 release. So 4.44 and 4.50 are in effect two different branches that both start
2719 I have left the 4.50 change log unchanged; it contains all the changes since
2720 4.43. The change log for 4.44 is below; many of its items are identical to
2721 those for 4.50. This seems to be the most sensible way to preserve the
2722 historical information.
2728 1. Minor wording change to the doc/README.SIEVE file.
2730 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2731 computation of the current number of files was incorrect.
2733 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2734 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2737 4. Give more explanation in the error message when the command for a transport
2738 filter fails to execute.
2740 5. There are several places where Exim runs a non-Exim command in a
2741 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2742 was being done only for the command run by the queryprogram router. It is
2743 now done for all such subprocesses. The other cases are: ${run, transport
2744 filters, and the commands run by the lmtp and pipe transports.
2746 6. Added CONFIGURE_GROUP build-time option.
2748 7. Some older OS have a limit of 256 on the maximum number of file
2749 descriptors. Exim was using setrlimit() to set 1000 as a large value
2750 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2751 systems. I've change it so that if it can't get 1000, it tries for 256.
2753 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2754 was an oversight, and furthermore, ever since the addition of extra
2755 controls (e.g. 4.43/32), the checks on when to allow different forms of
2756 "control" were broken. There should now be diagnostics for all cases when a
2757 control that does not make sense is encountered.
2759 9. Added the /retain_sender option to "control=submission".
2761 10. $recipients is now available in the predata ACL (oversight).
2763 11. Tidy the search cache before the fork to do a delivery from a message
2764 received from the command line. Otherwise the child will trigger a lookup
2765 failure and thereby defer the delivery if it tries to use (for example) a
2766 cached ldap connection that the parent has called unbind on.
2768 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2769 of $address_data from the recipient verification was clobbered by the
2770 sender verification.
2772 13. The value of address_data from a sender verification is now available in
2773 $sender_address_data in subsequent conditions in the ACL statement.
2775 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2777 15. Added a new option "connect=<time>" to callout options, to set a different
2780 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2781 was its contents. (It was OK if the option was not defined at all.)
2783 17. A "Completed" log line is now written for messages that are removed from
2784 the spool by the -Mrm option.
2786 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2787 information about exactly what failed.
2789 19. Added -dd to debug only the daemon process.
2791 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2792 handles timeouts, both on the server side and network timeouts. Renamed the
2793 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2795 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2796 It is now set to "smtps".
2798 22. $host_address is now set to the target address during the checking of
2799 ignore_target_hosts.
2801 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2802 being passed; this would have caused $sender_host_name to have been used if
2803 matching the list had actually called for a host name (not very likely,
2804 since this list is usually IP addresses). A host name is now passed as
2807 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2808 code that specifies a non-restarting handler (typically sigaction() in
2809 modern systems) in an attempt to fix a rare and obscure crash bug.
2811 25. Narrowed the window for a race in the daemon that could cause it to ignore
2812 SIGCHLD signals. This is not a major problem, because they are used only to
2813 wake it up if nothing else does.
2815 26. A malformed maildirsize file could cause Exim to calculate negative values
2816 for the mailbox size or file count. Odd effects could occur as a result.
2817 The maildirsize information is now recalculated if the size or filecount
2820 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2821 support for a long time. Removed HAVE_SYS_VFS_H.
2823 28. Installed the latest version of exipick from John Jetmore.
2825 29. In an address list, if the pattern was not a regular expression, an empty
2826 subject address (from a bounce message) matched only if the pattern was an
2827 empty string. Non-empty patterns were not even tested. This was the wrong
2828 because it is perfectly reasonable to use an empty address as part of a
2829 database query. An empty address is now tested by patterns that are
2830 lookups. However, all the other forms of pattern expect the subject to
2831 contain a local part and a domain, and therefore, for them, an empty
2832 address still always fails if the pattern is not itself empty.
2834 30. Exim went into a mad DNS loop when attempting to do a callout where the
2835 host was specified on an smtp transport, and looking it up yielded more
2836 than one IP address.
2838 31. Re-factored the code for checking spool and log partition space into a
2839 function that finds that data and another that does the check. The former
2840 is then used to implement four new variables: $spool_space, $log_space,
2841 $spool_inodes, and $log_inodes.
2843 32. The RFC2047 encoding function was originally intended for short strings
2844 such as real names; it was not keeping to the 75-character limit for
2845 encoded words that the RFC imposes. It now respects the limit, and
2846 generates multiple encoded words if necessary. To be on the safe side, I
2847 have increased the buffer size for the ${rfc2047: expansion operator from
2850 33. It is now permitted to omit both strings after an "if" condition; if the
2851 condition is true, the result is "true". As before, when the second string
2852 is omitted, a false condition yields an empty string. This makes it less
2853 cumbersome to write custom ACL and router conditions.
2855 34. Failure to deliver a bounce message always caused it to be frozen, even if
2856 there was an errors_to setting on the router. The errors_to setting is now
2859 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2860 canonical form (fully expanded) before being placed in
2861 $sender_host_address.
2863 36. The table in the code that translates DNS record types into text (T_A to
2864 "A" for instance) was missing entries for NS and CNAME. It is just possible
2865 that this could have caused confusion if both these types were looked up
2866 for the same domain, because the text type is used as part of Exim's
2867 per-process caching. But the chance of anyone hitting this buglet seems
2870 37. The dnsdb lookup has been extended in a number of ways.
2872 (1) There is a new type, "zns", which walks up the domain tree until it
2873 finds some nameserver records. It should be used with care.
2875 (2) There is a new type, "mxh", which is like "mx" except that it returns
2876 just the host names, not the priorities.
2878 (3) It is now possible to give a list of domains (or IP addresses) to be
2879 looked up. The behaviour when one of the lookups defers can be
2880 controlled by a keyword.
2882 (4) It is now possible to specify the separator character for use when
2883 multiple records are returned.
2885 38. The dnslists ACL condition has been extended: it is now possible to supply
2886 a list of IP addresses and/or domains to be looked up in a particular DNS
2889 39. Added log_selector=+queue_time_overall.
2891 40. When running the queue in the test harness, wait just a tad after forking a
2892 delivery process, to get repeatability of debugging output.
2894 41. Include certificate and key file names in error message when GnuTLS fails
2895 to set them up, because the GnuTLS error message doesn't include the name
2896 of the failing file when there is a problem reading it.
2898 42. Allow both -bf and -bF in the same test run.
2900 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2902 44. The "Exiscan patch" is now merged into the mainline Exim source.
2904 45. Sometimes the final signoff response after QUIT could fail to get
2905 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2906 < 0 before doing a fflush(). This bug looks as though it goes back to the
2907 introduction of TLS in release 3.20, but "sometimes" must have been rare
2908 because the tests only now provoked it.
2910 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2911 (this can affect the format of dates).
2913 47. exim_tidydb, when checking for the continued existence of a message for
2914 which it has found a message-specific retry record, was not finding
2915 messages that were in split spool directories. Consequently, it was
2916 deleting retry records that should have stayed in existence.
2918 48. Steve fixed some bugs in eximstats.
2920 49. The SPA authentication driver was not abandoning authentication and moving
2921 on to the next authenticator when an expansion was forced to fail,
2922 contradicting the general specification for all authenticators. Instead it
2923 was generating a temporary error. It now behaves as specified.
2925 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2926 (the order specifies the preference for clients). The order is now AES256,
2927 AES128, 3DES, ARCFOUR128.
2929 51. Small patch to Sieve code - explicitly set From: when generating an
2932 52. Exim crashed if a remote delivery caused a very long error message to be
2933 recorded - for instance if somebody sent an entire SpamAssassin report back
2934 as a large number of 550 error lines. This bug was coincidentally fixed by
2935 increasing the size of one of Exim's internal buffers (big_buffer) that
2936 happened as part of the Exiscan merge. However, to be on the safe side, I
2937 have made the code more robust (and fixed the comments that describe what
2940 53. Now that there can be additional text after "Completed" in log lines (if
2941 the queue_time_overall log selector is set), a one-byte patch to exigrep
2942 was needed to allow it to recognize "Completed" as not the last thing in
2945 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2946 patch that reportedly fixes this has been added. I am not expert enough to
2947 create a test for it. This is what the patch creator wrote:
2949 "I found a little strange behaviour of ldap code when working with
2950 Windows 2003 AD Domain, where users was placed in more than one
2951 Organization Units. When I tried to give exim partial DN, the exit code
2952 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2953 But simultaneously result of request was absolutely normal ldap result,
2954 so I produce this patch..."
2956 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2957 REFERENCE, so I have modified the code to exclude the patch when that macro
2960 55. Some experimental protocols are using DNS PTR records for new purposes. The
2961 keys for these records are domain names, not reversed IP addresses. The
2962 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2963 leaves it alone. Component reversal etc. now happens only for IP addresses.
2966 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2968 57. Double the size of the debug message buffer (to 2048) so that more of very
2969 long debug lines gets shown.
2971 58. The exicyclog utility now does better if the number of log files to keep
2972 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2974 59. Two changes related to the smtp_active_hostname option:
2976 (1) $smtp_active_hostname is now available as a variable.
2977 (2) The default for smtp_banner uses $smtp_active_hostname instead
2978 of $primary_hostname.
2980 60. The host_aton() function is supposed to be passed a string that is known
2981 to be a valid IP address. However, in the case of IPv6 addresses, it was
2982 not checking this. This is a hostage to fortune. Exim now panics and dies
2983 if the condition is not met. A case was found where this could be provoked
2984 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2985 components; fortuitously, this particular loophole had already been fixed
2986 by change 4.50/55 above.
2988 If there are any other similar loopholes, the new check in host_aton()
2989 itself should stop them being exploited. The report I received stated that
2990 data on the command line could provoke the exploit when Exim was running as
2991 exim, but did not say which command line option was involved. All I could
2992 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2993 running as the user.
2996 61. There was a buffer overflow vulnerability in the SPA authentication code
2997 (which came originally from the Samba project). I have added a test to the
2998 spa_base64_to_bits() function which I hope fixes it.
3001 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3002 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3004 63. The daemon start-up calls getloadavg() while still root for those OS that
3005 need the first call to be done as root, but it missed one case: when
3006 deliver_queue_load_max is set with deliver_drop_privilege. This is
3007 necessary for the benefit of the queue runner, because there is no re-exec
3008 when deliver_drop_privilege is set.
3010 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3011 This has been fixed.
3013 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3014 was in use, was not putting the data itself into the right store pool;
3015 consequently, it could be overwritten for a subsequent message in the same
3016 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3019 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3021 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3022 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3023 the function were treating the return as a boolean value, which happened to
3024 work because 0=false and not-0=true, but is not correct code.
3026 68. The host_aton() function was not handling scoped IPv6 addresses (those
3027 with, for example, "%eth0" on the end) correctly.
3029 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3030 negated items (that is, ~something) in unsigned ints. Some compilers
3031 apparently mutter when there is no cast.
3033 70. If an address verification called from an ACL failed, and did not produce a
3034 user-specific message (i.e. there was only a "system" message), nothing was
3035 put in $acl_verify_message. In this situation, it now puts the system
3038 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3039 session; change 4.31/43 added the unwanted input to the log line - except
3040 that it did not do this in the start of session case. It now does.
3042 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3043 This is wrong and can cause the other end to generate a synchronization
3044 error if it is another Exim or anything else that does the synchronization
3045 check. A QUIT command is no longer sent after a timeout.
3047 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3048 during host lookups.
3050 74. The defer_ok option of callout verification was not working if it was used
3051 when verifying addresses in header lines, that is, for this case:
3053 verify = header_sender/callout=defer_ok
3055 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3056 those file descriptors could be used for SMTP connections. If anything
3057 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3058 could be sent to the SMTP client, causing chaos. The daemon now opens
3059 stdin, stdout, and stderr to /dev/null when it puts itself into the
3062 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3063 log by default. The user can override this with suitable Perl magic.
3065 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3066 the log message when discard triggers, was not working for the DATA ACL or
3067 for the non-SMTP ACL.
3069 78. Error message wording change in sieve.c.
3071 79. If smtp_accept_max_per_host was set, the number of connections could be
3072 restricted to fewer than expected, because the daemon was trying to set up
3073 a new connection before checking whether the processes handling previous
3074 connections had finished. The check for completed processes is now done
3075 earlier. On busy systems, this bug wouldn't be noticed because something
3076 else would have woken the daemon, and it would have reaped the completed
3079 80. If a message was submitted locally by a user whose login name contained one
3080 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3081 It caused a spool format error. I have fixed the spool reading code. A
3082 related problem was that the "from" clause in the Received: line became
3083 illegal because of the space(s). It is now covered by ${quote_local_part.
3085 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3088 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3090 83. Patch from Sieve maintainer to fix unterminated string problem in
3091 "vacation" handling.
3093 84. Some minor changes to the Linux configuration files to help with other
3094 OS variants using glibc.
3096 85. One more patch for Sieve to update vacation handling to latest spec.
3099 ----------------------------------------------------
3100 See the note above about the 4.44 and 4.50 releases.
3101 ----------------------------------------------------
3107 1. Change 4.43/35 introduced a bug that caused file counts to be
3108 incorrectly computed when quota_filecount was set in an appendfile
3111 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3112 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3115 3. Give more explanation in the error message when the command for a transport
3116 filter fails to execute.
3118 4. There are several places where Exim runs a non-Exim command in a
3119 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3120 was being done only for the command run by the queryprogram router. It is
3121 now done for all such subprocesses. The other cases are: ${run, transport
3122 filters, and the commands run by the lmtp and pipe transports.
3124 5. Some older OS have a limit of 256 on the maximum number of file
3125 descriptors. Exim was using setrlimit() to set 1000 as a large value
3126 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3127 systems. I've change it so that if it can't get 1000, it tries for 256.
3129 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3130 was an oversight, and furthermore, ever since the addition of extra
3131 controls (e.g. 4.43/32), the checks on when to allow different forms of
3132 "control" were broken. There should now be diagnostics for all cases when a
3133 control that does not make sense is encountered.
3135 7. $recipients is now available in the predata ACL (oversight).
3137 8. Tidy the search cache before the fork to do a delivery from a message
3138 received from the command line. Otherwise the child will trigger a lookup
3139 failure and thereby defer the delivery if it tries to use (for example) a
3140 cached ldap connection that the parent has called unbind on.
3142 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3143 of $address_data from the recipient verification was clobbered by the
3144 sender verification.
3146 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3147 was its contents. (It was OK if the option was not defined at all.)
3149 11. A "Completed" log line is now written for messages that are removed from
3150 the spool by the -Mrm option.
3152 12. $host_address is now set to the target address during the checking of
3153 ignore_target_hosts.
3155 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3156 being passed; this would have caused $sender_host_name to have been used if
3157 matching the list had actually called for a host name (not very likely,
3158 since this list is usually IP addresses). A host name is now passed as
3161 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3162 code that specifies a non-restarting handler (typically sigaction() in
3163 modern systems) in an attempt to fix a rare and obscure crash bug.
3165 15. Narrowed the window for a race in the daemon that could cause it to ignore
3166 SIGCHLD signals. This is not a major problem, because they are used only to
3167 wake it up if nothing else does.
3169 16. A malformed maildirsize file could cause Exim to calculate negative values
3170 for the mailbox size or file count. Odd effects could occur as a result.
3171 The maildirsize information is now recalculated if the size or filecount
3174 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3175 support for a long time. Removed HAVE_SYS_VFS_H.
3177 18. Updated exipick to current release from John Jetmore.
3179 19. Allow an empty sender to be matched against a lookup in an address list.
3180 Previously the only cases considered were a regular expression, or an
3183 20. Exim went into a mad DNS lookup loop when doing a callout where the
3184 host was specified on the transport, if the DNS lookup yielded more than
3187 21. The RFC2047 encoding function was originally intended for short strings
3188 such as real names; it was not keeping to the 75-character limit for
3189 encoded words that the RFC imposes. It now respects the limit, and
3190 generates multiple encoded words if necessary. To be on the safe side, I
3191 have increased the buffer size for the ${rfc2047: expansion operator from
3194 22. Failure to deliver a bounce message always caused it to be frozen, even if
3195 there was an errors_to setting on the router. The errors_to setting is now
3198 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3199 canonical form (fully expanded) before being placed in
3200 $sender_host_address.
3202 24. Updated eximstats to version 1.33
3204 25. Include certificate and key file names in error message when GnuTLS fails
3205 to set them up, because the GnuTLS error message doesn't include the name
3206 of the failing file when there is a problem reading it.
3208 26. Expand error message when OpenSSL has problems setting up cert/key files.
3211 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3212 (this can affect the format of dates).
3214 28. exim_tidydb, when checking for the continued existence of a message for
3215 which it has found a message-specific retry record, was not finding
3216 messages that were in split spool directories. Consequently, it was
3217 deleting retry records that should have stayed in existence.
3219 29. eximstats updated to version 1.35
3220 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3221 1.35 - bugfix such that pie charts by volume are generated correctly
3223 30. The SPA authentication driver was not abandoning authentication and moving
3224 on to the next authenticator when an expansion was forced to fail,
3225 contradicting the general specification for all authenticators. Instead it
3226 was generating a temporary error. It now behaves as specified.
3228 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3229 (the order specifies the preference for clients). The order is now AES256,
3230 AES128, 3DES, ARCFOUR128.
3232 31. Small patch to Sieve code - explicitly set From: when generating an
3235 32. Exim crashed if a remote delivery caused a very long error message to be
3236 recorded - for instance if somebody sent an entire SpamAssassin report back
3237 as a large number of 550 error lines. This bug was coincidentally fixed by
3238 increasing the size of one of Exim's internal buffers (big_buffer) that
3239 happened as part of the Exiscan merge. However, to be on the safe side, I
3240 have made the code more robust (and fixed the comments that describe what
3243 33. Some experimental protocols are using DNS PTR records for new purposes. The
3244 keys for these records are domain names, not reversed IP addresses. The
3245 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3246 leaves it alone. Component reversal etc. now happens only for IP addresses.
3249 34. The host_aton() function is supposed to be passed a string that is known
3250 to be a valid IP address. However, in the case of IPv6 addresses, it was
3251 not checking this. This is a hostage to fortune. Exim now panics and dies
3252 if the condition is not met. A case was found where this could be provoked
3253 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3254 components; fortuitously, this particular loophole had already been fixed
3255 by change 4.50/55 or 4.44/33 above.
3257 If there are any other similar loopholes, the new check in host_aton()
3258 itself should stop them being exploited. The report I received stated that
3259 data on the command line could provoke the exploit when Exim was running as
3260 exim, but did not say which command line option was involved. All I could
3261 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3262 running as the user.
3265 35. There was a buffer overflow vulnerability in the SPA authentication code
3266 (which came originally from the Samba project). I have added a test to the
3267 spa_base64_to_bits() function which I hope fixes it.
3270 36. The daemon start-up calls getloadavg() while still root for those OS that
3271 need the first call to be done as root, but it missed one case: when
3272 deliver_queue_load_max is set with deliver_drop_privilege. This is
3273 necessary for the benefit of the queue runner, because there is no re-exec
3274 when deliver_drop_privilege is set.
3276 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3277 was in use, was not putting the data itself into the right store pool;
3278 consequently, it could be overwritten for a subsequent message in the same
3279 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3282 38. Sometimes the final signoff response after QUIT could fail to get
3283 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3284 < 0 before doing a fflush(). This bug looks as though it goes back to the
3285 introduction of TLS in release 3.20, but "sometimes" must have been rare
3286 because the tests only now provoked it.
3292 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3293 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3294 Now it returns an integer. A number of calls were still expecting a T/F
3295 return. Fortuitously, in all cases, the tests worked in OK situations,
3296 which is the norm. However, things would have gone wrong on any write
3297 failures on the smtp file descriptor. This function is used when sending
3298 messages over SMTP and also when doing verify callouts.
3300 2. When Exim is called to do synchronous delivery of a locally submitted
3301 message (the -odf or -odi options), it no longer closes stderr before doing
3304 3. Implemented the mua_wrapper option.
3306 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3308 5. Implemented the functions header_remove(), header_testname(),
3309 header_add_at_position(), and receive_remove_recipient(), and exported them
3312 6. If an ACL "warn" statement specified the addition of headers, Exim already
3313 inserted X-ACL-Warn: at the start if there was no header name. However, it
3314 was not making this test for the second and subsequent header lines if
3315 there were newlines in the string. This meant that an invalid header could
3316 be inserted if Exim was badly configured.
3318 7. Allow an ACL "warn" statement to add header lines at the start or after all
3319 the Received: headers, as well as at the end.
3321 8. Added the rcpt_4xx retry error code.
3323 9. Added postmaster_mailfrom=xxx to callout verification option.
3325 10. Added mailfrom=xxxx to the callout verification option, for verify=
3328 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3329 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3331 12. Inserted some casts to stop certain compilers warning when using pointer
3332 differences as field lengths or precisions in printf-type calls (mostly
3333 affecting debugging statements).
3335 13. Added optional readline() support for -be (dynamically loaded).
3337 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3338 same clock tick as a message's arrival, so that its received time was the
3339 same as the "first fail" time on the retry record, and that message
3340 remained on the queue past the ultimate address timeout, every queue runner
3341 would try a delivery (because it was past the ultimate address timeout) but
3342 after another failure, the ultimate address timeout, which should have then
3343 bounced the address, did not kick in. This was a "< instead of <=" error;
3344 in most cases the first failure would have been in the next clock tick
3345 after the received time, and all would be well.
3347 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3348 being recognized when the domain list was tested by the match_domain
3349 condition in an expansion string.
3351 16. Added the ${str2b64: operator.
3353 17. Exim was always calling setrlimit() to set a large limit for the number of
3354 processes, without checking whether the existing limit was already
3355 adequate. (It did check for the limit on file descriptors.) Furthermore,
3356 errors from getrlimit() and setrlimit() were being ignored. Now they are
3357 logged to the main and panic logs, but Exim does carry on, to try to do its
3358 job under whatever limits there are.
3360 18. Imported PCRE 5.0.
3362 19. Trivial typo in log message " temporarily refused connection" (the leading
3365 20. If the log selector return_path_on_delivery was set and an address was
3366 redirected to /dev/null, the delivery process crashed because it assumed
3367 that a return path would always be set for a "successful" delivery. In this
3368 case, the whole delivery is bypassed as an optimization, and therefore no
3371 21. Internal re-arrangement: the function for sending a challenge and reading
3372 a response while authentication was assuming a zero-terminated challenge
3373 string. It's now changed to take a pointer and a length, to allow for
3374 binary data in such strings.
3376 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3378 23. Exim was not respecting finduser_retries when seeking the login of the
3379 uid under which it was called; it was always trying 10 times. (The default
3380 setting of finduser_retries is zero.) Also, it was sleeping after the final
3381 failure, which is pointless.
3383 24. Implemented tls_on_connect_ports.
3385 25. Implemented acl_smtp_predata.
3387 26. If the domain in control=submission is set empty, Exim assumes that the
3388 authenticated id is a complete email address when it generates From: or
3389 Sender: header lines.
3391 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3392 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3393 chown and chgrp in /bin and hostname in /usr/bin.
3395 28. Exim was keeping the "process log" file open after each use, just as it
3396 does for the main log. This opens the possibility of it remaining open for
3397 long periods when the USR1 signal hits a daemon. Occasional processlog
3398 errors were reported, that could have been caused by this. Anyway, it seems
3399 much more sensible not to leave this file open at all, so that is what now
3402 29. The long-running daemon process does not normally write to the log once it
3403 has entered its main loop, and it closes the log before doing so. This is
3404 so that log files can straightforwardly be renamed and moved. However,
3405 there are a couple of unusual error situations where the daemon does write
3406 log entries, and I had neglected to close the log afterwards.
3408 30. The text of an SMTP error response that was received during a remote
3409 delivery was being truncated at 512 bytes. This is too short for some of
3410 the long messages that one sometimes sees. I've increased the limit to
3413 31. It is now possible to make retry rules that apply only when a message has a
3414 specific sender, in particular, an empty sender.
3416 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3417 it possible to be selective about when SMTP synchronization is enforced.
3419 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3421 32. Implemented hosts_connection_nolog.
3423 33. Added an ACL for QUIT.
3425 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3428 35. Added mailbox_size and mailbox_filecount to appendfile.
3430 36. Added control = no_multiline_responses to ACLs.
3432 37. There was a bug in the logic of the code that waits for the clock to tick
3433 in the case where the clock went backwards by a substantial amount such
3434 that the microsecond fraction of "now" was more than the microsecond
3435 fraction of "then" (but the whole seconds number was less).
3437 38. Added support for the libradius Radius client library this is found on
3438 FreeBSD (previously only the radiusclient library was supported).
3444 1. When certain lookups returned multiple values in the form name=value, the
3445 quoting of the values was not always being done properly. Specifically:
3446 (a) If the value started with a double quote, but contained no whitespace,
3448 (b) If the value contained whitespace other than a space character (i.e.
3449 tabs or newlines or carriage returns) it was not quoted.
3450 This fix has been applied to the mysql and pgsql lookups by writing a
3451 separate quoting function and calling it from the lookup code. The fix
3452 should probably also be applied to nisplus, ibase and oracle lookups, but
3453 since I cannot test any of those, I have not disturbed their existing code.
3455 2. A hit in the callout cache for a specific address caused a log line with no
3456 reason for rejecting RCPT. Now it says "Previous (cached) callout
3457 verification failure".
3459 3. There was an off-by-one bug in the queryprogram router. An over-long
3460 return line was truncated at 256 instead of 255 characters, thereby
3461 overflowing its buffer with the terminating zero. As well as fixing this, I
3462 have increased the buffer size to 1024 (and made a note to document this).
3464 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3465 when Exim is waiting for an SMTP response from a remote server, Exim
3466 restarts its select() call on the socket, thereby resetting its timeout.
3467 This is not a problem when such interrupts are rare. Somebody set up a cron
3468 job to run exiwhat every 2 minutes, which is less than the normal select()
3469 timeout (5 or 10 minutes). This meant that the select() timeout never
3470 kicked in because it was always reset. I have fixed this by comparing the
3471 time when an interrupt arrives with the time at the start of the first call
3472 to select(). If more time than the timeout has elapsed, the interrupt is
3473 treated as a timeout.
3475 5. Some internal re-factoring in preparation for the addition of Sieve
3476 extensions (by MH). In particular, the "personal" test is moved to a
3477 separate function, and given an option for scanning Cc: and Bcc: (which is
3478 not set for Exim filters).
3480 6. When Exim created an email address using the login of the caller as the
3481 local part (e.g. when creating a From: or Sender: header line), it was not
3482 quoting the local part when it contained special characters such as @.
3484 7. Installed new OpenBSD configuration files.
3486 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3487 try to make them clearer.
3489 9. Callout options, other than the timeout value, were being ignored when
3490 verifying sender addresses in header lines. For example, when using
3492 verify = header_sender/callout=no_cache
3494 the cache was (incorrectly) being used.
3496 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3497 only the Cygwin environment.
3499 11. When return_path_on_delivery was set as a log selector, if different remote
3500 addresses in the same message used different return paths and parallel
3501 remote delivery occurred, the wrong values would sometimes be logged.
3502 (Whenever a remote delivery process finished, the return path value from
3503 the most recently started remote delivery process was logged.)
3505 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3506 lines when AUTH and/or TLS are in use. This is the "received protocol"
3507 field. Exim used to use "asmtp" for authenticated SMTP, without any
3508 indication (in the protocol name) for TLS use. Now it follows the RFC and
3509 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3510 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3511 names appear in log lines as well as in Received: header lines.
3513 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3514 extensions, and comparison tests, and to fix some bugs.
3516 14. Changes to the "personal" filter test:
3518 (1) The test was buggy in that it was just doing the equivalent of
3519 "contains" tests on header lines. For example, if a user's address was
3520 anne@some.where, the "personal" test would incorrectly be true for
3522 To: susanne@some.where
3524 This test is now done by extracting each address from the header in turn,
3525 and checking the entire address. Other tests that are part of "personal"
3526 are now done using regular expressions (for example, to check local parts
3527 of addresses in From: header lines).
3529 (2) The list of non-personal local parts in From: addresses has been
3530 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3531 taken from the Sieve specification recommendations.
3533 (3) If the message contains any header line starting with "List-" it is
3534 treated as non-personal.
3536 (4) The test for "circular" in the Subject: header line has been removed
3537 because it now seems ill-conceived.
3539 15. Minor typos in src/EDITME comments corrected.
3541 16. Installed latest exipick from John Jetmore.
3543 17. If headers_add on a router specified a text string that was too long for
3544 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3545 of string_sprintf() is now avoided.
3547 18. $message_body_size was not set (it was always zero) when running the DATA
3548 ACL and the local_scan() function.
3550 19. For the "mail" command in an Exim filter, no default was being set for
3551 the once_repeat time, causing a random time value to be used if "once" was
3552 specified. (If the value happened to be <= 0, no repeat happened.) The
3553 default is now 0s, meaning "never repeat". The "vacation" command was OK
3554 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3555 (I found it when inspecting the code).
3557 20. There is now an overall timeout for performing a callout verification. It
3558 defaults to 4 times the callout timeout, which applies to individual SMTP
3559 commands during the callout. The overall timeout applies when there is more
3560 than one host that can be tried. The timeout is checked before trying the
3561 next host. This prevents very long delays if there are a large number of
3562 hosts and all are timing out (e.g. when the network connections are timing
3563 out). The value of the overall timeout can be changed by specifying an
3564 additional sub-option for "callout", called "maxwait". For example:
3566 verify = sender/callout=5s,maxwait=20s
3568 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3569 to the end before writing, but this should make it even safer).
3571 22. Exim was forgetting that it had advertised PIPELINING for the second and
3572 subsequent messages on an SMTP connection. It was also not resetting its
3573 memory on STARTTLS and an internal HELO.
3575 23. When Exim logs an SMTP synchronization error within a session, it now
3576 records whether PIPELINING has been advertised or not.
3578 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3579 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3580 rather than long int.
3582 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3584 26. Added the never_mail option to autoreply.
3590 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3591 crash if the getsockname() call failed; this can happen if a connection is
3592 closed very soon after it is established. The problem was simply in the
3593 order in which certain operations were done, causing Exim to try to write
3594 to the SMTP stream before it had set up the file descriptor. The bug has
3595 been fixed by making things happen in the correct order.
3601 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3602 before the connection was closed, thus losing the rejection response.
3604 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3605 some early Solaris releases, but causes trouble in current releases where
3606 socklen_t is defined.
3608 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3611 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3614 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3615 that releasing the top of it at the end releases what was used for sub-
3616 expansions (unless the block got too big). However, discard this block if
3617 the first thing is a variable or header, so that we can use its block when
3618 it is dynamic (useful for very large $message_headers, for example).
3620 6. Lookups now cache *every* query, not just the most recent. A new, separate
3621 store pool is used for this. It can be recovered when all lookup caches are
3622 flushed. Lookups now release memory at the end of their result strings.
3623 This has involved some general refactoring of the lookup sources.
3625 7. Some code has been added to the store_xxx() functions to reduce the amount
3626 of flapping under certain conditions.
3628 8. log_incoming_interface used to affect only the <= reception log lines. Now
3629 it causes the local interface and port to be added to several more SMTP log
3630 lines, for example "SMTP connection from", and rejection lines.
3632 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3634 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3636 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3637 could be overwritten at the end of the current message (or the start of a
3638 new message if it was set in a HELO ACL). The value is now preserved for
3639 the duration of the SMTP connection.
3641 12. If a transport had a headers_rewrite setting, and a matching header line
3642 contained an unqualified address, that address was qualified, even if it
3643 did not match any rewriting rules. The underlying bug was that the values
3644 of the flags that permit the existence of unqualified sender and recipient
3645 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3646 non-local messages, and by -bnq for local messages) were not being
3647 preserved with the message after it was received.
3649 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3650 "next input=" as part of the text comprising the host identity instead of
3651 the correct text. The code was using the same buffer for two different
3652 strings. However, depending on which order the printing function evaluated
3653 its arguments, the bug did not always show up. Under Linux, for example, my
3654 test suite worked just fine.
3656 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3657 doesn't work with some older versions of Perl. It has been changed to "my",
3658 which in any case is probably the better facility to use.
3660 15. A really picky compiler found some instances of statements for creating
3661 error messages that either had too many or two few arguments for the format
3664 16. The size of the buffer for calls to the DNS resolver has been increased
3665 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3666 for addresses that have a lot of PTR records. This alleviates a problem; it
3667 does not fully solve it.
3669 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3670 buffer now truncates the list and logs the incident, which is the same
3671 action as happens when Exim is looking up a host name and its aliases.
3672 Previously in this situation something unpredictable would happen;
3673 sometimes it was "internal error: store_reset failed".
3675 18. If a server dropped the connection unexpectedly when an Exim client was
3676 using GnuTLS and trying to read a response, the client delivery process
3677 crashed while trying to generate an error log message.
3679 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3680 string, for example:
3682 warn message = H1: something\nH2: something
3684 the text was added as a single header line from Exim's point of view
3685 though it ended up OK in the delivered message. However, searching for the
3686 second and subsequent header lines using $h_h2: did not work. This has been
3687 fixed. Similarly, if a system filter added multiple headers in this way,
3688 the routers could not see them.
3690 20. Expanded the error message when iplsearch is called with an invalid key to
3691 suggest using net-iplsearch in a host list.
3693 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3694 ACLs are no longer actually imposed (and a message to that effect is
3697 22. If a "gecos" field in a passwd entry contained escaped characters, in
3698 particular, if it contained a \" sequence, Exim got it wrong when building
3699 a From: or a Sender: header from that name. A second bug also caused
3700 incorrect handling when an unquoted " was present following a character
3701 that needed quoting.
3703 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3704 was not being matched caselessly.
3706 24. Arranged for all hyphens in the exim.8 source to be escaped with
3709 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3710 properly. Recipient callout cache records were still being keyed to include
3711 the sender, even when use_sender was set false. This led to far more
3712 callouts that were necessary. The sender is no longer included in the key
3713 when use_sender is false.
3715 26. Added "control = submission" modifier to ACLs.
3717 27. Added the ${base62d: operator to decode base 62 numbers.
3719 28. dnsdb lookups can now access SRV records.
3721 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3722 the configuration file.
3724 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3725 (-v) mode. This makes the output for a verbose queue run more intelligible.
3727 31. Added a use_postmaster feature to recipient callouts.
3729 32. Added the $body_zerocount variable, containing the number of binary zero
3730 bytes in the message body.
3732 33. The time of last modification of the "new" subdirectory is now used as the
3733 "mailbox time last read" when there is a quota error for a maildir
3736 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3738 35. Added +ignore_unknown as a special item in host lists.
3740 36. Code for decoding IPv6 addresses in host lists is now included, even if
3741 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3742 address was recognized as an IP address, but was then not correctly decoded
3743 into binary, causing unexpected and incorrect effects when compared with
3750 1. Very minor rewording of debugging text in manualroute to say "list of
3751 hosts" instead of "hostlist".
3753 2. If verify=header_syntax was set, and a header line with an unqualified
3754 address (no domain) and a large number of spaces between the end of the
3755 name and the colon was received, the reception process suffered a buffer
3756 overflow, and (when I tested it) crashed. This was caused by some obsolete
3757 code that should have been removed. The fix is to remove it!
3759 3. When running in the test harness, delay a bit after writing a bounce
3760 message to get a bit more predictability in the log output.
3762 4. Added a call to search_tidyup() just before forking a reception process. In
3763 theory, someone could use a lookup in the expansion of smtp_accept_max_
3764 per_host which, without the tidyup, could leave open a database connection.
3766 5. Added the variables $recipient_data and $sender_data which get set from a
3767 lookup success in an ACL "recipients" or "senders" condition, or a router
3768 "senders" option, similar to $domain_data and $local_part_data.
3770 6. Moved the writing of debug_print from before to after the "senders" test
3773 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3774 problems for message scanning, either using a data ACL, or using
3775 local_scan() because the Received: header was not generated till after they
3776 were called (in order to set the time as the time of reception completion).
3777 I have revised the way this works. The header is now generated after the
3778 body is received, but before the ACL or local_scan() are called. After they
3779 are run, the timestamp in the header is updated.
3785 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3786 before starting a queue runner without re-exec. This happened only when
3787 deliver_drop_privilege was set or when the Exim user was set to root. The
3788 effect of the bug was that timeouts during subsequent deliveries caused
3789 crashes instead of being properly handled. The handler is now left at its
3790 default (and expected) setting.
3792 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3793 message, again when deliver_drop_privilege is set or Exim is run as root.
3794 The bug described in (1) was not present in this case, but the tidying up
3795 of the other signals was missing. I have made the two cases consistent.
3797 3. The ignore_target_hosts setting on a manualroute router was being ignored
3798 for hosts that were looked up using the /MX notation.
3800 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3803 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3804 operated on the sender address. After changing the $sender_address to <>
3805 for the sender address verify, Exim was re-instated it as the original
3806 (before rewriting) address, but remembering that it had rewritten it, so it
3807 wasn't rewriting it again. This bug also had the effect of breaking the
3808 sender address verification caching when the sender address was rewritten.
3810 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3811 This has been changed so that if the ip literal address matches
3812 ignore_target_hosts, the router declines.
3814 7. Added expansion conditions match_domain, match_address, and match_local_
3815 part (NOT match_host).
3817 8. The placeholder for the Received: header didn't have a length field set.
3819 9. Added code to Exim itself and to exim_lock to test for a specific race
3820 condition that could lead to file corruption when using MBX delivery. The
3821 issue is with the lockfile that is created in /tmp. If this file is removed
3822 after a process has opened it but before that process has acquired a lock,
3823 there is the potential for a second process to recreate the file and also
3824 acquire a lock. This could lead to two Exim processes writing to the file
3825 at the same time. The added code performs the same test as UW imapd; it
3826 checks after acquiring the lock that its file descriptor still refers to
3827 the same named file.
3829 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3830 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3833 11. Added the smtp_active_hostname option. If used, this will typically be made
3834 to depend on the incoming interface address. Because $interface_address is
3835 not set up until the daemon has forked a reception process, error responses
3836 that can happen earlier (such as "too many connections") no longer contain
3839 12. If an expansion in a condition on a "warn" statement fails because a lookup
3840 defers, the "warn" statement is abandoned, and the next ACL statement is
3841 processed. Previously this caused the whole ACL to be aborted.
3843 13. Added the iplsearch lookup type.
3845 14. Added ident_timeout as a log selector.
3847 15. Added tls_certificate_verified as a log selector.
3849 16. Added a global option tls_require_ciphers (compare the smtp transport
3850 option of the same name). This controls incoming TLS connections.
3852 17. I finally figured out how to make tls_require_ciphers do a similar thing
3853 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3854 before starting the TLS session.
3856 18. Tabs are now shown as \t in -bP output.
3858 19. If the log selector return_path_on_delivery was set, Exim crashed when
3859 bouncing a message because it had too many Received: header lines.
3861 20. If two routers both had headers_remove settings, and the first one included
3862 a superfluous trailing colon, the final name in the first list and the
3863 first name in the second list were incorrectly joined into one item (with a
3864 colon in the middle).
3870 1. Added -C and -D options to the exinext utility, mainly to make it easier
3871 to include in the automated testing, but these could be helpful when
3872 multiple configurations are in use.
3874 2. The exinext utility was not formatting the output nicely when there was
3875 an alternate port involved in the retry record key, nor when there was a
3876 message id as well (for retries that were specific to a specific message
3877 and a specific host). It was also confused by IPv6 addresses, because of
3878 the additional colons they contain. I have fixed the IPv4 problem, and
3879 patched it up to do a reasonable job for IPv6.
3881 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3882 delivery, the log line now contains "pipelined" if PIPELINING was used.
3884 4. An SMTP transport process used to panic and die if the bind() call to set
3885 an explicit outgoing interface failed. This has been changed; it is now
3886 treated in the same way as a connect() failure.
3888 5. A reference to $sender_host_name in the part of a conditional expansion
3889 that was being skipped was still causing a DNS lookup. This no longer
3892 6. The def: expansion condition was not recognizing references to header lines
3893 that used bh_ and bheader_.
3895 7. Added the _cache feature to named lists.
3897 8. The code for checking quota_filecount in the appendfile transport was
3898 allowing one more file than it should have been.
3900 9. For compatibility with Sendmail, the command line option
3908 and sets the incoming protocol and host name (for trusted callers). The
3909 host name and its colon can be omitted when only the protocol is to be set.
3910 Note the Exim already has two private options, -pd and -ps, that refer to
3911 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3912 "s", but I don't think that's a major issue.
3914 10. A number of refactoring changes to the code, none of which should affect
3917 (a) The number of logging options was getting close to filling up the
3918 32-bit word that was used as a bit map. I have split them into two classes:
3919 those that are passed in the argument to log_write(), and those that are
3920 only ever tested independently outside of that function. These are now in
3921 separate 32-bit words, so there is plenty of room for expansion again.
3922 There is no change in the user interface or the logging behaviour.
3924 (b) When building, for example, log lines, the code previously used a
3925 macro that called string_cat() twice, in order to add two strings. This is
3926 not really sufficiently general. Furthermore, there was one instance where
3927 it was actually wrong because one of the argument was used twice, and in
3928 one call a function was used. (As it happened, calling the function twice
3929 did not affect the overall behaviour.) The macro has been replaced by a
3930 function that can join an arbitrary number of extra strings onto a growing
3933 (c) The code for expansion conditions now uses a table and a binary chop
3934 instead of a serial search (which was left over from when there were very
3935 few conditions). Also, it now recognizes conditions like "pam" even when
3936 the relevant support is not compiled in: a suitably worded error message is
3937 given if an attempt is made to use such a condition.
3939 11. Added ${time_interval:xxxxx}.
3941 12. A bug was causing one of the ddress fields not to be passed back correctly
3942 from remote delivery subprocesses. The field in question was not being
3943 subsequently used, so this caused to problems in practice.
3945 13. Added new log selectors queue_time and deliver_time.
3947 14. Might have fixed a bug in maildirsizefile handling that threw up
3948 "unexpected character" debug warnings, and recalculated the data
3949 unnecessarily. In any case, I expanded the warning message to give more
3952 15. Added the message "Restricted characters in address" to the statements in
3953 the default ACL that block characters like @ and % in local parts.
3955 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3956 Three changes have been made:
3958 (a) There was a serious bug; a negative response to MAIL caused the whole
3959 recipient domain to be cached as invalid, thereby blocking all messages
3960 to all local parts at the same domain, from all senders. This bug has
3961 been fixed. The domain is no longer cached after a negative response to
3962 MAIL if the sender used is not empty.
3964 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3967 (c) A new callout option, "use_sender" has been added for people who want
3968 the modified behaviour.
3974 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3977 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3978 indeed breaks things for older releases.
3980 3. Added additional logging to the case where there is a problem reading data
3981 from a filter that is running in a subprocess using a pipe, in order to
3982 try to track down a specific problem.
3984 4. Testing facility fudge: when running in the test harness and attempting
3985 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3986 getting "No route to host". Convert this to a timeout.
3988 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3991 6. Some OS don't have socklen_t but use size_t instead. This affects the
3992 fifth argument of getsockopt() amongst other things. This is now
3993 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3994 can be set for individual OS. I have set it for SunOS5, OSF1, and
3995 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3996 some earlier ones do not.
3998 7. Change 4.30/15 was not doing the test caselessly.
4000 8. The standard form for an IPv6 address literal was being rejected by address
4001 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4002 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4003 this, as well as the form without the "IPv6" on the front (but only when
4004 address literals are enabled, of course).
4006 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4008 10. Exim crashed if a message with an empty sender address specified by -f
4009 encountered a router with an errors_to setting. This could be provoked only
4010 by a command such as
4014 where an empty string was supplied; "<>" did not hit this bug.
4016 11. Installed PCRE release 4.5.
4018 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4019 remained set. It is now erased.
4021 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4022 times from message ids (which are base 36 rather than the normal 62).
4024 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4025 were being counted as actual protocol errors, and logged if the log
4026 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4027 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4028 and DATA following a set of rejected RCPTs do not count as protocol errors.
4029 In other words, Exim assumes they were pipelined, though this may not
4030 actually be the case. Of course, in all cases the client gets an
4031 appropriate error code.
4033 15. If a lookup fails in an ACL condition, a message about the failure may
4034 be available; it is used if testing the ACL cannot continue, because most
4035 such messages specify what the cause of the deferral is. However, some
4036 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4037 that caused an old message to be retained and used if a later statement
4038 caused a defer, replacing the real cause of the deferral.
4040 16. If an IP address had so many PTR records that the DNS lookup buffer
4041 was not large enough to hold them, Exim could crash while trying to process
4042 the truncated data. It now detects and logs this case.
4044 17. Further to 4.21/58, another change has been made: if (and only if) the
4045 first line of a message (the first header line) ends with CRLF, a bare LF
4046 in a subsequent header line has a space inserted after it, so as not to
4047 terminate the header.
4049 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4050 unnecessarily, used atoi() instead of strtol(), and didn't check the
4051 termination when getting file sizes from file names by regex.
4053 19. Completely re-implemented the support for maildirsize files, in the light
4054 of a number of problems with the previous contributed implementation
4055 (4.30/29). In particular:
4057 . If the quota is zero, the maildirsize file is maintained, but no quota is
4060 . If the maildir directory does not exist, it is created before any attempt
4061 to write a maildirsize file.
4063 . The quota value in the file is just a cache; if the quota is changed in
4064 the transport, the new value overrides.
4066 . A regular expression is available for excluding directories from the
4069 20. The autoreply transport checks the characters in options that define the
4070 message's headers; it allows continued headers, but it was checking with
4071 isspace() after an embedded newline instead of explicitly looking for a
4074 21. If all the "regular" hosts to which an address was routed had passed their
4075 expiry times, and had not reached their retry times, the address was
4076 bounced, even if fallback hosts were defined. Now Exim should go on to try
4079 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4080 equivalent code in the SMTP transport. Some hosts send humungous responses
4081 to HELO/EHLO, more than 1024 it seems.
4083 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4084 gives compiler warnings in some environments. I've now done it "properly",
4087 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4088 (because of problems with the built-in one) was declared to return uschar *
4089 instead of char *, causing compiler failure.
4091 25. Fixed a file descriptor leak when processing alias/forward files.
4093 26. Fixed a minor format string issue in dbfn.c.
4095 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4097 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4098 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4101 29. When a sender address is verified, it is cached, to save repeating the test
4102 when there is more than one recipient in a message. However, when the
4103 verification involves a callout, it is possible for different callout
4104 options to be set for different recipients. It is too complicated to keep
4105 track of this in the cache, so now Exim always runs a verification when a
4106 callout is required, relying on the callout cache for the optimization.
4107 The overhead is duplication of the address routing, but this should not be
4110 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4111 to be verified with callout=postmaster, and the main callout worked but the
4112 postmaster check failed, the verification correctly failed. However, if a
4113 subsequent RCPT command asked for sender verification *without* the
4114 postmaster check, incorrect caching caused this verification also to fail,
4117 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4118 it was not caching the DNS options (qualify_single, search_parents) that
4119 were used when the lookup failed. A subsequent lookup with different
4120 options therefore always gave the same answer, though there were cases
4121 where it should not have. (Example: a "domains = !$mx_any" option on a
4122 dnslookup router: the "domains" option is always processed without any
4123 widening, but the router might have qualify_single set.) Now Exim uses the
4124 cached value only when the same options are set.
4126 32. Added John Jetmore's "exipick" utility to the distribution.
4128 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4129 than a timeout (e.g. a certificate is required, and is not provided), an
4130 Exim server now closes the connection immediately. Previously it waited for
4131 the client to close - but if the client is SSL, it seems that they each
4132 wait for each other, leading to a delay before one of them times out.
4134 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4135 maintained 0.8.x compatibility because I don't think many are using it, and
4136 it is clearly obsolete.
4138 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4141 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4142 client certificate was expired. A simple patch fixes this, though I don't
4143 understand the full logic of why the verify callback is called multiple
4146 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4147 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4148 which causes problems with some clients (such as the Certicom SSL Plus
4149 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4150 disables the coutermeasure allowing Eudora to connect."
4152 38. Exim was not checking that a write() to a log file succeeded. This could
4153 lead to Bad Things if a log got too big, in particular if it hit a file
4154 size limit. Exim now panics and dies if it cannot write to a log file, just
4155 as it does if it cannot open a log file.
4157 39. Modified OS/Makefile-Linux so that it now contains
4159 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4161 The two -D definitions ensure that Exim is compiled with large file
4162 support, which makes it possible to handle log files that are bigger than
4165 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4166 instance) a domain was checked against a named list that involved a lookup,
4167 causing $domain_data to be set, then another domain was checked against the
4168 same list, then the first domain was re-checked, the value of $domain_data
4169 after the final check could be wrong. In particular, if the second check
4170 failed, it could be set empty. This bug probably also applied to
4173 41. The strip_trailing_dot option was not being applied to the address given
4174 with the -f command-line option.
4176 42. The code for reading a message's header from the spool was incrementing
4177 $received_count, but never initializing it. This meant that the value was
4178 incorrect (doubled) while delivering a message in the same process in which
4179 it was received. In the most common configuration of Exim, this never
4180 happens - a fresh exec is done - but it can happen when
4181 deliver_drop_privilege is set.
4183 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4184 it now includes up to 150 characters of the unexpected data in the log
4187 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4188 and building data strings. The size of both of these buffers was 10 000
4189 bytes - far larger than anybody would *ever* want, thought I. Needless to
4190 say, somebody hit the limit. I have increased the maximum line length to
4191 20 000 and the maximum data length of concatenated lines to 100 000. I have
4192 also fixed two bugs, because there was no checking on these buffers. Tsk,
4193 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4194 buffer is too small.
4196 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4197 lsearch lookups. Now it does.
4199 46. When parsing a route_list item in a manualroute router, a fixed-length
4200 buffer was used for the list of hosts. I made this 1024 bytes long,
4201 thinking that nobody would ever have a list of hosts that long. Wrong.
4202 Somebody had a whole pile of complicated expansion conditions, and the
4203 string was silently truncated, leading to an expansion error. It turns out
4204 that it is easier to change to an unlimited length (owing to other changes
4205 that have happened since this code was originally written) than to build
4206 structure for giving a limitation error. The length of the item that
4207 expands into the list of hosts is now unlimited.
4209 47. The lsearch lookup could not handle data where the length of text line was
4210 more than 4095 characters. Such lines were truncated, leading to shortened
4211 data being returned. It should now handle lines of any length.
4213 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4214 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4217 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4218 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4219 no longer happens in comment lines. A list of replacements is now placed
4220 at the head of all of the source files, except those whose only change is
4221 to replace PERL_COMMAND in the very first #! line.
4223 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4224 messages on the queue, with a bottom-up merge sort, using code contributed
4225 by Michael Haardt. This should make operations like -bp somewhat faster on
4226 large queues. It won't affect queue runners, except when queue_run_in_order
4229 51. Installed eximstats 1.31 in the distribution.
4231 52. Added support for SRV lookups to the dnslookup router.
4233 53. If an ACL referred to $message_body or $message_body_end, the value was not
4234 reset for any messages that followed in the same SMTP session.
4236 54. The store-handling optimization for building very long strings was not
4237 differentiating between the different store pools. I don't think this
4238 actually made any difference in practice, but I've tidied it.
4240 55. While running the routers to verify a sender address, $sender_address
4241 was still set to the sender address. This is wrong, because when routing to
4242 send a bounce to the sender, it would be empty. Therefore, I have changed
4243 it so that, while verifying a sender address, $sender_address is set to <>.
4244 (There is no change to what happens when verifying a recipient address.)
4246 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4247 target A or AAAA records (if not already returned) without resetting the
4248 qualify_single or search_parents options of the DNS resolver. These are
4249 inappropriate in this case because the targets of MX and SRV records must
4250 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4251 target that, when qualified, matched something in the local domain. These
4252 two options are now turned off when doing these lookups.
4254 57. It seems that at least some releases of Reiserfs (which does not have the
4255 concept of a fixed number of inodes) returns zero and not -1 for the
4256 number of available inodes. This interacted badly with check_spool_inodes,
4257 which assumed that -1 was the "no such thing" setting. What I have done is
4258 to check that the total number of inodes is greater than zero before doing
4259 the test of how many are available.
4261 58. When a "warn" ACL statement has a log_message modifier, the message is
4262 remembered, and not repeated. This is to avoid a lot of repetition when a
4263 message has many recipients that cause the same warning to be written.
4264 Howewer, Exim was preserving the list of already written lines for an
4265 entire SMTP session, which doesn't seem right. The memory is now reset if a
4266 new message is started.
4268 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4269 debugging output unless log_rewrite was also set.
4271 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4272 of (int)(handle) when we know that handle contains (void *)(-1).
4274 61. The Exim daemon panic-logs an error return when it closes the incoming
4275 connection. However "connection reset by peer" seems to be common, and
4276 isn't really an error worthy of noting specially, so that particular error
4279 62. When Exim is trying to find all the local interfaces, it used to panic and
4280 die if the ioctl to get the interface flags failed. However, it seems that
4281 on at least one OS (Solaris 9) it is possible to have an interface that is
4282 included in the list of interfaces, but for which you get a failure error
4283 for this call. This happens when the interface is not "plumbed" into a
4284 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4285 failure of the "get flags" call assumes that the interface is down.
4287 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4288 makes life easier for people who are doing arithmetic on fields extracted
4289 from dates, where you often get leading zeros that should not be
4290 interpreted as octal.
4292 64. Added qualify_domain to the redirect router, to override the global
4295 65. If a pathologically long header line contained very many addresses (the
4296 report of this problem mentioned 10 000) and each of them was rewritten,
4297 Exim could use up a very large amount of memory. (It kept on making new
4298 copies of the header line as it rewrote, and never released the old ones.)
4299 At the expense of a bit more processing, the header rewriting function has
4300 been changed so that it no longer eats memory in this way.
4302 66. The generation of the Received: header has been moved from the time that a
4303 message starts to be received, to the time that it finishes. The timestamp
4304 in the Received: header should now be very close to that of the <= log
4305 line. There are two side-effects of this change:
4307 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4308 logged header lines no longer include the local Received: line, because
4309 it has not yet been created. The same applies to a copy of the message
4310 that is returned to a non-SMTP sender when a message is rejected.
4312 (b) When a filter file is tested using -bf, no additional Received: header
4313 is added to the test message. After some thought, I decided that this
4316 This change does not affect the value of $received_for. It is still set
4317 after address rewriting, but before local_scan() is called.
4319 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4321 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4322 gave an unhelpful panic error message, and a defer error. I have managed to
4323 change this behaviour so that it now rejects any supplied certificate,
4324 which seems right, as the list of acceptable certificates is empty.
4326 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4327 gave an unhelpful defer error. I have not managed to make this reject any
4328 supplied certificates, but the error message it gives is "no certificate
4329 supplied", which is not helpful.
4331 70. exigrep's output now also includes lines that are not associated with any
4332 message, but which match the given pattern. Implemented by a patch from
4333 Martin Sluka, which also tidied up the Perl a bit.
4335 71. Recipient callout verification, like sender verification, was using <> in
4336 the MAIL FROM command. This isn't really the right thing, since the actual
4337 sender may affect whether the remote host accepts the recipient or not. I
4338 have changed it to use the actual sender in the callout; this means that
4339 the cache record is now keyed on a recipient/sender pair, not just the
4340 recipient address. There doesn't seem to be a real danger of callout loops,
4341 since a callout by the remote host to check the sender would use <>.
4342 [SEE ABOVE: changed after hitting problems.]
4344 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4345 temporary errors. However, in the case of such a code being given after
4346 the end of a data transmission (i.e. after ".") Exim was failing to write
4347 a retry record for the message. (Yes, there was some broken host that was
4348 actually sending 8xx at this point.)
4350 73. An unknown lookup type in a host list could cause Exim to panic-die when
4351 the list was checked. (An example that provoked this was putting <; in the
4352 middle of a list instead of at the start.) If this happened during a DATA
4353 ACL check, a -D file could be left lying around. This kind of configuration
4354 error no longer causes Exim to die; instead it causes a defer errror. The
4355 incident is still logged to the main and panic logs.
4357 74. Buglet left over from Exim 3 conversion. The message "too many messages
4358 in one connection" was written to the rejectlog but not the mainlog, except
4359 when address rewriting (yes!) was being logged.
4361 75. Added write_rejectlog option.
4363 76. When a system filter was run not as root (that is, when system_filter_user
4364 was set), the values of the $n variables were not being returned to the
4365 main process; thus, they were not subsequently available in the $sn
4368 77. Added +return_path_on_delivery log selector.
4370 78. A connection timeout was being treated differently from recipients deferred
4371 when testing hosts_max_try with a message that was older than the host's
4372 retry timeout. (The host should not be counted, thus allowing all hosts to
4373 be tried at least once before bouncing.) This may have been the cause of an
4374 occasionally reported bug whereby a message would remain on the queue
4375 longer than the retry timeout, but would be bounced if a delivery was
4376 forced. I say "may" because I never totally pinned down the problem;
4377 setting up timeout/retry tests is difficult. See also the next item.
4379 79. The ultimate address timeout was not being applied to errors that involved
4380 a combination of host plus message (for example, a timeout on a MAIL
4381 command). When an address resolved to a number of possible hosts, and they
4382 were not all tried for each delivery (e.g. because of hosts_max_try), a
4383 message could remain on the queue longer than the retry timeout.
4385 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4388 81. Fixed an obscure SMTP outgoing bug which required at least the following
4389 conditions: (a) there was another message waiting for the same server;
4390 (b) the server returned 5xx to all RCPT commands in the first message so
4391 that the message was not completed; (c) the server dropped the connection
4392 or gave a negative response to the RSET that Exim sends to abort the
4393 transaction. The observed case was a dropped connection after DATA that had
4394 been sent in pipelining mode. That is, the server had advertised PIPELINING
4395 but was not implementing it correctly. The effect of the bug was incorrect
4396 behaviour, such as trying another host, and this could lead to a crash.
4402 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4403 and daemon.c were passed as pointers to ints; they should have been
4404 pointers to socklen_t variables (which are typically unsigned ints).
4406 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4409 3. Fixed a really odd bug that affected only the testing scheme; patching a
4410 certain fixed string in the binary changed the value of another string that
4411 happened to be identical to the end of the original first string.
4413 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4414 name", it returns that address as the IP address. On some operating
4415 systems (e.g. Solaris), it also passes back the IP address string as the
4416 "host name". However, on others (e.g. Linux), it passes back an empty
4417 string. Exim wasn't checking for this, and was changing the host name to an
4418 empty string, assuming it had been canonicized.
4420 5. Although rare, it is permitted to have more than one PTR record for a given
4421 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4422 all the names associated with an address, because they do in Solaris.
4423 However, it seems that they do not in Linux for data that comes from the
4424 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4425 I found this out when I moved to a new Linux workstation and tried to run
4426 the Exim test suite.
4428 To get round this problem I have changed the code so that it now does its
4429 own call to the DNS to look up PTR records when searching for a host name.
4430 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4431 addresses that are only in /etc/hosts are still found.
4433 This behaviour is, however, controlled by an option called host_lookup_
4434 order, which defaults to "bydns:byaddr". If people want to use the other
4435 order, or indeed, just use one or the other means of lookup, they can
4436 specify it in this variable.
4438 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4439 some operating systems, this comes back from gethostbyaddr() as an empty
4440 string, and this is what Exim used to test for. However, it seems that in
4441 other systems, "." is yielded. Exim now tests for this case too.
4443 7. The values of check_spool_space and check_log_space are now held internally
4444 as a number of kilobytes instead of an absolute number of bytes. If a
4445 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4446 kilobyte. This means that much larger values can be stored.
4448 8. Exim monitor: an attempt to get the action menu when not actually pointing
4449 at a message produces an empty menu entitled "No message selected". This
4450 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4451 no entries in it ("Shell widget menu has zero width and/or height"). So I
4452 have added a single, blank menu entry in this case.
4454 9. Added ${quote_local_part.
4456 10. MIME decoding is now applied to the contents of Subject: header lines when
4459 11. Now that a reference to $sender_host_address automatically causes a reverse
4460 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4461 host lookup before query-style lookups in lists that might use this
4462 variable. This has therefore been abolished, and the "net-" prefix is no
4463 longer necessary for query-style lookups.
4465 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4466 have been a typo for LFLAGS, so it has been changed.
4468 13. The install script calls Exim with "-C /dev/null" in order to find the
4469 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4470 to be output. Howeve, since Exim outputs its version number before the
4471 error, it didn't break the script. It just looked ugly. I fixed this by
4472 always allowing "-C /dev/null" if the caller is root.
4474 14. Ignore overlarge ACL variable number when reading spool file - insurance
4475 against a later release with more variables having written the file.
4477 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4478 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4479 this, as well as the form without the "IPv6" on the front.
4481 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4482 OS/Makefile-Darwin file.
4484 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4485 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4487 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4488 variable. However, some OS use ints for inodes. I've added cast to long int
4489 to get rid of the compiler warning.
4491 19. I had forgotten to lock out "/../" in configuration file names when
4492 ALT_CONFIG_PREFIX was set.
4494 20. Routers used for verification do not need to specify transports. However,
4495 if such a router generated a host list, and callout was configured, Exim
4496 crashed, because it could not find a port number from the (non-existent)
4497 transport. It now assumes port 25 in this circumstance.
4499 21. Added the -t option to exigrep.
4501 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4502 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4503 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4504 them is set, LOOKUP_LSEARCH is forced.
4506 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4508 24. Added sender and host information to the "rejected by local_scan()" log
4509 line; previously there was no indication of these.
4511 25. Added .include_if_exists.
4513 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4514 Linux. It turns out that not all file systems support this. Apparently some
4515 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4516 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4517 not supported on the file descriptor, is now ignored when Exim is trying to
4518 sync a directory. This applies only to Linux.
4520 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4522 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4523 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4526 29. Added support for maildirsize files from supplied patch (modified a bit).
4528 30. The use of :fail: followed by an empty string could lead Exim to respond to
4529 sender verification failures with (e.g.):
4531 550 Verification failed for <xxx>
4532 550 Sender verify failed
4534 where the first response line was missing the '-' that indicates it is not
4535 the final line of the response.
4537 31. The loop for finding the name of the user that called Exim had a hardwired
4538 limit of 10; it now uses the value of finduser_retries, which is used for
4539 all other user lookups.
4541 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4544 33. Exim was neglecting to zero errno before one call of strtol() when
4545 expanding a string and expecting an integer value. On some systems this
4546 resulted in spurious "integer overflow" errors. Also, it was casting the
4547 result into an int without checking.
4549 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4550 did not work. The code looks as if it has *never* worked, though it appears
4551 to have been documented since at least releast 1.62. I have made it work.
4553 35. The "timeout_DNS" error in retry rules, also documented since at least
4554 1.62, also never worked. As it isn't clear exactly what this means, and
4555 clearly it isn't a major issue, I have abolished the feature by treating it
4556 as "timeout", and writing a warning to the main and panic logs.
4558 36. The display of retry rules for -brt wasn't always showing the error code
4561 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4562 timeout_connect_A, timeout_connect_MX.
4564 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4565 to the empty sender.
4567 39. The daemon was not analysing the content of -oX till after it had closed
4568 stderr and disconnected from the controlling terminal. This meant that any
4569 syntax errors were only noted on the panic log, and the return code from
4570 the command was 0. By re-arranging the code a little, I've made the
4571 decoding happen first, so such errors now appear on stderr, and the return
4572 code is 1. However, the actual setting up of the sockets still happens in
4573 the disconnected process, so errors there are still only recorded on the
4576 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4577 connections (as happens on some IP stacks) was logged at start up time as
4578 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4579 it from "IPv6 and IPv4", which means that two separate sockets are being
4582 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4583 says whether AF_INET or AF_INET6 was passed as an argument.
4585 42. Exiwhat output was messed up when time zones were included in log
4588 43. Exiwhat now gives more information about the daemon's listening ports,
4589 and whether -tls-on-connect was used.
4591 44. The "port" option of the smtp transport is now expanded.
4593 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4594 silently ignored. Now an error message is written to the main and panic
4597 46. There's a new ACL modifier called "logwrite" which writes to a log file
4598 as soon as it is encountered.
4600 47. Added $local_user_uid and $local_user_gid at routing time.
4602 48. Exim crashed when trying to verify a sender address that was being
4605 49. Exim was recognizing only a space character after ".include". It now also
4606 recognizes a tab character.
4608 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4609 extracting the relevant information from the specification. The man page no
4610 longer contains scrambled data for the -d option, and I've added a section
4611 at the front about calling Exim under different names.
4613 51. Added "extra_headers" argument to the "mail" command in filter files.
4615 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4618 53. Installed eximstats 1.29.
4620 54. Added transport_filter_timeout as a generic transport option.
4622 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4623 Cc: header lines. This was required by RFC 822, but it not required by RFC
4626 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4627 incoming messages that did not have them. Now it does so only if the
4628 message originates locally, that is, if there is no associated remote host
4629 address. When Resent- header lines are present, this applies to the Resent-
4630 lines rather than the non-Resent- lines.
4632 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4633 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4635 58. Messages for configuration errors now include the name of the main
4636 configuration file - useful now that there may be more than one file in a
4637 list (.included file names were always shown).
4639 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4640 for those rare installations that do not start the daemon as root or run it
4641 setuid root. I've cut out the call to initgroups() if the daemon is not
4644 60. The Exim user and group can now be bound into the binary as text strings
4645 that are looked up at the start of Exim's processing.
4647 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4649 62. Added $mailstore_basename variable.
4651 63. Installed patch to sieve.c from Michael Haardt.
4653 64. When Exim failed to open the panic log after failing to open the main log,
4654 the original message it was trying to log was written to stderr and debug
4655 output, but if they were not available (the usual case in production), it
4656 was lost. Now it is written to syslog before the two lines that record the
4657 failures to open the logs.
4659 65. Users' Exim filters run in subprocesses under the user's uid. It is
4660 possible for a "deliver" command or an alias in a "personal" command to
4661 provoke an address rewrite. If logging of address rewriting is configured,
4662 this fails because the process is not running as root or exim. There may be
4663 a better way of dealing with this, but for the moment (because 4.30 needs
4664 to be released), I have disabled address rewrite logging when running a
4665 filter in a non-root, non-exim process.
4671 1. The buildconfig auxiliary program wasn't quoting the value set for
4672 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4673 not defined. This bug was masked in 4.22 by the effect that was fixed in
4676 2. Some messages that were rejected after a message id was allocated were
4677 shown as "incomplete" by exigrep. It no longer does this for messages that
4678 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4680 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4681 have allow_domain_literals set, the ID did not get logged in the <= line.
4682 Domain literals are now always recognized in Message-ID: header lines.
4684 4. The first argument for a ${extract expansion item is the key name or field
4685 number. Leading and trailing spaces in this item were not being ignored,
4686 causing some misleading effects.
4688 5. When deliver_drop_privilege was set, single queue runner processes started
4689 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4690 same command in the process it spins off) were not dropping privilege.
4692 6. When the daemon running as "exim" started a queue runner, it always
4693 re-executed Exim in the spun-off process. This is a waste of effort when
4694 deliver_drop_privilege is set. The new process now just calls the
4695 queue-runner function directly.
4701 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4704 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4705 ignored. Though the use of -oP was forcing the writing of a pid file, it
4706 was always written to the default place.
4708 3. If the message "no IP address found for host xxxx" is generated during
4709 incoming verification, it is now followed by identification of the incoming
4710 connection (so you can more easily find what provoked it).
4712 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4714 5. Added some features to "harden" Exim a bit more against certain attacks:
4716 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4717 be put in Local/Makefile. This is like the never_users runtime option,
4718 but it cannot be overridden. The default setting is "root".
4720 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4721 prefix string with which any file named in a -C command line option
4724 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4725 is retained for -C and -D only if the caller of Exim is root. Without
4726 it, the exim user may also use -C and -D and retain privilege.
4728 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4729 command line option is disabled.
4731 6. Macro names set by the -D option must start with an upper case letter, just
4732 like macro names defined in the configuration file.
4734 7. Added "dereference=" facility to LDAP.
4736 8. Two instances of the typo "uknown" in the source files are fixed.
4738 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4739 the Configure-Makefile script screwed up while processing it.
4741 10. Incorporated PCRE 4.4.
4743 11. The SMTP synchronization check was not operating right at the start of an
4744 SMTP session. For example, it could not catch a HELO sent before the client
4745 waited for the greeting. There is now a check for outstanding input at the
4746 point when the greeting is written. Because of the duplex, asynchronous
4747 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4748 way, but not yet received, when the check is performed.
4750 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4751 on TCP/IP sockets, because this apparently causes some broken clients to
4754 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4755 unchanged) from the Cygwin maintainer.
4757 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4758 was supported instead of testing for mbx. Effectively a typo.
4760 15. The spa authenticator server code was not checking that the input it
4761 received was valid base64.
4763 16. The debug output line for the "set" modifier in ACLs was not showing the
4764 name of the variable that was being set.
4766 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4768 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4769 lookup on its own. Something else was needed to trigger the lookup. For
4770 example, a match in host_lookup or the need for a host name in a host list.
4771 Now, if $sender_host_name is referenced and the host name has not yet been
4772 looked up, a lookup is performed. If the lookup fails, the variable remains
4773 empty, and $host_lookup_failed is set to "1".
4775 19. Added "eqi" as a case-independent comparison operator.
4777 20. The saslauthd authentication condition could segfault if neither service
4778 nor realm was specified.
4780 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4781 error message that was logged was misleading, and incoming SMTP
4782 connections were dropped. The message is now more accurate, and temporary
4783 errors are given to SMTP connections.
4785 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4786 (except RSET) with a 421 error, until QUIT is received. However, it was
4787 failing to send a response to QUIT.
4789 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4790 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4791 ACL is now run after the helo_try_verify_hosts code.
4793 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4794 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4795 case-independent, so other case variants are also recognized). Apparently
4796 some systems use these upper case variants.
4798 25. If more than two messages were waiting for the same host, and a transport
4799 filter was specified for the transport, Exim sent two messages over the
4800 same TCP/IP connection, and then failed with "socket operation on non-
4801 socket" when it tried to send the third.
4803 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4805 27. The extern definition of crypt16() in expand.c was not being excluded when
4806 the OS had its own crypt16() function.
4808 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4809 as a preferred synonym for return_size_limit, both as an option and as an
4812 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4814 30. Changed the default configuration ACL to relax the local part checking rule
4815 for addresses that are not in any local domains. For these addresses,
4816 slashes and pipe symbols are allowed within local parts, but the sequence
4817 /../ is explicitly forbidden.
4819 31. SPA server authentication was not clearing the challenge buffer before
4822 32. log_message in a "warn" ACL statement was writing to the reject log as
4823 well as to the main log, which contradicts the documentation and doesn't
4824 seem right (because no rejection is happening). So I have stopped it.
4826 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4827 However, I am unable to do any testing of this.
4829 34. Fixed an infelicity in the appendfile transport. When checking directories
4830 for a mailbox, to see if any needed to be created, it was accidentally
4831 using path names with one or more superfluous leading slashes; tracing
4832 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4834 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4835 contents are added to the log line that is written for every discarded
4836 recipient. (Previously a log_message setting was ignored.)
4838 36. The ${quote: operator now quotes the string if it is empty.
4840 37. The install script runs exim in order to find its version number. If for
4841 some reason other than non-existence or emptiness, which it checks, it
4842 could not run './exim', it was installing it with an empty version number,
4843 i.e. as "exim-". This error state is now caught, and the installation is
4846 38. An argument was missing from the function that creates an error message
4847 when Exim fails to connect to the socket for saslauthd authentication.
4848 This could cause Exim to crash, or give a corrupted message.
4850 39. Added isip, isip4, and isip6 to ${if conditions.
4852 40. The ACL variables $acl_xx are now saved with the message, and can be
4853 accessed later in routers, transports, and filters.
4855 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4856 strings in the file are not string-expanded.
4858 42. If a MAIL command specified a SIZE value that was too large to fit into an
4859 int variable, the check against message_size_limit failed. Such values are
4860 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4861 day this will have to be increased, but I don't think I want to be around
4862 when emails are that large.
4869 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4870 iconv() is not standard in FreeBSD.
4872 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4873 IPv6 enabled. The observed symptom was a segmentation fault on return from
4874 the function os_common_find_running_interfaces() in src/os.c.
4876 3. In the check_special_case() function in daemon.c I had used "errno" as an
4877 argument name, which causes warnings on some systems. This was basically a
4878 typo, since it was named "eno" in the comments!
4880 4. The code that waits for the clock to tick (at a resolution of some fraction
4881 of a second) so as to ensure message-id uniqueness was always waiting for
4882 at least one whole tick, when it could have waited for less. [This is
4883 almost certainly not relevant at current processor speeds, where it is
4884 unlikely to ever wait at all. But we try to future-proof.]
4886 5. The function that sleeps for a time interval that includes fractions of a
4887 second contained a race. It did not block SIGALRM between setting the
4888 timer, and suspending (a couple of lines later). If the interval was short
4889 and the sigsuspend() was delayed until after it had expired, the suspension
4890 never ended. On busy systems this could lead to processes getting stuck for
4893 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4894 process, before it forks any delivery processes. The open lookup caching
4895 mechanism meant that the open file or database connection was passed into
4896 the delivery process. The problem was that delivery processes always tidy
4897 up cached lookup data. This could cause a problem for the next delivery
4898 process started by the queue runner, because the external queue runner
4899 process does not know about the closure. So the next delivery process
4900 still has data in the lookup cache. In the case of a file lookup, there was
4901 no problem because closing a file descriptor in a subprocess doesn't affect
4902 the parent. However, if the lookup was caching a connection to a database,
4903 the connection was closed, and the second delivery process was likely to
4904 see errors such as "PGSQL: query failed: server closed the connection
4905 unexpectedly". The problem has been fixed by closing all cached lookups
4906 in a queue runner before running a delivery process.
4908 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4909 seem to have the "const" qualifier which it has on other OS. I've
4912 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4913 *of the same type* (client or server) with the same public name that an
4914 error should be diagnosed.
4916 9. When Exim looked up a host name for an IP address, but failed to find the
4917 original IP address when looking up the host name (a safety check), it
4918 output the message "<ip address> does not match any IP for NULL", which was
4919 confusing, to say the least. The bug was that the host name should have
4920 appeared instead of "NULL".
4922 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4923 user that is built into the binary, and the -C or -D options is used, root
4924 privilege is dropped before the configuration file is read. In addition,
4925 logging is switched to stderr instead of the normal log files. If the
4926 configuration then re-defines the Exim user, the unprivileged environment
4927 is probably not what is expected, so Exim logs a panic warning message (but
4930 However, if deliver_drop_privilege is set, the unprivileged state may well
4931 be exactly what is intended, so the warning has been cut out in that case,
4932 and Exim is allowed to try to write to its normal log files.
4938 1. smtp_return_error_details was not giving details for temporary sender
4939 or receiver verification errors.
4941 2. Diagnose a configuration error if two authenticators have the same public
4944 3. Exim used not to create the message log file for a message until the first
4945 delivery attempt. This could be confusing when incoming messages were held
4946 for policy or load reasons. The message log file is now created at the time
4947 the message is received, and an initial "Received" line is written to it.
4949 4. The automatically generated man page for command line options had a minor
4950 bug that caused no ill effects; however, a more serious problem was that
4951 the procedure for building the man page automatically didn't always
4952 operate. Consequently, release 4.20 contains an out-of-date version. This
4953 shouldn't happen again.
4955 5. When building Exim with embedded Perl support, the script that builds the
4956 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4957 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4959 6. The freeze_tell option was not being used for messages that were frozen on
4960 arrival, either by an ACL or by local_scan().
4962 7. Added the smtp_incomplete_transaction log selector.
4964 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4965 was accepting AUTH without a new EHLO.
4967 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4968 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4971 10. Logging of TCP/IP connections (when configured) now happens in the main
4972 daemon process instead of the child process, so that the TCP/IP connection
4973 count is more accurate (but it can never be perfect).
4975 11. The use of "drop" in a nested ACL was not being handled correctly in the
4976 outer ACL. Now, if condition failure induced by the nested "drop" causes
4977 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4978 or "require"), the connection is dropped.
4980 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4981 that yield "discard" can now be used with an "accept" or a "discard" verb,
4982 but an error is generated for any others (because I can't see a useful way
4983 to define what should happen).
4985 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4986 are now processed in the same way as lines in the Exim configuration file.
4987 In particular, continuation lines are supported.
4989 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4991 15. Added -ti meaning -t -i.
4993 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4994 domains, and warn by logging if others are found.
4996 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4997 returned from GIFCONF when Exim is trying to find the list of interfaces on
4998 a host. The code in os.c has been modified to copy each ifreq to an aligned
4999 structure in all cases.
5001 Also, in some cases, the returned ifreq's were being copied to a 'struct
5002 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5003 means the last couple of bytes of an IPv6 address could be chopped if the
5004 ifreq contained only a normal sockaddr (14 bytes storage).
5006 18. Named domain lists were not supported in the hosts_treat_as_local option.
5007 An entry such as +xxxx was not recognized, and was treated as a literal
5010 19. Ensure that header lines added by a DATA ACL are included in the reject log
5011 if the ACL subsequently rejects the message.
5013 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5014 MD5 (which is deprecated).
5016 21. When testing a filter file using -bf, Exim was writing a message when it
5017 took the sender from a "From " line in the message, but it was not doing so
5018 when it took $return_path from a Return-Path: header line. It now does.
5020 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5021 with a valid header line field name (a series of printing characters
5022 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5024 23. Changed "disc" in the source to "disk" to conform to the documentation and
5025 the book and for uniformity.
5027 24. Ignore Sendmail's -Ooption=value command line item.
5029 25. When execve() failed while trying to run a command in a pipe transport,
5030 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5031 could be confused with a return value of 69 from the command itself. This
5032 has been changed to 127, the value the shell returns if it is asked to run
5033 a non-existent command. The wording for the related log line suggests a
5034 non-existent command as the problem.
5036 26. If received_header_text expands to an empty string, do not add a Received:
5037 header line to the message. (Well, it adds a token one on the spool, but
5038 marks it "old" so that it doesn't get used or transmitted.)
5040 27. Installed eximstats 1.28 (addition of -nt option).
5042 28. There was no check for failure on the call to getsockname() in the daemon
5043 code. This can fail if there is a shortage of resources on the system, with
5044 ENOMEM, for example. A temporary error is now given on failure.
5046 29. Contrary to the C standard, it seems that in some environments, the
5047 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5048 program. Exim now does this explicitly; it affects the formatting of
5049 timestamps using strftime().
5051 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5052 complaints. I've now initialized all the variables, to avoid this.
5054 32. Header lines added by a system filter were not being "seen" during
5055 transport-time rewrites.
5057 33. The info_callback() function passed to OpenSSL is set up with type void
5058 (*)(SSL *, int, int), as described somewhere. However, when calling the
5059 function (actually a macro) that sets it up, the type void(*)() is
5060 expected. I've put in a cast to prevent warnings from picky compilers.
5062 34. If a DNS black list lookup found a CNAME record, but there were no A
5063 records associated with the domain it pointed at, Exim crashed.
5065 35. If a DNS black list lookup returned more than one A record, Exim ignored
5066 all but the first. It now scans all returned addresses if a particular IP
5067 value is being sought. In this situation, the contents of the
5068 $dnslist_value variable are a list of all the addresses, separated by a
5071 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5072 to accept a host name and all its aliases if the forward lookup for any of
5073 them yielded the IP address of the incoming connection. Now it accepts only
5074 those names whose forward lookup yields the correct IP address. Any other
5075 names are discarded. This closes a loophole whereby a rogue DNS
5076 administrator could create reverse DNS records to break through a
5077 wildcarded host restriction in an ACL.
5079 37. If a user filter or a system filter that ran in a subprocess used any of
5080 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5081 the wrong values were passed to the pipe command ($thisaddress had the
5082 value of $0, $0 had the value of $1, etc). This bug was introduced by
5083 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5085 38. Improved the line breaking for long SMTP error messages from ACLs.
5086 Previously, if there was no break point between 40 and 75 characters, Exim
5087 left the rest of the message alone. Two changes have been made: (a) I've
5088 reduced the minimum length to 35 characters; (b) if it can't find a break
5089 point between 35 and 75 characters, it looks ahead and uses the first one
5090 that it finds. This may give the occasional overlong line, but at least the
5091 remaining text gets split now.
5093 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5094 file descriptors might be low, and that setting 1000 would always raise it.
5095 It turns out that in some environments, the limit is already over 1000 and
5096 that lowering it causes trouble. So now Exim takes care not to decrease it.
5098 40. When delivering a message, the value of $return_path is set to $sender_
5099 address at the start of routing (routers may change the value). By an
5100 oversight, this default was not being set up when an address was tested by
5101 -bt or -bv, which affected the outcome if any router or filter referred to
5104 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5105 log only when "message" or "log_message" are set. However, if one of the
5106 conditions was an address verification, or a call to a nested ACL, the
5107 messages generated by the underlying test were being passed through. This
5108 no longer happens. The underlying message is available in $acl_verify_
5109 message for both "message" and "log_message" expansions, so it can be
5110 passed through if needed.
5112 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5113 new expansion $bh_ to give the encoded byte string without charset
5114 translation. Translation happens only if iconv() is available; HAVE_ICONV
5115 indicates this at build time. HEADERS_CHARSET gives the charset to
5116 translate to; headers_charset can change it in the configuration, and
5117 "headers charset" can change it in an individual filter file.
5119 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5120 that creates RFC 2047 encoded "words" labels them as that charset instead
5121 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5122 expansion operator; (ii) when Exim creates a From: line for a local
5123 message; (iii) when a header line is rewritten to include a "phrase" part.
5125 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5126 buggy, causing it to skip the first lines of messages whose message ID
5127 ended in 'D'. This would not have bitten before Exim release 4.14, because
5128 message IDs were unlikely to end in 'D' before then. The effect was to have
5129 incorrect size information for certain domains.
5131 45. #include "config.h" was missing at the start of the crypt16.c module. This
5132 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5135 46. If there was a timeout during a "random" callout check, Exim treated it as
5136 a failure of the random address, and carried on sending RSET and the real
5137 address. If the delay was just some slowness somewhere, the response to the
5138 original RCPT would be taken as a response to RSET and so on, causing
5139 mayhem of various kinds.
5141 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5142 when I implemented it. It didn't allow for the fact that some option values
5143 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5144 the right test for positive values.
5146 48. Domain names in DNS records are case-independent. Exim always looks them up
5147 in lower case. Some resolvers return domain names in exactly the case they
5148 appear in the zone file, that is, they may contain uppercase letters. Not
5149 all resolvers do this - some return always lower case. Exim was treating a
5150 change of case by a resolver as a change of domain, similar to a widening
5151 of a domain abbreviation. This triggered its re-routing code and so it was
5152 trying to route what was effectively the same domain again. This normally
5153 caused routing to fail (because the router wouldn't handle the domain
5154 twice). Now Exim checks for this case specially, and just changes the
5155 casing of the domain that it ultimately uses when it transmits the message
5158 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5161 50. If a filter generated a file delivery with a non-absolute name (possible if
5162 no home directory exists for the router), the forbid_file option was not
5165 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5166 the existing equality matching.
5168 52. Exim was using ints instead of ino_t variables in some places where it was
5169 dealing with inode numbers.
5171 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5172 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5173 called TMPDIR, and if it finds it is different, it changes its value.
5175 54. The smtp_printf() function is now made available to local_scan() so
5176 additional output lines can be written before returning. There is also an
5177 smtp_fflush() function to enable the detection of a dropped connection.
5178 The variables smtp_input and smtp_batched_input are exported to
5181 55. Changed the default runtime configuration: the message "Unknown user"
5182 has been removed from the ACL, and instead placed on the localuser router,
5183 using the cannot_route_message feature. This means that any verification
5184 failures that generate their own messages won't get overridden. Similarly,
5185 the "Unrouteable address" message that was in the ACL for unverifiable
5186 relay addresses has also been removed.
5188 56. Added hosts_avoid_esmtp to the smtp transport.
5190 57. The exicyclog script was not checking for the esoteric option
5191 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5192 will work only if exicyclog is run under the appropriate euid.
5194 58. Following a discussion on the list, the rules by which Exim recognises line
5195 endings on incoming messages have been changed. The -dropcr and drop_cr
5196 options are now no-ops, retained only for backwards compatibility. The
5197 following line terminators are recognized: LF CRLF CR. However, special
5198 processing applies to CR:
5200 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5201 nor a local message in the state where . is a terminator.
5203 (ii) If a bare CR is encountered in a header line, an extra space is added
5204 after the line terminator so as not to end the header. The reasoning
5205 behind this is that bare CRs in header lines are most likely either
5206 to be mistakes, or people trying to play silly games.
5208 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5209 was being incorrectly given as 18 bytes larger than it should have been.
5210 This is a VOB (very old bug).
5212 60. This may never have affected anything current, but just in case it has:
5213 When the local host is found other than at the start of a list of hosts,
5214 the local host, those with the same MX, and any that follow, are discarded.
5215 When the list in question was part of a longer list of hosts, the following
5216 hosts (not currently being processed) were also being discarded. This no
5217 longer happens. I'm not sure if this situation could ever has previously
5220 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5223 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5224 header. This is something that is recommended in a new Internet Draft, and
5225 is something that is documented as being done by Sendmail. There are two
5226 possible values. For messages generated by the autoreply transport, Exim
5229 Auto-Submitted: auto-replied
5231 whereas for all other generated messages (e.g. bounces) it adds
5233 Auto-Submitted: auto-generated
5235 63. The "personal" condition in filters now includes a test for the
5236 Auto-Submitted: header. If it contains the string "auto-" the message it
5237 not considered personal.
5239 64. Added rcpt_include_affixes as a generic transport option.
5241 65. Added queue_only_override (default true).
5243 66. Added the syslog_duplication option.
5245 67. If what should have been the first header line of a message consisted of
5246 a space followed by a colon, Exim was mis-interpreting it as a header line.
5247 It isn't of course - it is syntactically invalid and should therefore be
5248 treated as the start of the message body. The misbehaviour could have
5249 caused a number of strange effects, including loss of data in subsequent
5250 header lines, and spool format errors.
5252 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5253 client host had authenticated. This control can now be exercised by an ACL
5254 for more flexibility.
5256 69. By default, callouts do not happen when testing with -bh. There is now a
5257 variant, -bhc, which does actually run the callout code, including
5258 consulting and updating the callout cache.
5260 70. Added support for saslauthd authentication, courtesy of Alexander
5263 71. If statvfs() failed on the spool or log directories while checking their
5264 size for availability, Exim confusingly gave the error "space shortage".
5265 Furthermore, in debugging mode it crashed with a floating point exception.
5266 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5267 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5268 serious problem, Exim now writes to the main and panic logs when this
5269 happens, with details of the failure. It then refuses to accept the
5270 incoming message, giving the message "spool directory problem" or "log
5271 directory problem" with a 421 code for SMTP messages.
5273 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5274 0, 1, and 2 exist, because some OS complain for execs without them (see
5275 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5276 descriptors. However, the code omitted to check that the open succeeded,
5277 causing mysterious errors if for some reason the permissions on /dev/null
5278 got screwed. Now Exim writes a message to the main and panic logs, and
5279 bombs out if it can't open /dev/null.
5281 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5282 interact so that it is all more flexible. It is supposed to remain
5283 backwards compatible. Also added extra_local_interfaces.
5285 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5286 to bomb out with an assertion failure - to the client this appears as a
5287 connection drop. This problem occurs in the part of the code that was taken
5288 from the Samba project. Fortunately, the assertion is in a very simple
5289 function, so I have fixed this by reproducing the function inline in the
5290 one place where it is called, and arranging for authentication to fail
5291 instead of killing the process with assert().
5293 75. The SPA client code was not working when the server requested OEM rather
5294 than Unicode encoding.
5296 76. Added code to make require_files with a specific uid setting more usable in
5297 the case where statting the file as root fails - usually a non-root-mounted
5298 NFS file system. When this happens and the failure is EACCES, Exim now
5299 forks a subprocess and does the per-uid checking as the relevant uid.
5301 77. Added process_log_path.
5303 78. If log_file_path was not explicitly set, a setting of check_log_space or
5304 check_log_inodes was ignored.
5306 79. If a space check for the spool or log partitions fails, the incident is now
5307 logged. Of course, in the latter case the data may get lost...
5309 80. Added the %p formatting code to string_format() so that it can be used to
5310 print addresses in debug_print(). Adjusted all the address printing in the
5311 debugging in store.c to use %p rather than %d.
5313 81. There was a concern that a line of code in smtp_in.c could overflow a
5314 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5315 initially expressed, the concern was not well-founded, because trailing
5316 spaces are removed early. However, if the trailing spaces were followed by
5317 a NULL, they did not get removed, so the overflow was possible. Two fixes
5320 (a) I re-wrote the offending code in a cleaner fashion.
5321 (b) If an incoming SMTP command contains a NULL character, it is rejected
5324 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5325 runs initgroups(), so that if the Exim user is in any additional groups,
5326 they will be used during message reception.
5332 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob