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.
18 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
21 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
22 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
24 PP/07 Catch divide-by-zero in ${eval:...}.
27 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
34 NM/01 Workround for PCRE version dependency in version reporting
37 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
38 This fixes portability to compilers other than gcc, notably
39 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
41 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
42 makefiles for portability to HP-UX and POSIX correctness.
44 PP/01 Permit LOOKUP_foo enabling on the make command-line.
45 Also via indented variable definition in the Makefile.
46 (Debugging by Oliver Heesakkers).
48 PP/02 Restore caching of spamd results with expanded spamd_address.
49 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
51 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
52 Improves build reliability. Fix from: Frank Elsner
54 NM/02 Fix wide character breakage in the rfc2047 coding
55 Fixes bug 1064. Patch from Andrey N. Oktyabrski
57 NM/03 Allow underscore in dnslist lookups
58 Fixes bug 1026. Patch from Graeme Fowler
60 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
61 Code patches from Adam Ciarcinski of NetBSD.
63 NM/04 Fixed exiqgrep to cope with mailq missing size issue
66 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
67 is logged, to avoid truncation. Patch from John Horne.
69 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
70 Patch from Jakob Hirsch.
72 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
73 SQL string expansion failure details.
74 Patch from Andrey Oktyabrski.
76 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
77 Patch from Simon Arlott.
79 PP/09 New lookups functionality failed to compile on old gcc which rejects
80 extern declarations in function scope.
81 Patch from Oliver Fleischmann
83 PP/10 Use sig_atomic_t for flags set from signal handlers.
84 Check getgroups() return and improve debugging.
85 Fixed developed for diagnosis in bug 927 (which turned out to be
88 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
89 Patch from Mark Zealey.
91 PP/12 Bugzilla 1056: Improved spamd server selection.
92 Patch from Mark Zealey.
94 PP/13 Bugzilla 1086: Deal with maildir quota file races.
95 Based on patch from Heiko Schlittermann.
97 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
98 Patch from Uwe Doering, sign-off by Michael Haardt.
100 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
101 variable declaration deep within a block. Bug and patch from
104 PP/15 lookups-Makefile IRIX compatibilty coercion.
106 PP/16 Make DISABLE_DKIM build knob functional.
108 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
109 Patch by Simon Arlott
111 TF/03 Fix valgrind.h portability to C89 compilers that do not support
112 variable argument macros. Our copy now differs from upstream.
118 TF/01 Failure to get a lock on a hints database can have serious
119 consequences so log it to the panic log.
121 TF/02 Log LMTP confirmation messages in the same way as SMTP,
122 controlled using the smtp_confirmation log selector.
124 TF/03 Include the error message when we fail to unlink a spool file.
126 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
127 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
128 for maintaining out-of-tree patches for some time.
130 PP/01 Bugzilla 139: Documentation and portability issues.
131 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
132 Handle per-OS dynamic-module compilation flags.
134 PP/02 Let /dev/null have normal permissions.
135 The 4.73 fixes were a little too stringent and complained about the
136 permissions on /dev/null. Exempt it from some checks.
137 Reported by Andreas M. Kirchwitz.
139 PP/03 Report version information for many libraries, including
140 Exim version information for dynamically loaded libraries. Created
141 version.h, now support a version extension string for distributors
142 who patch heavily. Dynamic module ABI change.
144 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
145 privilege escalation vulnerability whereby the Exim run-time user
146 can cause root to append content of the attacker's choosing to
149 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
152 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
153 If dropping privileges for untrusted macros, we disabled normal logging
154 on the basis that it would fail; for the Exim run-time user, this is not
155 the case, and it resulted in successful deliveries going unlogged.
156 Fixed. Reported by Andreas Metzler.
162 PP/01 Date: & Message-Id: revert to normally being appended to a message,
163 only prepend for the Resent-* case. Fixes regression introduced in
164 Exim 4.70 by NM/22 for Bugzilla 607.
166 PP/02 Include check_rfc2047_length in configure.default because we're seeing
167 increasing numbers of administrators be bitten by this.
169 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
171 PP/03 Bugzilla 994: added openssl_options main configuration option.
173 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
175 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
177 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
179 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
180 without a peer certificate, leading to a segfault because of an
181 assumption that peers always have certificates. Be a little more
182 paranoid. Problem reported by Martin Tscholak.
184 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
185 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
186 NB: ClamAV planning to remove STREAM in "middle of 2010".
187 CL also introduces -bmalware, various -d+acl logging additions and
188 more caution in buffer sizes.
190 PP/09 Implemented reverse_ip expansion operator.
192 PP/10 Bugzilla 937: provide a "debug" ACL control.
194 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
196 PP/12 Bugzilla 973: Implement --version.
198 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
200 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
202 PP/15 Bugzilla 816: support multiple condition rules on Routers.
204 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
205 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
206 ignore trailing whitespace.
208 JJ/02 prevent non-panic DKIM error from being sent to paniclog
210 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
213 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
214 Notification from Dr Andrew Aitchison.
216 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
217 ExtendedDetectionInfo response format.
218 Notification from John Horne.
220 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
223 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
224 XSL and documented dependency on system catalogs, with examples of how
227 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
230 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
231 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
232 configuration file which is writeable by the Exim user or group.
234 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
235 of configuration files to cover files specified with the -C option if
236 they are going to be used with root privileges, not just the default
239 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
240 option (effectively making it always true).
242 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
243 files to be used while preserving root privileges.
245 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
246 that rogue child processes cannot use them.
248 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
249 run-time user, instead of root.
251 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
252 Exim run-time user without dropping privileges.
254 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
255 result string, instead of calling string_vformat() twice with the same
258 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
259 for other users. Others should always drop root privileges if they use
260 -C on the command line, even for a whitelisted configure file.
262 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
264 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
270 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
271 $data_path, and $header_path variables; fixed documentation bugs and
274 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
275 exipick to access non-standard spools, including the "frozen" queue
278 NM/01 Bugzilla 965: Support mysql stored procedures.
279 Patch from Alain Williams
281 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
283 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
284 Patch from Andreas Metzler
286 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
287 Patch from Kirill Miazine
289 NM/05 Bugzilla 671: Added umask to procmail example.
291 JJ/03 installed exipick 20100323.0, fixing doc bug
293 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
294 directory. Notification and patch from Dan Rosenberg.
296 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
298 TK/02 Improve log output when DKIM signing operation fails.
300 MH/01 Treat the transport option dkim_domain as a colon separated
301 list, not as a single string, and sign the message with each element,
302 omitting multiple occurences of the same signer.
304 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
305 Bugzilla 985, 986. Patch by Simon Arlott
307 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
308 Patch by Simon Arlott
310 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
311 MBX locking. Notification from Dan Rosenberg.
317 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
319 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
321 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
323 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
325 NM/04 Bugzilla 915: Flag broken perl installation during build.
331 TK/01 Added patch by Johannes Berg that expands the main option
332 "spamd_address" if it starts with a dollar sign.
334 TK/02 Write list of recipients to X-Envelope-Sender header when building
335 the mbox-format spool file for content scanning (suggested by Jakob
338 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
339 (http://www.dcc-servers.net/) support via dccifd. Activated by
340 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
342 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
343 by Mark Daniel Reidel <mr@df.eu>.
345 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
346 When building exim an external PCRE library is now needed -
347 PCRE is a system library on the majority of modern systems.
348 See entry on PCRE_LIBS in EDITME file.
350 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
351 conversation. Added nologin parameter to request.
352 Patch contributed by Kirill Miazine.
354 TF/01 Do not log submission mode rewrites if they do not change the address.
356 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
358 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
359 log files in place. Contributed by Roberto Lima.
361 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
363 TF/03 Bugzilla 615: When checking the local_parts router precondition
364 after a local_part_suffix or local_part_prefix option, Exim now
365 does not use the address's named list lookup cache, since this
366 contains cached lookups for the whole local part.
368 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
369 Robert Millan. Documentation is in experimental-spec.txt.
371 TF/04 Bugzilla 668: Fix parallel build (make -j).
373 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
375 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
376 Patch provided by Jan Srzednicki.
378 TF/05 Leading white space used to be stripped from $spam_report which
379 wrecked the formatting. Now it is preserved.
381 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
382 that they are available at delivery time.
384 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
386 TF/08 TLS error reporting now respects the incoming_interface and
387 incoming_port log selectors.
389 TF/09 Produce a more useful error message if an SMTP transport's hosts
390 setting expands to an empty string.
392 NM/06 Bugzilla 744: EXPN did not work under TLS.
393 Patch provided by Phil Pennock.
395 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
396 Patch provided by Richard Godbee.
398 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
399 acl_smtp_notquit, added index entry.
401 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
402 Patch provided by Eugene Bujak.
404 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
405 accept(). Patch provided by Maxim Dounin.
407 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
408 Patch provided by Phil Pennock.
410 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
412 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
413 Patch provided by Brad "anomie" Jorsch.
415 NM/14 Bugzilla 622: Added timeout setting to transport filter.
416 Patch provided by Dean Brooks.
418 TK/05 Add native DKIM support (does not depend on external libraries).
420 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
421 Patch provided by Graeme Fowler.
423 NM/16 Bugzilla 851: Documentation example syntax fix.
425 NM/17 Changed NOTICE file to remove references to embedded PCRE.
427 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
430 NM/19 Bugzilla 745: TLS version reporting.
431 Patch provided by Phil Pennock.
433 NM/20 Bugzilla 167: bool: condition support.
434 Patch provided by Phil Pennock.
436 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
437 clients. Patch provided by Phil Pennock.
439 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
440 Patch provided by Brad "anomie" Jorsch.
442 NM/23 Bugzilla 687: Fix misparses in eximstats.
443 Patch provided by Heiko Schlittermann.
445 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
446 Patch provided by Heiko Schlittermann.
448 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
449 plus update to original patch.
451 NM/26 Bugzilla 799: Documentation correction for ratelimit.
453 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
454 Patch provided by David Brownlee.
456 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
458 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
460 NM/30 Bugzilla 888: TLS documentation bugfixes.
462 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
464 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
465 Unlike the original bugzilla I have changed all shell scripts in src tree.
467 NM/33 Bugzilla 898: Transport filter timeout fix.
468 Patch by Todd Rinaldo.
470 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
471 Patch by Serge Demonchaux.
473 NM/35 Bugzilla 39: Base64 decode bug fixes.
474 Patch by Jakob Hirsch.
476 NM/36 Bugzilla 909: Correct connect() call in dcc code.
478 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
480 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
482 NM/39 Bugzilla 911: Fixed MakeLinks build script.
488 TK/01 Add preliminary DKIM support. Currently requires a forked version of
489 ALT-N's libdkim that I have put here:
490 http://duncanthrax.net/exim-experimental/
492 Note to Michael Haardt: I had to rename some vars in sieve.c. They
493 were called 'true' and it seems that C99 defines that as a reserved
494 keyword to be used with 'bool' variable types. That means you could
495 not include C99-style headers which use bools without triggering
496 build errors in sieve.c.
498 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
499 as mailq or other aliases. Changed the --help handling significantly
500 to do whats expected. exim_usage() emits usage/help information.
502 SC/01 Added the -bylocaldomain option to eximstats.
504 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
506 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
508 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
514 PH/01 Another patch from the Sieve maintainer.
516 PH/02 When an IPv6 address is converted to a string for single-key lookup
517 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
518 dots are used instead of colons so that keys in lsearch files need not
519 contain colons. This was done some time before quoting was made available
520 in lsearch files. However, iplsearch files do require colons in IPv6 keys
521 (notated using the quote facility) so as to distinguish them from IPv4
522 keys. This meant that lookups for IP addresses in host lists did not work
523 for iplsearch lookups.
525 This has been fixed by arranging for IPv6 addresses to be expressed with
526 colons if the lookup type is iplsearch. This is not incompatible, because
527 previously such lookups could never work.
529 The situation is now rather anomolous, since one *can* have colons in
530 ordinary lsearch keys. However, making the change in all cases is
531 incompatible and would probably break a number of configurations.
533 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
536 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
537 conversion specification without a maximum field width, thereby enabling
538 a rogue spamd server to cause a buffer overflow. While nobody in their
539 right mind would setup Exim to query an untrusted spamd server, an
540 attacker that gains access to a server running spamd could potentially
541 exploit this vulnerability to run arbitrary code as the Exim user.
543 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
544 $primary_hostname instead of what libspf2 thinks the hosts name is.
546 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
547 a directory entry by the name of the lookup key. Previously, if a
548 symlink pointed to a non-existing file or a file in a directory that
549 Exim lacked permissions to read, a lookup for a key matching that
550 symlink would fail. Now it is enough that a matching directory entry
551 exists, symlink or not. (Bugzilla 503.)
553 PH/03 The body_linecount and body_zerocount variables are now exported in the
556 PH/04 Added the $dnslist_matched variable.
558 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
559 This means they are set thereafter only if the connection becomes
562 PH/06 Added the client_condition to authenticators so that some can be skipped
563 by clients under certain conditions.
565 PH/07 The error message for a badly-placed control=no_multiline_responses left
566 "_responses" off the end of the name.
568 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
570 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
571 (without spaces) instead of just copying the configuration text.
573 PH/10 Added the /noupdate option to the ratelimit ACL condition.
575 PH/11 Added $max_received_linelength.
577 PH/12 Added +ignore_defer and +include_defer to host lists.
579 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
580 way in which PCRE > 7.0 is built.
582 PH/14 Implemented queue_only_load_latch.
584 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
585 MAIL command. The effect was to mangle the value on 64-bit systems.
587 PH/16 Another patch from the Sieve maintainer.
589 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
591 PH/18 If a system quota error occurred while trying to create the file for
592 a maildir delivery, the message "Mailbox is full" was not appended to the
593 bounce if the delivery eventually timed out. Change 4.67/27 below applied
594 only to a quota excession during the actual writing of the file.
596 PH/19 It seems that peer DN values may contain newlines (and other non-printing
597 characters?) which causes problems in log lines. The DN values are now
598 passed through string_printing() before being added to log lines.
600 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
601 and InterBase are left for another time.)
603 PH/21 Added message_body_newlines option.
605 PH/22 Guard against possible overflow in moan_check_errorcopy().
607 PH/23 POSIX allows open() to be a macro; guard against that.
609 PH/24 If the recipient of an error message contained an @ in the local part
610 (suitably quoted, of course), incorrect values were put in $domain and
611 $local_part during the evaluation of errors_copy.
617 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
618 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
621 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
622 issue a MAIL command.
624 PH/02 In an ACL statement such as
626 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
628 if a client was not listed at all, or was listed with a value other than
629 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
630 the condition was not true (as it should be), so access was not denied.
631 The bug was that the ! inversion was incorrectly passed on to the second
632 item. This has been fixed.
634 PH/03 Added additional dnslists conditions == and =& which are different from
635 = and & when the dns lookup returns more than one IP address.
637 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
638 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
640 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
641 FSYNC, which compiles an option called disable_fsync that allows for
642 bypassing fsync(). The documentation is heavily laced with warnings.
644 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
646 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
647 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
648 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
649 including adding "make clean"; (3) Added -fPIC when compiling the test
650 dynamically loaded module, to get rid of a warning.
652 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
653 message fails, move_frozen_messages = true and ignore_bounce_errors_after
654 = 0s. The bug is otherwise harmless.
656 PH/07 There was a bug in the dovecot authenticator such that the value of
657 $auth1 could be overwritten, and so not correctly preserved, after a
658 successful authentication. This usually meant that the value preserved by
659 the server_setid option was incorrect.
661 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
663 PH/09 Installed PCRE release 7.0.
665 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
666 run for batched SMTP input. It is now run at the start of every message
667 in the batch. While fixing this I discovered that the process information
668 (output by running exiwhat) was not always getting set for -bs and -bS
669 input. This is fixed, and it now also says "batched" for BSMTP.
671 PH/11 Added control=no_pipelining.
673 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
674 patch, slightly modified), and move the expansion of helo_data till after
675 the connection is made in the smtp transport (so it can use these
678 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
680 PH/14 Added log_selector = +pid.
682 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
684 PH/16 Add ${if forany and ${if forall.
686 PH/17 Added dsn_from option to vary the From: line in DSNs.
688 PH/18 Flush SMTP output before performing a callout, unless control =
689 no_callout_flush is set.
691 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
692 was true (the default) a successful delivery failed to delete the retry
693 item, thus causing premature timeout of the address. The bug is now
696 PH/20 Added hosts_avoid_pipelining to the smtp transport.
698 PH/21 Long custom messages for fakedefer and fakereject are now split up
699 into multiline reponses in the same way that messages for "deny" and
700 other ACL rejections are.
702 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
703 with slight modification.
705 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
706 draft, changing the syntax and factoring some duplicate code".
708 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
709 for deliveries of the second and subsequent messages over the same SMTP
712 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
713 ${reduce, with only minor "tidies".
715 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
717 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
718 expansion side effects.
720 PH/27 When a message times out after an over-quota error from an Exim-imposed
721 quota, the bounce message says "mailbox is full". This message was not
722 being given when it was a system quota that was exceeded. It now should
725 MH/03 Made $recipients available in local_scan(). local_scan() already has
726 better access to the recipient list through recipients_list[], but
727 $recipients can be useful in postmaster-provided expansion strings.
729 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
730 in the case of a MAIL command with additional options following the
731 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
732 were accidentally chopped off.
734 PH/29 SMTP synchronization checks are implemented when a command is read -
735 there is a check that no more input is waiting when there shouldn't be
736 any. However, for some commands, a delay in an ACL can mean that it is
737 some time before the response is written. In this time, more input might
738 arrive, invalidly. So now there are extra checks after an ACL has run for
739 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
740 pipelining has not been advertised.
742 PH/30 MH's patch to allow iscntrl() characters to be list separators.
744 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
745 returned in the SMTP response when smtp_return_error_details was false.
748 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
749 instead of the C I/O that was originally supplied, because problems were
752 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
753 Exim which did not show up earlier: it was assuming that a call to
754 SSL_CTX_set_info_callback() might give an error value. In fact, there is
755 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
756 was a macro that became an assignment, so it seemed to work. This has
757 changed to a proper function call with a void return, hence the compile
758 error. Exim's code has been fixed.
760 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
763 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
765 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
767 PH/37 If a message is not accepted after it has had an id assigned (e.g.
768 because it turns out to be too big or there is a timeout) there is no
769 "Completed" line in the log. When some messages of this type were
770 selected by exigrep, they were listed as "not completed". Others were
771 picked up by some special patterns. I have improved the selection
772 criteria to be more general.
774 PH/38 The host_find_failed option in the manualroute router can now be set
775 to "ignore", to completely ignore a host whose IP address cannot be
776 found. If all hosts are ignored, the behaviour is controlled by the new
777 host_all_ignored option.
779 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
780 homing or because of multiple MX records with /mx) generated more than
781 one IP address, and the following item turned out to be the local host,
782 all the secondary addresses of the first item were incorrectly removed
783 from the list, along with the local host and any following hosts (which
784 is what is supposed to happen).
786 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
787 whoever called Exim into the -H file. In the case of the daemon it was
788 behaving confusingly. When first started, it used values for whoever
789 started the daemon, but after a SIGHUP it used the Exim user (because it
790 calls itself on a restart). I have changed the code so that it now always
793 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
794 message are rejected with the same error (e.g. no authentication or bad
795 sender address), and a DATA command is nevertheless sent (as can happen
796 with PIPELINING or a stupid MUA), the error message that was given to the
797 RCPT commands is included in the rejection of the DATA command. This is
798 intended to be helpful for MUAs that show only the final error to their
801 PH/42 Another patch from the Sieve maintainer.
803 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
804 Eximstats - Fixed some broken HTML links and added missing column headers
806 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
807 columns for Rejects, Temp Rejects, Ham, and Spam rows.
809 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
811 PH/43 Yet another patch from the Sieve maintainer.
813 PH/44 I found a way to check for a TCP/IP connection going away before sending
814 the response to the final '.' that terminates a message, but only in the
815 case where the client has not sent further data following the '.'
816 (unfortunately, this is allowed). However, in many cases there won't be
817 any further data because there won't be any more messages to send. A call
818 to select() can be used: if it shows that the input is "ready", there is
819 either input waiting, or the socket has been closed. An attempt to read
820 the next input character can distinguish the two cases. Previously, Exim
821 would have sent an OK response which the client would never have see.
822 This could lead to message repetition. This fix should cure that, at
823 least in a lot of common cases.
825 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
826 advertised in response to EHLO.
832 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
833 fixed by 4.65/MH/01 (is this a record?) are fixed:
835 (i) An empty string was always treated as zero by the numeric comparison
836 operators. This behaviour has been restored.
838 (ii) It is documented that the numeric comparison operators always treat
839 their arguments as decimal numbers. This was broken in that numbers
840 starting with 0 were being interpreted as octal.
842 While fixing these problems I realized that there was another issue that
843 hadn't been noticed. Values of message_size_limit (both the global option
844 and the transport option) were treated as octal if they started with 0.
845 The documentation was vague. These values are now always treated as
846 decimal, and I will make that clear in the documentation.
852 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
853 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
856 MH/01 Don't check that the operands of numeric comparison operators are
857 integers when their expansion is in "skipping" mode (fixes bug
858 introduced by 4.64-PH/07).
860 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
861 child addresses, Exim now panics and dies. Previously, because the count
862 is held in a short int, deliveries were likely to be lost. As such a
863 large number of recipients for a single message is ridiculous
864 (performance will be very, very poor), I have chosen to impose a limit
865 rather than extend the field.
871 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
872 leftover -K file (the existence of which was triggered by #402).
873 While we were at it, introduced process PID as part of the -K
874 filename. This should rule out race conditions when creating
877 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
878 processing considerably. Previous code took too long for large mails,
879 triggering a timeout which in turn triggers #401.
881 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
882 in the DK code in transports.c. sendfile() is not really portable,
883 hence the _LINUX specificness.
885 TF/01 In the add_headers option to the mail command in an Exim filter,
886 there was a bug that Exim would claim a syntax error in any
887 header after the first one which had an odd number of characters
890 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
891 callout verification, Exim cached a "reject" for the entire domain. This
892 is correct for most verifications, but it is not correct for a recipient
893 verification with use_sender or use_postmaster set, because in that case
894 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
895 case of MAIL FROM:<> rejection from other early rejections (e.g.
896 rejection of HELO). When verifying a recipient using a non-null MAIL
897 address, the cache is ignored if it shows MAIL FROM:<> rejection.
898 Whatever the result of the callout, the value of the domain cache is
899 left unchanged (for any other kind of callout, getting as far as trying
900 RCPT means that the domain itself is ok).
902 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
905 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
906 manifest itself as EPIPE rather than ECONNECT. When tidying away a
907 session, the daemon ignores ECONNECT errors and logs others; it now
908 ignores EPIPE as well.
910 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
911 (quoted-printable decoding).
913 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
914 later the small subsequent patch to fix an introduced bug.
916 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
918 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
920 PH/08 An error is now given if message_size_limit is specified negative.
922 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
923 to be given (somewhat) arbitrary names.
925 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
928 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
929 miscellaneous code fixes
931 PH/10 Added the log_reject_target ACL modifier to specify where to log
934 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
935 hostname. This is wrong, because it relates to the incoming message (and
936 probably the interface on which it is arriving) and not to the outgoing
937 callout (which could be using a different interface). This has been
938 changed to use the value of the helo_data option from the smtp transport
939 instead - this is what is used when a message is actually being sent. If
940 there is no remote transport (possible with a router that sets up host
941 addresses), $smtp_active_hostname is used.
943 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
944 tweaks were necessary in order to get it to work (see also 21 below):
945 (a) The code assumed that strncpy() returns a negative number on buffer
946 overflow, which isn't the case. Replaced with Exim's string_format()
948 (b) There were several signed/unsigned issues. I just did the minimum
949 hacking in of casts. There is scope for a larger refactoring.
950 (c) The code used strcasecmp() which is not a standard C function.
951 Replaced with Exim's strcmpic() function.
952 (d) The code set only $1; it now sets $auth1 as well.
953 (e) A simple test gave the error "authentication client didn't specify
954 service in request". It would seem that Dovecot has changed its
955 interface. Fortunately there's a specification; I followed it and
956 changed what the client sends and it appears to be working now.
958 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
961 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
962 address A is aliased to B and C, where B exists and C does not. Without
963 -v the output is "A verified" because verification stops after a
964 successful redirection if more than one address is generated. However,
965 with -v the child addresses are also verified. Exim was outputting "A
966 failed to verify" and then showing the successful verification for C,
967 with its parentage. It now outputs "B failed to verify", showing B's
968 parentage before showing the successful verification of C.
970 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
971 look up a TXT record in a specific list after matching in a combined
974 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
975 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
976 they consult the DNS. I had assumed they would set it the way they
977 wanted; and indeed my experiments on Linux seem to show that in some
978 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
979 To be on the safe side, however, I have now made the interface to
980 host_find_byname() similar to host_find_bydns(), with an argument
981 containing the DNS resolver options. The host_find_byname() function now
982 sets these options at its start, just as host_find_bydns() does. The smtp
983 transport options dns_qualify_single and dns_search_parents are passed to
984 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
985 of host_find_byname() use the default settings of RES_DEFNAMES
986 (qualify_single) but not RES_DNSRCH (search_parents).
988 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
989 spool_read_header() do less string testing, by means of a preliminary
990 switch on the second character of optional "-foo" lines. (This is
991 overdue, caused by the large number of possibilities that now exist.
992 Originally there were few.) While I was there, I also converted the
993 str(n)cmp tests so they don't re-test the leading "-" and the first
994 character, in the hope this might squeeze out yet more improvement.
996 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
997 flag allowing group syntax was set by the header_syntax check but not
998 turned off, possible causing trouble later; (2) The flag was not being
999 set at all for the header_verify test, causing "group"-style headers to
1000 be rejected. I have now set it in this case, and also caused header_
1001 verify to ignore an empty address taken from a group. While doing this, I
1002 came across some other cases where the code for allowing group syntax
1003 while scanning a header line wasn't quite right (mostly, not resetting
1004 the flag correctly in the right place). These bugs could have caused
1005 trouble for malformed header lines. I hope it is now all correct.
1007 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1008 with the "reply" argument non-NULL. The code, however (which originally
1009 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1010 but it didn't always do it. This confused somebody who was copying the
1011 code for some other use. I have removed all the tests.
1013 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1014 feature that was used to support insecure browsers during the U.S. crypto
1015 embargo. It requires special client support, and Exim is probably the
1016 only MTA that supported it -- and would never use it because real RSA is
1017 always available. This code has been removed, because it had the bad
1018 effect of slowing Exim down by computing (never used) parameters for the
1019 RSA_EXPORT functionality.
1021 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1022 authenticator to fail if there's a tab character in the incoming data
1023 (there should never be unless someone is messing about, as it's supposed
1024 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1025 if the connection is using TLS or if the remote IP is the same as the
1026 local IP, and the "valid-client-cert option" if a client certificate has
1029 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1030 authenticators. This can be used for authorization after authentication
1031 succeeds. (In the case of plaintext, it servers for both authentication
1034 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1035 if any retry times were supplied.
1037 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1038 connection, where there is no client IP address to check. In this
1039 situation, the verify now always succeeds.
1041 PH/25 Applied John Jetmore's -Mset patch.
1043 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1045 PH/27 In a string expansion for a processed (not raw) header when multiple
1046 headers of the same name were present, leading whitespace was being
1047 removed from all of them, but trailing whitespace was being removed only
1048 from the last one. Now trailing whitespace is removed from each header
1049 before concatenation. Completely empty headers in a concatenation (as
1050 before) are ignored.
1052 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1053 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1055 PH/29 [Removed. This was a change that I later backed out, and forgot to
1056 correct the ChangeLog entry (that I had efficiently created) before
1057 committing the later change.]
1059 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1060 address errors (4xx response to RCPT) over the same connection as other
1061 messages routed to the same hosts. Such deliveries are always "forced",
1062 so retry times are not inspected. This resulted in far too many retries
1063 for the affected addresses. The effect occurred only when there were more
1064 hosts than the hosts_max_try setting in the smtp transport when it had
1065 the 4xx errors. Those hosts that it had tried were not added to the list
1066 of hosts for which the message was waiting, so if all were tried, there
1067 was no problem. Two fixes have been applied:
1069 (i) If there are any address or message errors in an SMTP delivery, none
1070 of the hosts (tried or untried) are now added to the list of hosts
1071 for which the message is waiting, so the message should not be a
1072 candidate for sending over the same connection that was used for a
1073 successful delivery of some other message. This seems entirely
1074 reasonable: after all the message is NOT "waiting for some host".
1075 This is so "obvious" that I'm not sure why it wasn't done
1076 previously. Hope I haven't missed anything, but it can't do any
1077 harm, as the worst effect is to miss an optimization.
1079 (ii) If, despite (i), such a delivery is accidentally attempted, the
1080 routing retry time is respected, so at least it doesn't keep
1081 hammering the server.
1083 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1084 in ${readsocket because some servers need this prod.
1086 PH/32 Added some extra debug output when updating a wait-xxx database.
1088 PH/33 The hint "could be header name not terminated by colon", which has been
1089 given for certain expansion errors for a long time, was not being given
1090 for the ${if def:h_colon_omitted{... case.
1092 PH/34 The spec says: "With one important exception, whenever a domain list is
1093 being scanned, $domain contains the subject domain." There was at least
1094 one case where this was not true.
1096 PH/35 The error "getsockname() failed: connection reset by peer" was being
1097 written to the panic log as well as the main log, but it isn't really
1098 panic-worthy as it just means the connection died rather early on. I have
1099 removed the panic log writing for the ECONNRESET error when getsockname()
1102 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1103 runs only) independently of the message's sender address. This meant
1104 that, if the 4xx error was in fact related to the sender, a different
1105 message to the same recipient with a different sender could confuse
1106 things. In particualar, this can happen when sending to a greylisting
1107 server, but other circumstances could also provoke similar problems.
1108 I have changed the default so that the retry time for these errors is now
1109 based a combination of the sender and recipient addresses. This change
1110 can be overridden by setting address_retry_include_sender=false in the
1113 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1114 remote server are returned as part of bounce messages. This was not
1115 happening for LMTP over a pipe (the lmtp transport), but now it is the
1116 same for both kinds of LMTP.
1118 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1119 in header lines that were in fact CNAMEs. This is no longer the case.
1121 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1122 and queue runs started by the daemon processed all messages. This has
1123 been fixed so that -R and -S can now usefully be given with -q<time>.
1125 PH/40 Import PCRE release 6.7 (fixes some bugs).
1127 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1129 PH/42 Give an error if -q is specified more than once.
1131 PH/43 Renamed the variables $interface_address and $interface_port as
1132 $received_ip_address and $received_port, to make it clear that these
1133 values apply to message reception, and not to the outgoing interface when
1134 a message is delivered. (The old names remain recognized, of course.)
1136 PH/44 There was no timeout on the connect() call when using a Unix domain
1137 socket in the ${readsocket expansion. There now is.
1139 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1140 be meaningful with "accept".
1142 SC/01 Eximstats V1.43
1143 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1145 SC/02 Eximstats V1.44
1146 Use a glob alias rather than an array ref in the generated
1147 parser. This improves both readability and performance.
1149 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1150 Collect SpamAssassin and rejection statistics.
1151 Don't display local sender or destination tables unless
1152 there is data to show.
1153 Added average volumes into the top table text output.
1155 SC/04 Eximstats V1.46
1156 Collect data on the number of addresses (recipients)
1157 as well as the number of messages.
1159 SC/05 Eximstats V1.47
1160 Added 'Message too big' to the list of mail rejection
1161 reasons (thanks to Marco Gaiarin).
1163 SC/06 Eximstats V1.48
1164 Mainlog lines which have GMT offsets and are too short to
1165 have a flag are now skipped.
1167 SC/07 Eximstats V1.49 (Alain Williams)
1168 Added the -emptyok flag.
1170 SC/08 Eximstats V1.50
1171 Fixes for obtaining the IP address from reject messages.
1173 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1174 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1175 whitesspace changes from 4.64-PH/27
1177 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1180 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1181 are found, allow negative numbers in numeric criteria)
1183 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1185 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1186 to match changes made in 4.64-PH/43
1188 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1190 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1191 "domain" is actually an IP address, and force a failure. This locks out
1192 those revolvers/nameservers that support "A-for-A" lookups, in
1193 contravention of the specifications.
1195 PH/48 When a host name was looked up from an IP address, and the subsequent
1196 forward lookup of the name timed out, the host name was left in
1197 $sender_host_name, contrary to the specification.
1199 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1200 restricted to single-key lookups, Exim was not diagnosing an error if
1201 * or *@ was used with a query-style lookup.
1203 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1205 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1206 long ago, but noone interested enough thought of it. Let's just say that
1207 the "1.1" means that there are some new functions that weren't there at
1208 some point in the past.
1210 PH/51 Error processing for expansion failure of helo_data from an smtp
1211 transport during callout processing was broken.
1213 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1214 tested/used via the -bh/-bhc/-bs options.
1216 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1217 bug, fixed in subsequent PCRE releases).
1219 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1220 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1222 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1228 SC/01 Use a glob alias rather than an array ref in eximstats generated
1229 parser. This improves both readability and performance.
1231 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1232 Don't display local sender or destination tables in eximstats unless
1233 there is data to show.
1234 Added average volumes into the eximstats top table text output.
1236 SC/03 Collect data on the number of addresses (recipients) as well
1237 as the number of messages in eximstats.
1239 TF/01 Correct an error in the documentation for the redirect router. Exim
1240 does (usually) call initgroups() when daemonizing.
1242 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1243 with consistent privilege compared to when running as a daemon.
1245 TF/03 Note in the spec that $authenticated_id is not set for local
1246 submissions from trusted users.
1248 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1249 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1251 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1252 by adding some example configuration directives to the default
1253 configuration file. A little bit of work is required to uncomment the
1254 directives and define how usernames and passwords are checked, but
1255 there is now a framework to start from.
1257 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1258 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1259 without this. I don't know how relevant this is to other LDAP libraries.
1261 PH/02 Add the verb name to the "unknown ACL verb" error.
1263 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1265 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1267 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1268 directory not expanded when it should be if an expanded home directory
1269 was set for the address (which is overridden by the transport).
1271 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1274 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1275 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1276 because it is too late at that time, and has no effect.
1278 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1279 security issue with \' (bugzilla #107). I could not use the
1280 PQescapeStringConn() function, because it needs a PGconn value as one of
1283 PH/08 When testing addresses using -bt, indicate those final addresses that
1284 are duplicates that would not cause an additional delivery. At least one
1285 person was confused, thinking that -bt output corresponded to deliveries.
1286 (Suppressing duplicates isn't a good idea as you lose the information
1287 about possibly different redirections that led to the duplicates.)
1289 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1290 systems where poll() doesn't work, in particular OS X.
1292 PH/10 Added more information to debugging output for retry time not reached.
1294 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1295 operations in malware.c.
1297 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1300 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1301 syslog_duplication set false, log lines that would normally be written
1302 both the the main log and to the reject log were not written to syslog at
1305 PH/14 In the default configuration, change the use of "message" in ACL warn
1306 statements to "add_header".
1308 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1309 not followed by a command (e.g. "seen endif").
1311 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1312 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1315 PH/17 Added extra conditions to the default value of delay_warning_condition
1319 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1320 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1321 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1324 The Auto-Submitted: and various List- headers are standardised, whereas I
1325 don't think Precedence: ever was.
1327 PH/18 Refactored debugging code in route_finduser() to show more information,
1328 in particular, the error code if getpwnam() issues one.
1330 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1331 This is apparently needed in addition to the PH/07 change above to avoid
1332 any possible encoding problems.
1334 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1335 but not after initializing Perl.
1337 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1338 output them only if debugging. By default they are written stderr,
1339 apparently, which is not desirable.
1341 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1344 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1347 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1349 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1350 authenticated or an ident call has been made. Suppress the default
1351 values for $authenticated_id and $authenticated_sender (but permit -oMai
1352 and -oMas) when testing with -bh.
1354 PH/24 Re-jigged the order of the tests in the default configuration so that the
1355 tests for valid domains and recipients precede the DNS black list and CSA
1356 tests, on the grounds that those ones are more expensive.
1358 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1359 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1360 This bug exists in every version of Exim that I still have, right back to
1363 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1364 However, an attempt to turn on case-sensitivity in a regex key by
1365 including (?-i) didn't work because the subject string was already
1366 lowercased, and the effects were non-intuitive. It turns out that a
1367 one-line patch can be used to allow (?-i) to work as expected.
1373 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1374 other effects) broke the use of negated acl sub-conditions.
1376 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1379 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1380 "Deny" causes Exim to reject the incoming connection with a 554 error.
1381 Unfortunately, if there is a major crisis, such as a disk failure,
1382 tcp-wrappers gives "deny", whereas what one would like would be some
1383 kind of temporary error. A kludge has been added to help with this.
1384 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1385 554 error is used if errno is still zero or contains ENOENT (which occurs
1386 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1389 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1391 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1392 errors. Otherwise a message that provokes a temporary error (when other
1393 messages do not) can cause a whole host to time out.
1395 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1396 addresses were routed directly to files or pipes from a redirect router.
1397 File deliveries just didn't batch; pipe deliveries might have suffered
1400 PH/06 A failure to get a lock for a hints database would erroneously always say
1401 "Failed to get write lock", even when it was really a read lock.
1403 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1404 of 0600. This has been changed to use the value of the lockfile_mode
1405 option (which defaults to 0600).
1407 PH/08 Applied small patch from the Sieve maintainer.
1409 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1410 folder from quota calculations, a direct delivery into this folder messed
1411 up the contents of the maildirsize file. This was because the regex was
1412 used only to exclude .Trash (or whatever) when the size of the mailbox
1413 was calculated. There was no check that a delivery was happening into an
1414 excluded directory. This bug has been fixed by ignoring all quota
1415 processing for deliveries into excluded directories.
1417 PH/10 Added the maildirfolder_create_regex option to appendfile.
1423 PH/01 The code for finding all the local interface addresses on a FreeBSD
1424 system running IPv6 was broken. This may well have applied to all BSD
1425 systems, as well as to others that have similar system calls. The broken
1426 code found IPv4 interfaces correctly, but gave incorrect values for the
1427 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1428 that it would not match correctly against @[] and not recognize the IPv6
1431 PH/02 The ipliteral router was not recognizing addresses of the form user@
1432 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1434 PH/03 Added disable_ipv6.
1436 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1437 decoded form, because it is most often used to construct To: headers
1438 lines in autoreplies, and the decoded form may well be syntactically
1439 invalid. However, $reply_address has leading white space removed, and all
1440 newlines turned into spaces so that the autoreply transport does not
1443 PH/05 If group was specified without a user on a router, and no group or user
1444 was specified on a transport, the group from the router was ignored.
1446 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1447 for visible compile-time settings that can be used to change these
1448 numbers, for those that want even more. Backwards compatibility with old
1449 spool files has been maintained. However, going back to a previous Exim
1450 release will lost any variables that are in spool files.
1452 PH/07 Two small changes when running in the test harness: increase delay when
1453 passing a TCP/IP connection to a new process, in case the original
1454 process has to generate a bounce, and remove special handling of
1455 127.0.0.2 (sic), which is no longer necessary.
1457 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1458 be the same on different OS.
1460 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1463 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1464 whether --show-vars was specified or not
1466 JJ/02 exipick: Added support for new ACL variable spool format introduced
1469 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1470 syntactically invalid From: or Reply-to: line, and a filter used this to
1471 generate an autoreply, and therefore failed to obtain an address for the
1472 autoreply, Exim could try to deliver to a non-existent relative file
1473 name, causing unrelated and misleading errors. What now happens is that
1474 it logs this as a hard delivery error, but does not attempt to create a
1477 PH/11 The exinext utility has a -C option for testing purposes, but although
1478 the given file was scanned by exinext itself; it wasn't being passed on
1479 when Exim was called.
1481 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1482 an end-of-file indication when reading a command response.
1484 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1485 compiled. In many other places in Exim, IPv6 addresses are always
1486 recognized, so I have changed this. It also means that IPv4 domain
1487 literals of the form [IPV4:n.n.n.n] are now always recognized.
1489 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1490 used if the router is not running as root, for example, when verifying at
1491 ACL time, or when using -bh. The debugging output from this situation was
1492 non-existent - all you got was a failure to exec. I have made two
1495 (a) Failures to set uid/gid, the current directory, or a process leader
1496 in a subprocess such as that created by queryprogram now generate
1497 suitable debugging ouput when -d is set.
1499 (b) The queryprogram router detects when it is not running as root,
1500 outputs suitable debugging information if -d is set, and then runs
1501 the subprocess without attempting to change uid/gid.
1503 PH/15 Minor change to Makefile for building test_host (undocumented testing
1506 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1507 additional section of a DNS packet that returns MX or SRV records.
1508 Instead, it always explicitly searches for A/AAAA records. This avoids
1509 major problems that occur when a DNS server includes only records of one
1510 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1511 fixed another bug: if SRV records were looked up and the corresponding
1512 address records were *not* found in the additional section, the port
1513 values from the SRV records were lost.
1515 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1516 using the correct key (the original address) when searching the retry
1517 rules in order to find which one to use for generating the retry hint.
1519 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1520 adding the default one. Similarly, if it contains a Reply-To: header, the
1521 errors_reply_to option, if set, is not used.
1523 PH/19 When calculating a retry time, Exim used to measure the "time since
1524 failure" by looking at the "first failed" field in the retry record. Now
1525 it does not use this if it is later than than the arrival time of the
1526 message. Instead it uses the arrival time. This makes for better
1527 behaviour in cases where some deliveries succeed, thus re-setting the
1528 "first failed" field. An example is a quota failure for a huge message
1529 when small messages continue to be delivered. Without this change, the
1530 "time since failure" will always be short, possible causing more frequent
1531 delivery attempts for the huge message than are intended.
1532 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1534 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1535 $1, $2, $3) because the numerical variables can be reset during some
1536 expansion items (e.g. "match"), thereby losing the authentication data.
1538 PH/21 Make -bV show the size of off_t variables so that the test suite can
1539 decide whether to run tests for quotas > 2G.
1541 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1542 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1543 filecount value is greater than 2G or if a quota value is greater than 2G
1544 on a system where the size of off_t is not greater than 4, a panic error
1547 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1548 never match. The debug and -bh output now contains an explicit error
1549 message indicating a malformed IPv4 address or mask.
1551 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1552 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1553 PH/23 above applies.
1555 PH/25 Do not write to syslog when running in the test harness. The only
1556 occasion when this arises is a failure to open the main or panic logs
1557 (for which there is an explicit test).
1559 PH/26 Added the /no_tell option to "control=freeze".
1561 PH/27 If a host name lookup failed very early in a connection, for example, if
1562 the IP address matched host_lookup and the reverse lookup yielded a name
1563 that did not have a forward lookup, an error message of the form "no IP
1564 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1565 could be logged. Now it outputs the IP address instead of "NULL".
1567 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1568 allows the sender and the authenticated sender to be set when
1569 submitting a message from within Exim. Since child_open_exim() is
1570 documented for local_scan(), the new function should be too.
1572 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1573 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1574 results in an empty string is now treated as unset.
1576 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1578 PH/31 Added sender_verify_fail logging option.
1580 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1581 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1582 tidied the source and removed it altogether.
1584 PH/33 When a queue run was abandoned because the load average was too high, a
1585 log line was always written; now it is written only if the queue_run log
1586 selector is set. In addition, the log line for abandonment now contains
1587 information about the queue run such as the pid. This is always present
1588 in "start" and "stop" lines but was omitted from the "abandon" line.
1590 PH/34 Omit spaces between a header name and the colon in the error message that
1591 is given when verify = headers_syntax fails (if there are lots of them,
1592 the message gets confusing).
1594 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1595 names, as there are now some PTR records that contain slashes. This check
1596 is only to protect against broken name servers that fall over on strange
1597 characters, so the fact that it applies to all lookups doesn't matter.
1599 PH/36 Now that the new test suite is complete, we can remove some of the
1600 special code in Exim that was needed for the old test suite. For example,
1601 sorting DNS records because real resolvers return them in an arbitrary
1602 order. The new test suite's fake resolver always returns records in the
1605 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1606 bounces) except when queue_only is set, to avoid logging races between
1607 the different processes.
1609 PH/38 Panic-die if .include specifies a non-absolute path.
1611 PH/39 A tweak to the "H" retry rule from its user.
1613 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1614 a label. They prevented compilation on older perls.
1616 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1617 a warning to be raised on newish perls.
1619 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1620 on queue. Changes to match documented behaviour of showing count of
1621 messages matching specified criteria.
1623 PH/40 Changed the default ident timeout from 30s to 5s.
1625 PH/41 Added support for the use of login_cap features, on those BSD systems
1626 that have them, for controlling the resources used by pipe deliveries.
1628 PH/42 The content-scanning code uses fopen() to create files in which to put
1629 message data. Previously it was not paying any attention to the mode of
1630 the files. Exim runs with umask(0) because the rest of the code creates
1631 files with open(), and sets the required mode explicitly. Thus, these
1632 files were ending up world-writeable. This was not a big issue, because,
1633 being within the spool directory, they were not world-accessible. I have
1634 created a function called modefopen, which takes an additional mode
1635 argument. It sets umask(777), creates the file, chmods it to the required
1636 mode, then resets the umask. All the relevant calls to fopen() in the
1637 content scanning code have been changed to use this function.
1639 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1640 to 24 hours. This avoids potential overflow problems when processing G
1641 and H retry rules. I suspect nobody ever tinkers with this value.
1643 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1645 PH/45 When the plaintext authenticator is running as a client, the server's
1646 challenges are checked to ensure they are valid base64 strings. By
1647 default, the authentication attempt is cancelled if an invalid string is
1648 received. Setting client_ignore_invalid_base64 true ignores these errors.
1649 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1650 they are received. Thus, the responses can be made to depend on the
1651 challenges. If an invalid string is ignored, an empty string is placed in
1654 PH/46 Messages that are created by the autoreply transport now contains a
1655 References: header, in accordance with RFCs 2822 and 3834.
1657 PH/47 Added authenticated_sender_force to the smtp transport.
1659 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1661 PH/49 Installed latest patch from the Sieve maintainer.
1663 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1664 was also set, the appendfile transport was unnecessarily scanning a
1665 directory of message files (e.g. for maildir delivery) to find the count
1666 of files (along with the size), even though it did not need this
1667 information. It now does the scan only if it needs to find either the
1668 size of the count of files.
1670 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1672 PH/52 Two bugs concerned with error handling when the smtp transport is
1675 (i) Exim was not creating retry information for temporary errors given
1676 for individual recipients after the DATA command when the smtp transport
1677 was used in LMTP mode. This meant that they could be retried too
1678 frequently, and not timed out correctly.
1680 (ii) Exim was setting the flag that allows error details to be returned
1681 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1682 recipients that were returned after the DATA command.
1684 PH/53 This is related to PH/52, but is more general: for any failing address,
1685 when detailed error information was permitted to be returned to the
1686 sender, but the error was temporary, then after the final timeout, only
1687 "retry timeout exceeded" was returned. Now it returns the full error as
1688 well as "retry timeout exceeded".
1690 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1691 do this, and (what is worse) MTAs that accept it.
1693 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1694 will now be deprecated.
1696 PH/56 New os.c-cygwin from the Cygwin maintainer.
1698 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1699 formats (previously only available in exim formats via -bpr, -bpru,
1700 and -bpra. Now also available in native and exiqgrep formats)
1702 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1703 with very large, slow to parse queues
1705 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1707 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1709 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1710 responses to authentication challenges, though it was showing the
1711 challenges; (ii) I've removed the CR characters from the debug output for
1714 PH/58 Allow for the insertion of a newline as well as a space when a string
1715 is turned into more than one encoded-word during RFC 2047 encoding. The
1716 Sieve code now uses this.
1718 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1719 data_4xx, lost_connection, tls_required.
1721 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1722 message was being sent as an SMTP response.
1724 PH/61 Add -l and -k options to exicyclog.
1726 PH/62 When verifying, if an address was redirected to one new address, so that
1727 verification continued, and the new address failed or deferred after
1728 having set something in $address_data, the value of $address_data was not
1729 passed back to the ACL. This was different to the case when no
1730 redirection occurred. The value is now passed back in both cases.
1732 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1733 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1734 use login_cap.h, so on its own it isn't the distinguishing feature. The
1735 new name refers directly to the setclassresources() function.
1737 PH/65 Added configuration files for NetBSD3.
1739 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1741 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1742 is preferred over IPv4.
1744 PH/68 The bounce_return_message and bounce_return_body options were not being
1745 honoured for bounces generated during the reception of non-SMTP messages.
1746 In particular, this applied to messages rejected by the ACL. This bug has
1747 been fixed. However, if bounce_return_message is true and bounce_return_
1748 body is false, the headers that are returned for a non-SMTP message
1749 include only those that have been read before the error was detected.
1750 (In the case of an ACL rejection, they have all been read.)
1752 PH/69 The HTML version of the specification is now built in a directory called
1753 spec_html instead of spec.html, because the latter looks like a path with
1754 a MIME-type, and this confuses some software.
1756 PH/70 Catch two compiler warnings in sieve.c.
1758 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1759 function verify_get_ident() calls ip_connect() to connect a socket, but
1760 if the "connect()" function timed out, ip_connect() used to close the
1761 socket. However, verify_get_ident() also closes the socket later, and in
1762 between Exim writes to the log, which may get opened at this point. When
1763 the socket was closed in ip_connect(), the log could get the same file
1764 descriptor number as the socket. This naturally causes chaos. The fix is
1765 not to close the socket in ip_connect(); the socket should be closed by
1766 the function that creates it. There was only one place in the code where
1767 this was missing, in the iplookup router, which I don't think anybody now
1768 uses, but I've fixed it anyway.
1770 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1771 well as to direct DNS lookups. Otherwise the handling of names in host
1772 lists is inconsistent and therefore confusing.
1778 PH/01 Two changes to the default runtime configuration:
1780 (1) Move the checks for relay_from_hosts and authenticated clients from
1781 after to before the (commented out) DNS black list checks.
1783 (2) Add control=submission to the relay_from_hosts and authenticated
1784 clients checks, on the grounds that messages accepted by these
1785 statements are most likely to be submissions.
1787 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1789 (1) Generate an error if the third argument for the ${prvs expansion is
1792 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1795 (3) Reset the variables that are obtained from the first argument of
1796 ${prvscheck and used in the second argument before leaving the code,
1797 because their memory is reclaimed, so using them afterwards may do
1800 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1801 one (it's much easier than Tom thought :-).
1803 (5) Because of (4), we can now allow for the use of $prvscheck_result
1804 inside the third argument.
1806 PH/03 For some reason, the default setting of PATH when running a command from
1807 a pipe transport was just "/usr/bin". I have changed it to
1810 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1811 anything to be listed in the output from -bV.
1813 PH/05 When a filter generated an autoreply, the entire To: header line was
1814 quoted in the delivery log line, like this:
1816 => >A.N.Other <ano@some.domain> <original@ddress> ...
1818 This has been changed so that it extracts the operative address. There
1819 may be more than one such address. If so, they are comma-separated, like
1822 => >ano@some.domain,ona@other.domain <original@ddress> ...
1824 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1825 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1826 not being looked up in the rDNS to get a host name, Exim was showing the
1827 IP address twice in Received: lines, even though the IP addresses were
1828 identical. For example:
1830 Received: from [1.2.3.4] (helo=[1.2.3.4])
1832 However, if the real host name was known, it was omitting the HELO data
1833 if it matched the actual IP address. This has been tidied up so that it
1834 doesn't show the same IP address twice.
1836 PH/07 When both +timestamp and +memory debugging was on, the value given by
1837 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1838 called by the debug printing, thereby overwriting the timestamp buffer.
1839 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1841 PH/08 When the original message was included in an autoreply transport, it
1842 always said "this is a copy of the message, including all the headers",
1843 even if body_only or headers_only was set. It now gives an appropriate
1846 PH/09 Applied a patch from the Sieve maintainer which:
1848 o fixes some comments
1849 o adds the (disabled) notify extension core
1850 o adds some debug output for the result of if/elsif tests
1851 o points to the current vacation draft in the documentation
1852 and documents the missing references header update
1856 o fixes a bug in processing the envelope test (when testing
1857 multiple envelope elements, the last element determinted the
1860 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1861 Electronic Mail") by including:
1863 Auto-submitted: auto-generated
1865 in the messages that it generates (bounce messages and others, such as
1866 warnings). In the case of bounce messages for non-SMTP mesages, there was
1867 also a typo: it was using "Auto_submitted" (underscore instead of
1868 hyphen). Since every message generated by Exim is necessarily in response
1869 to another message, thes have all been changed to:
1871 Auto-Submitted: auto-replied
1873 in accordance with these statements in the RFC:
1875 The auto-replied keyword:
1877 - SHOULD be used on messages sent in direct response to another
1878 message by an automatic process,
1880 - MUST NOT be used on manually-generated messages,
1882 - MAY be used on Delivery Status Notifications (DSNs) and Message
1883 Disposition Notifications (MDNs),
1885 - MUST NOT be used on messages generated by automatic or periodic
1886 processes, except for messages which are automatic responses to
1889 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1890 to the default Received: header definition.
1892 PH/12 Added log selector acl_warn_skipped (default on).
1894 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1895 variables because (a) they are in the wrong storage pool and (b) even if
1896 they were copied, it wouldn't work properly because of the caching.
1898 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1899 checking when decoding. Apparently there are clients that generate
1900 overlong encoded strings. Why am I not surprised?
1902 PH/15 If the first argument of "${if match_address" was not empty, but did not
1903 contain an "@" character, Exim crashed. Now it writes a panic log message
1904 and treats the condition as false.
1906 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1908 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1909 extension "envelope-auth". The code is finished and in agreement with
1910 other implementations, but there is no documentation so far and in fact,
1911 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1912 not changing the active code.
1914 Print executed "if" and "elsif" statements when debugging is used. This
1915 helps a great deal to understand what a filter does.
1917 Document more things not specified clearly in RFC3028. I had all this
1918 sorted out, when out of a sudden new issues came to my mind. Oops."
1920 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1923 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1924 canonical form (as documented). However, after a host name lookup from
1925 the IP address, check_host() was doing a simple string comparison with
1926 addresses acquired from the DNS when checking that the found name did
1927 have the original IP as one of its addresses. Since any found IPv6
1928 addresses are likely to be in abbreviated form, the comparison could
1929 fail. Luckily, there already exists a function for doing the comparison
1930 by converting both addresses to binary, so now that is used instead of
1931 the text comparison.
1933 PH/20 There was another similar case to PH/19, when a complete host name was
1934 given in a host list; looking up its IP address could give an abbreviated
1935 form, whereas the current host's name might or might not be abbreviated.
1936 The same fix has been applied.
1942 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1943 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1946 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1947 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1949 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1951 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1952 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1953 Also, for "Auto-Submitted", treat anything other than "no" as
1954 non-personal, in accordance with RFC 3834. (Previously it treated
1955 anything starting "auto-" as non-personal.)
1957 TF/01 The control=submission/name=... option had a problem with syntax
1958 errors if the name included a slash character. The /name= option
1959 now slurps the rest of the string, so it can include any characters
1960 but it must come last in the list of options (after /sender_retain
1963 PH/05 Some modifications to the interface to the fake nameserver for the new
1971 TK/01 Added the "success_on_redirect" address verification option. See
1972 NewStuff for rationale and an example.
1974 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1976 PH/02 Patch to exigrep to allow it to work on syslog lines.
1978 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1979 fread() to skip over the body file's header line, because in Cygwin the
1980 header line is locked and is inaccessible.
1982 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1983 co-exist for some time) to make it clear that it is the Exim ID that is
1984 referenced, not the Message-ID: header line.
1986 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1987 string_format() function, because snprintf() does not exist on all
1990 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1991 ${stat: expansion item.
1993 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1994 protocol synchronization error", to keep the pedants happy.
1996 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1997 well as for IRIX systems, when gcc is being used. See the host.c source
2000 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2002 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2005 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2006 transport and to the smtp transport in LMTP mode.
2008 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2010 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2011 run a filter in a subprocess. This could lead to confusion in subsequent
2012 lookups in the parent process. There should also be a search_tidyup() at
2013 the end of the subprocess.
2015 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2016 only if the host matched helo_try_verify_hosts, which caused the
2017 verification to occur when the EHLO/HELO command was issued. The ACL just
2018 tested the remembered result. Now, if a previous verification attempt has
2019 not happened, "verify = helo" does it there and then.
2021 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2023 TK/03 Fix log output including CR from clamd.
2025 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2026 exist provoked a memory error which could cause a segfault.
2028 PH/15 Installed PCRE 6.2
2030 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2032 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2033 of the problem. Specifically, suggested +O2 rather than +O1 for the
2036 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2038 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2039 record, the port was not correctly logged when the outgoing_port log
2040 selector was set (it logged the transort's default port).
2042 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2043 fallback_hosts, and "hosts" in the smtp transport.
2045 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2046 host errors such as "Connection refused".
2048 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2049 authentication with radiusclient 0.4.9:
2051 - Error returned from rc_read_config was caught wrongly
2052 - Username/password not passed on to radius server due to wrong length.
2054 The presumption is that some radiusclient API changes for 4.51/PH/17
2055 were not taken care of correctly. The code is still untested by me (my
2056 Linux distribution still has 0.3.2 of radiusclient), but it was
2057 contributed by a Radius user.
2059 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2060 expanding the "port" option of the smtp transport.
2062 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2063 while reading a MIME header. Thanks to Tom Hughes for a patch.
2065 PH/24 Include config.h inside local_scan.h so that configuration settings are
2068 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2069 that in an ACL for RCPT (for example), you can examine exactly what was
2072 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2073 commands, but it was not correctly comparing the address with the actual
2074 client host address. Thus, it would show the EHLO address in Received:
2075 header lines when this was not necessary.
2077 PH/27 Added the % operator to ${eval:}.
2079 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2080 it should be ignoring failures (because with -C, for example, it has lost
2081 privilege). It wasn't ignoring creation failures other than "already
2084 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2087 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2088 because an input line was too long, either on its own, or by virtue of
2089 too many continuations, the temporary file was not being removed, and the
2090 return code was incorrect.
2092 PH/31 Missing "BOOL" in function definition in filtertest.c.
2094 PH/32 Applied Sieve patches from the maintainer.
2096 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2098 PH/33 Added "verify = not_blind".
2100 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2101 Local/Makefile (with some defaults set). These are used in built scripts
2102 such as exicyclog, but they have never been used in the exim_install
2103 script (though there are many overriding facilities there). I have
2104 arranged that the exim_install script now takes note of these two
2107 PH/35 Installed configuration files for Dragonfly.
2109 PH/36 When a locally submitted message by a trusted user did not contain a
2110 From: header, and the sender address was obtained from -f or from an SMTP
2111 MAIL command, and the trusted user did not use -F to supply a sender
2112 name, $originator_name was incorrectly used when constructing a From:
2113 header. Furthermore, $originator_name was used for submission mode
2114 messages from external hosts without From: headers in a similar way,
2115 which is clearly wrong.
2117 PH/37 Added control=suppress_local_fixups.
2119 PH/38 When log_selector = +received_sender was set, and the addition of the
2120 sender made the log line's construction buffer exactly full, or one byte
2121 less than full, an overflow happened when the terminating "\n" was
2124 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2125 when the result of a list match is failure because a DNS lookup failed.
2127 PH/40 RM_COMMAND is now used in the building process.
2129 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2130 the "build-* directories that it finds.
2132 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2133 address in a domain literal was a prefix of an interface address.
2135 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2136 when verifying a sender address, unless rewrite_headers is false.
2138 PH/44 Wrote a long comment about why errors_to addresses are verified as
2139 recipients, not senders.
2141 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2142 the ratelimit ACL was added.
2144 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2146 PH/46 Added extra information about PostgreSQL errors to the error string.
2148 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2149 avoiding the need to install special zones in a real server. This is
2150 backwards compatible; if it can't find the fake resolver, it drops back.
2151 Thus, both old and new test suites can be run.
2153 TF/02 Added util/ratelimit.pl
2155 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2156 clock is set back in time.
2158 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2159 Candler <B.Candler@pobox.com>.
2161 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2162 OK for addresses that are the result of redirections.
2164 PH/48 A number of further additions for the benefit of the new test suite,
2165 including a fake gethostbyname() that interfaces to the fake DNS resolver
2168 TF/06 The fix for widen_domains has also been applied to qualify_single and
2169 search_parents which are the other dnslookup options that can cause
2172 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2175 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2177 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2178 service" if it encountered an MX record with an empty target hostname.
2179 The message is now "an MX or SRV record indicated no SMTP service".
2181 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2182 if the DNS of the sending site is misconfigured. This is quite a
2183 common situation. This change restores the behaviour of treating a
2184 helo verification defer as a failure.
2186 PH/51 If self=fail was set on a router, the bounce message did not include the
2187 actual error message.
2193 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2195 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2196 command itself ended in error, the underlying message about the transport
2197 filter timeout was being overwritten with the pipe command error. Now the
2198 underlying error message should be appended to the second error message.
2200 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2202 PH/02 Reduce the amount of output that "make" produces by default. Full output
2203 can still be requested.
2205 PH/03 The warning log line about a condition test deferring for a "warn" verb
2206 was being output only once per connection, rather than after each
2207 occurrence (because it was using the same function as for successful
2208 "warn" verbs). This seems wrong, so I have changed it.
2210 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2211 it should not have, which might have caused a crash in the right
2212 circumstances, but probably never did.
2214 PH/04 Installed a modified version of Tony Finch's patch to make submission
2215 mode fix the return path as well as the Sender: header line, and to
2216 add a /name= option so that you can make the user's friendly name appear
2219 TF/03 Added the control = fakedefer ACL modifier.
2221 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2222 Mark Lowes for thorough testing.
2224 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2226 TK/03 Merged latest SRS patch from Miles Wilton.
2228 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2229 with the definition in sysexits.h (which is #included earlier).
2230 Fortunately, Exim does not actually use EX_OK. The code used to try to
2231 preserve the sysexits.h value, by assumimg that macro definitions were
2232 scanned for macro replacements. I have been disabused of this notion,
2233 so now the code just undefines EX_OK before #including unistd.h.
2235 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2236 in the smtp transport. When a block could not be written in a single
2237 write() function, the timeout was being re-applied to each part-write.
2238 This seems wrong - if the receiver was accepting one byte at a time it
2239 would take for ever. The timeout is now adjusted when this happens. It
2240 doesn't have to be particularly precise.
2242 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2243 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2245 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2246 without a domain if the check to <postmaster@domain> fails.
2248 SC/01 Eximstats: added -xls and the ability to specify output files
2249 (patch written by Frank Heydlauf).
2251 SC/02 Eximstats: use FileHandles for outputing results.
2253 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2255 SC/04 Eximstats: fixed display of large numbers with -nvr option
2257 SC/05 Eximstats: fixed merging of reports with empty tables.
2259 SC/06 Eximstats: added the -include_original_destination flag
2261 SC/07 Eximstats: removed tabs and trailing whitespace.
2263 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2265 TK/06 MBOX spool code: Add real "From " MBOX separator line
2266 so the .eml file is really in mbox format (even though
2267 most programs do not really care). Patch from Alex Miller.
2269 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2270 The latter is generated from $received_to and is only set if the
2271 message has one envelope recipient. SA can use these headers,
2272 obviously out-of-the-box. Patch from Alex Miller.
2274 PH/08 The ${def test on a variable was returning false if the variable's
2275 value was "0", contrary to what the specification has always said!
2276 The result should be true unless the variable is empty.
2278 PH/09 The syntax error of a character other than { following "${if
2279 def:variable_name" (after optional whitespace) was not being diagnosed.
2280 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2281 accidental colon was present, for example, could give incorrect results.
2283 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2284 result is used (not including appendfile, where other changes are about
2287 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2288 This involved changing a lot of size variables from int to off_t. It
2289 should work with maildirs and everything.
2291 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2292 spamd dying while we are connected to it.
2294 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2297 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2298 function for BDB 4.3.
2300 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2302 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2303 the sources for PCRE have been re-arranged and more files are now
2306 PH/15 The code I had for printing potentially long long variables in PH/11
2307 above was not the best (it lost precision). The length of off_t variables
2308 is now inspected at build time, and an appropriate printing format (%ld
2309 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2310 to be "long long int" or "long int". This is needed for the internal
2311 formatting function string_vformat().
2313 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2314 the configuration file to be ":syslog", then the script "guesses" where
2315 the logs files are, rather than using the compiled in default. In our
2316 case the guess is not the same as the compiled default, so the script
2317 suddenly stopped working when I started to use syslog. The patch checks
2318 to see if log_file_path is "". If so, it attempts to read it from exim
2319 with no configuration file to get the compiled in version, before it
2320 falls back to the previous guessing code."
2322 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2323 implementing BATV in an Exim configuration. See NewStuff for the gory
2326 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2327 Makefile that are specific to HP-UX.
2329 PH/18 If the "use_postmaster" option was set for a recipient callout together
2330 with the "random" option, the postmaster address was used as the MAIL
2331 FROM address for the random test, but not for the subsequent recipient
2332 test. It is now used for both.
2334 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2335 patch removes a few documentation additions to RFC 3028, because the
2336 latest draft now contains them. It adds the new en;ascii-case comparator
2337 and a new error check for 8bit text in MIME parts. Comparator and
2338 require names are now matched exactly. I enabled the subaddress
2339 extension, but it is not well tested yet (read: it works for me)."
2341 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2342 rework some of the code of TK/09 above to avoid the hardwired use of
2343 "%lld" and "long long". Replaced the call to snprintf() with a call to
2346 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2347 records point to non-existent hosts", "retry timeout exceeded", and
2348 "retry time not reached for any host after a long failure period".
2350 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2351 experimental DomainKeys support:
2353 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2354 (2) On an error such as an illegally used "control", the wrong name for
2355 the control was given.
2357 These problems did NOT occur unless DomainKeys support was compiled.
2359 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2361 PH/24 Added ${if match_ip condition.
2363 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2364 databases so that it will be absolutely obvious if a crash occurs in the
2365 DB library. This is a regular occurrence (often caused by mis-matched
2368 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2369 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2370 picked up on a user's system that detects such things. There doesn't seem
2371 to be a gcc warning option for this - only an attribute that has to be
2372 put on the function's prototype. It seems that in Fedora Core 4 they have
2373 set this on a number of new functions. No doubt there will be more in due
2376 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2377 specify a transport. However, if an address that was verified by such a
2378 router was the subject of a callout, Exim crashed because it tried to
2379 read the rcpt_include_affixes from the non-existent transport. Now it
2380 just assumes that the setting of that option is false. This bug was
2381 introduced by 4.51/PH/31.
2383 PH/28 Changed -d+all to exclude +memory, because that information is very
2384 rarely of interest, but it makes the output a lot bigger. People tend to
2385 do -d+all out of habit.
2387 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2388 code in os-type was giving problems when libc.so lives in lib64, like on
2391 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2392 aren't the modern standard, and it seems that some systems' include files
2393 don't always have them. Exim was already checking for some of the newer
2394 ones like T_AAAA, and defining it itself. I've added checks for all the
2395 record types that Exim uses.
2397 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2398 not automatically generating a new one, as it is supposed to. This
2399 prevented TLS from working. If the file did exist, but contained invalid
2400 data, a new version was generated, as expected. It was only the case of a
2401 non-existent file that was broken.
2403 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2404 with a change in libdomainkeys > 0.64.
2406 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2407 from DNS. If the selector record carries the flag, it now has
2408 precedence over the domain-wide flag.
2410 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2412 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2413 the use of a port name that isn't defined in /etc/services) occurred, the
2414 message was deferred as in a normal delivery, and thus remained on the
2415 spool, instead of being failed because of the mua_wrapper setting. This
2416 is now fixed, and I tidied up some of the mua_wrapper messages at the
2419 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2420 the messages in a hash of arrays rather than using individual hashes.
2421 This is a bit cleaner and results in dramatic memory savings, albeit
2422 at a slight CPU cost.
2424 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2425 as requested by Marc Sherman.
2427 SC/10 Eximstats: added histograms for user specified patterns as requested
2430 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2432 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2433 fopen() in the content-scanning modules that did not already have it.
2439 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2440 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2442 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2444 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2446 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2447 to test Sieve filters that use "vacation".
2449 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2450 that changes the way the GnuTLS parameters are stored in the cache file.
2451 The new format can be generated externally. For backward compatibility,
2452 if the data in the cache doesn't make sense, Exim assumes it has read an
2453 old-format file, and it generates new data and writes a new file. This
2454 means that you can't go back to an older release without removing the
2457 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2458 work if there are any delivery delays because "one_time" forces the
2459 parent to be marked "delivered", so its unseen clone is never tried
2460 again. For this reason, Exim now forbids the simultaneous setting of
2463 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2464 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2465 entry if you want to know the details. The fix, however, neglected to
2466 consider the case when local delivery batching is involved. The test for
2467 "previously delivered" was not happening when checking to see if an
2468 address could be batched with a previous (undelivered) one; under
2469 certain circumstances this could lead to multiple deliveries to the same
2472 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2473 in its include files, and this causes problems building Exim.
2475 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2476 header_syntax) but Exim was just ignoring anything given after a slash.
2477 In particular, this caused confusion with an attempt to use "verify =
2478 reverse_host_lookup/defer_ok". An error is now given when options are
2479 supplied for verify items that do not have them. (Maybe reverse_host_
2480 lookup should have a defer_ok option, but that's a different point.)
2482 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2483 defined by RFC 821) to 2048, because there were problems with some AUTH
2484 commands, and RFC 1869 says the size should be increased for extended
2485 SMTP commands that take arguments.
2487 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2490 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2491 "unknown" error; now it says that the functionality isn't in the binary.
2493 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2494 an address' error message when a string expansion fails (syntax or
2495 whatever). Otherwise the password may appear in the log. Following change
2496 PH/42 below, there is no longer a chance of it appearing in a bounce
2499 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2501 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2502 of its addresses was ever tried. (Bugzilla bug #2.)
2504 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2505 the result incorrectly in the debug output. (It correctly added a newline
2506 to what was transported.)
2508 TF/01 Added $received_time.
2510 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2511 commented out examples of how to interface to a virus scanner and to
2512 SpamAssassin. Also added commented examples of av_scanner and
2513 spamd_address settings.
2515 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2516 and controls are allowed in which ACLs. There were a couple of minor
2517 errors. Some of the entries in the conditions table (which is a table of
2518 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2519 negation of where the condition IS allowed.
2521 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2523 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2524 header file does not have a version number, so I've had to invent a new
2525 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2526 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2527 radiusclient), but it was contributed by a Radius user.
2529 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2530 files or directories, for OpenSSL.
2532 PH/19 When an Exim process that is running as root has to create an Exim log
2533 file, it does so in a subprocess that runs as exim:exim so as to get the
2534 ownership right at creation (otherwise, other Exim processes might see
2535 the file with the wrong ownership). There was no test for failure of this
2536 fork() call, which would lead to the process getting stuck as it waited
2537 for a non-existent subprocess. Forks do occasionally fail when resources
2538 run out. I reviewed all the other calls to fork(); they all seem to check
2541 PH/20 When checking for unexpected SMTP input at connect time (before writing
2542 the banner), Exim was not dealing correctly with a non-positive return
2543 from the read() function. If the client had disconnected by this time,
2544 the result was a log entry for a synchronization error with an empty
2545 string after "input=" when read() returned zero. If read() returned -1
2546 (an event I could not check), uninitialized data bytes were printed.
2547 There were reports of junk text (parts of files, etc) appearing after
2550 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2552 PH/22 Added support for macro redefinition, and (re)definition in between
2553 driver and ACL definitions.
2555 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2556 forgetting to use the resulting value; it was using the unexpanded value.
2558 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2559 hadn't been configured. The fix is from Juergen Kreileder, who
2560 understands it better than I do:
2562 "Here's what I see happening with three configured cyrus_sasl
2563 authenticators configured (plain, login, cram-md5):
2565 On startup auth_cyrus_sasl_init() gets called for each of these.
2566 This means three calls to sasl_listmech() without a specified mech_list.
2567 => SASL tests which mechs of all available mechs actually work
2568 => three warnings about OTP not working
2569 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2571 With the patch, sasl_listmech() also gets called three times. But now
2572 SASL's mech_list option is set to the server_mech specified in the the
2573 authenticator. Or in other words, the answer from sasl_listmech()
2574 gets limited to just the mech you're testing for (which is different
2576 => the return list contains just 'plain' or 'login', 'cram-md5' or
2577 nothing depending on the value of ob->server_mech.
2579 I've just tested the patch: Authentication still works fine,
2580 unavailable mechs specified in the exim configuration are still
2581 caught, and the auth.log warnings about OTP are gone."
2583 PH/25 When debugging is enabled, the contents of the command line are added
2584 to the debugging output, even when log_selector=+arguments is not
2587 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2588 answer is "GNU", and only if the return is "GNU/something" is the answer
2591 PH/27 $acl_verify_message is now set immediately after the failure of a
2592 verification in an ACL, and so is available in subsequent modifiers. In
2593 particular, the message can be preserved by coding like this:
2595 warn !verify = sender
2596 set acl_m0 = $acl_verify_message
2598 Previously, $acl_verify_message was set only while expanding "message"
2599 and "log_message" when a very denied access.
2601 PH/28 Modified OS/os.c-Linux with
2603 -#ifndef OS_LOAD_AVERAGE
2604 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2606 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2607 nomenclature these days.)
2609 PH/29 Installed patch from the Sieve maintainer that adds the options
2610 sieve_useraddress and sieve_subaddress to the redirect router.
2612 PH/30 In these circumstances:
2613 . Two addresses routed to the same list of hosts;
2614 . First host does not offer TLS;
2615 . First host accepts first address;
2616 . First host gives temporary error to second address;
2617 . Second host offers TLS and a TLS session is established;
2618 . Second host accepts second address.
2619 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2620 and peerdn, if requested) that were in fact used only for the second
2623 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2624 attention to any local part prefix or suffix that was matched by the
2625 router that accepted the address. It now behaves in the same way as it
2626 does for delivery: the affixes are removed from the local part unless
2627 rcpt_include_affixes is set on the transport.
2629 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2630 timeout during the DATA phase of an incoming message.
2632 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2633 applied a patch sanctioned by the Sieve maintainer.
2635 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2636 the uid or gid is negative. A case of a negative gid caused this to be
2637 noticed. The fix allows for either to be negative.
2639 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2640 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2643 PH/36 The change PH/12 above was broken. Fixed it.
2645 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2646 the grounds that routing the same address twice would always produce the
2647 same answer. This might have been true once, but it is certainly no
2648 longer true now. Routing a child address may depend on the previous
2649 routing that produced that child. Some complicated redirection strategies
2650 went wrong when messages had multiple recipients, and made Exim's
2651 behaviour dependent on the order in which the addresses were given.
2653 I have moved the duplicate checking until after the routing is complete.
2654 Exim scans the addresses that are assigned to local and remote
2655 transports, and removes any duplicates. This means that more work will be
2656 done, as duplicates will always all be routed, but duplicates are
2657 presumably rare, so I don't expect this is of any significance.
2659 For deliveries to pipes, files, and autoreplies, the duplicate checking
2660 still happens during the routing process, since they are not going to be
2663 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2664 It corrects a timeout issue with spamd. This is Ian's comment: "The
2665 background is that sometimes spamd either never reads data from a
2666 connection it has accepted, or it never writes response data. The exiscan
2667 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2668 blindly assumes that writes won't block so it may never time out."
2670 PH/39 Allow G after quota size as well as K and M.
2672 PH/40 The value set for $authenticated_id in an authenticator may not contain
2673 binary zeroes or newlines because the value is written to log lines and
2674 to spool files. There was no check on this. Now the value is run through
2675 the string_printing() function so that such characters are converted to
2676 printable escape sequences.
2678 PH/41 $message_linecount is a new variable that contains the total number of
2679 lines in the message. Compare $body_linecount, which is the count for the
2682 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2683 bounce and delay warning messages, except in certain special cases, which
2686 (a) An SMTP error message from a remote host;
2687 (b) A message specified in a :fail: redirection;
2688 (c) A message specified in a "fail" command in a system filter;
2689 (d) A message specified in a FAIL return from the queryprogram router;
2690 (e) A message specified by the cannot_route_message router option.
2692 In these cases only, Exim does include the error details in bounce and
2693 warning messages. There are also a few cases where bland messages such
2694 as "unrouteable address" or "local delivery error" are given.
2696 PH/43 $value is now also set for the "else" part of a ${run expansion.
2698 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2699 being worked on, but at least Exim now implements the latest version to
2702 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2703 process to complete was treated as a delivery failure, a timeout while
2704 writing the message to the pipe was logged, but erroneously treated as a
2705 successful delivery. Such timeouts include transport filter timeouts. For
2706 consistency with the overall process timeout, these timeouts are now
2707 treated as errors, giving rise to delivery failures by default. However,
2708 there is now a new Boolean option for the pipe transport called
2709 timeout_defer, which, if set TRUE, converts the failures into defers for
2710 both kinds of timeout. A transport filter timeout is now identified in
2713 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2714 systems where "make" and "gmake" are different, calling "gmake" at top
2715 level broke things. I've arranged for the value of $(MAKE) to be passed
2716 from the Makefile to this script so that it can call the same version of
2720 A note about Exim versions 4.44 and 4.50
2721 ----------------------------------------
2723 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2724 changes of various kinds. As a consequence, a big documentation update was
2725 needed. This delayed the release for rather longer than seemed good, especially
2726 in the light of a couple of (minor) security issues. Therefore, the changes
2727 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2728 release. So 4.44 and 4.50 are in effect two different branches that both start
2731 I have left the 4.50 change log unchanged; it contains all the changes since
2732 4.43. The change log for 4.44 is below; many of its items are identical to
2733 those for 4.50. This seems to be the most sensible way to preserve the
2734 historical information.
2740 1. Minor wording change to the doc/README.SIEVE file.
2742 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2743 computation of the current number of files was incorrect.
2745 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2746 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2749 4. Give more explanation in the error message when the command for a transport
2750 filter fails to execute.
2752 5. There are several places where Exim runs a non-Exim command in a
2753 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2754 was being done only for the command run by the queryprogram router. It is
2755 now done for all such subprocesses. The other cases are: ${run, transport
2756 filters, and the commands run by the lmtp and pipe transports.
2758 6. Added CONFIGURE_GROUP build-time option.
2760 7. Some older OS have a limit of 256 on the maximum number of file
2761 descriptors. Exim was using setrlimit() to set 1000 as a large value
2762 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2763 systems. I've change it so that if it can't get 1000, it tries for 256.
2765 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2766 was an oversight, and furthermore, ever since the addition of extra
2767 controls (e.g. 4.43/32), the checks on when to allow different forms of
2768 "control" were broken. There should now be diagnostics for all cases when a
2769 control that does not make sense is encountered.
2771 9. Added the /retain_sender option to "control=submission".
2773 10. $recipients is now available in the predata ACL (oversight).
2775 11. Tidy the search cache before the fork to do a delivery from a message
2776 received from the command line. Otherwise the child will trigger a lookup
2777 failure and thereby defer the delivery if it tries to use (for example) a
2778 cached ldap connection that the parent has called unbind on.
2780 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2781 of $address_data from the recipient verification was clobbered by the
2782 sender verification.
2784 13. The value of address_data from a sender verification is now available in
2785 $sender_address_data in subsequent conditions in the ACL statement.
2787 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2789 15. Added a new option "connect=<time>" to callout options, to set a different
2792 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2793 was its contents. (It was OK if the option was not defined at all.)
2795 17. A "Completed" log line is now written for messages that are removed from
2796 the spool by the -Mrm option.
2798 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2799 information about exactly what failed.
2801 19. Added -dd to debug only the daemon process.
2803 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2804 handles timeouts, both on the server side and network timeouts. Renamed the
2805 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2807 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2808 It is now set to "smtps".
2810 22. $host_address is now set to the target address during the checking of
2811 ignore_target_hosts.
2813 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2814 being passed; this would have caused $sender_host_name to have been used if
2815 matching the list had actually called for a host name (not very likely,
2816 since this list is usually IP addresses). A host name is now passed as
2819 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2820 code that specifies a non-restarting handler (typically sigaction() in
2821 modern systems) in an attempt to fix a rare and obscure crash bug.
2823 25. Narrowed the window for a race in the daemon that could cause it to ignore
2824 SIGCHLD signals. This is not a major problem, because they are used only to
2825 wake it up if nothing else does.
2827 26. A malformed maildirsize file could cause Exim to calculate negative values
2828 for the mailbox size or file count. Odd effects could occur as a result.
2829 The maildirsize information is now recalculated if the size or filecount
2832 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2833 support for a long time. Removed HAVE_SYS_VFS_H.
2835 28. Installed the latest version of exipick from John Jetmore.
2837 29. In an address list, if the pattern was not a regular expression, an empty
2838 subject address (from a bounce message) matched only if the pattern was an
2839 empty string. Non-empty patterns were not even tested. This was the wrong
2840 because it is perfectly reasonable to use an empty address as part of a
2841 database query. An empty address is now tested by patterns that are
2842 lookups. However, all the other forms of pattern expect the subject to
2843 contain a local part and a domain, and therefore, for them, an empty
2844 address still always fails if the pattern is not itself empty.
2846 30. Exim went into a mad DNS loop when attempting to do a callout where the
2847 host was specified on an smtp transport, and looking it up yielded more
2848 than one IP address.
2850 31. Re-factored the code for checking spool and log partition space into a
2851 function that finds that data and another that does the check. The former
2852 is then used to implement four new variables: $spool_space, $log_space,
2853 $spool_inodes, and $log_inodes.
2855 32. The RFC2047 encoding function was originally intended for short strings
2856 such as real names; it was not keeping to the 75-character limit for
2857 encoded words that the RFC imposes. It now respects the limit, and
2858 generates multiple encoded words if necessary. To be on the safe side, I
2859 have increased the buffer size for the ${rfc2047: expansion operator from
2862 33. It is now permitted to omit both strings after an "if" condition; if the
2863 condition is true, the result is "true". As before, when the second string
2864 is omitted, a false condition yields an empty string. This makes it less
2865 cumbersome to write custom ACL and router conditions.
2867 34. Failure to deliver a bounce message always caused it to be frozen, even if
2868 there was an errors_to setting on the router. The errors_to setting is now
2871 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2872 canonical form (fully expanded) before being placed in
2873 $sender_host_address.
2875 36. The table in the code that translates DNS record types into text (T_A to
2876 "A" for instance) was missing entries for NS and CNAME. It is just possible
2877 that this could have caused confusion if both these types were looked up
2878 for the same domain, because the text type is used as part of Exim's
2879 per-process caching. But the chance of anyone hitting this buglet seems
2882 37. The dnsdb lookup has been extended in a number of ways.
2884 (1) There is a new type, "zns", which walks up the domain tree until it
2885 finds some nameserver records. It should be used with care.
2887 (2) There is a new type, "mxh", which is like "mx" except that it returns
2888 just the host names, not the priorities.
2890 (3) It is now possible to give a list of domains (or IP addresses) to be
2891 looked up. The behaviour when one of the lookups defers can be
2892 controlled by a keyword.
2894 (4) It is now possible to specify the separator character for use when
2895 multiple records are returned.
2897 38. The dnslists ACL condition has been extended: it is now possible to supply
2898 a list of IP addresses and/or domains to be looked up in a particular DNS
2901 39. Added log_selector=+queue_time_overall.
2903 40. When running the queue in the test harness, wait just a tad after forking a
2904 delivery process, to get repeatability of debugging output.
2906 41. Include certificate and key file names in error message when GnuTLS fails
2907 to set them up, because the GnuTLS error message doesn't include the name
2908 of the failing file when there is a problem reading it.
2910 42. Allow both -bf and -bF in the same test run.
2912 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2914 44. The "Exiscan patch" is now merged into the mainline Exim source.
2916 45. Sometimes the final signoff response after QUIT could fail to get
2917 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2918 < 0 before doing a fflush(). This bug looks as though it goes back to the
2919 introduction of TLS in release 3.20, but "sometimes" must have been rare
2920 because the tests only now provoked it.
2922 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2923 (this can affect the format of dates).
2925 47. exim_tidydb, when checking for the continued existence of a message for
2926 which it has found a message-specific retry record, was not finding
2927 messages that were in split spool directories. Consequently, it was
2928 deleting retry records that should have stayed in existence.
2930 48. Steve fixed some bugs in eximstats.
2932 49. The SPA authentication driver was not abandoning authentication and moving
2933 on to the next authenticator when an expansion was forced to fail,
2934 contradicting the general specification for all authenticators. Instead it
2935 was generating a temporary error. It now behaves as specified.
2937 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2938 (the order specifies the preference for clients). The order is now AES256,
2939 AES128, 3DES, ARCFOUR128.
2941 51. Small patch to Sieve code - explicitly set From: when generating an
2944 52. Exim crashed if a remote delivery caused a very long error message to be
2945 recorded - for instance if somebody sent an entire SpamAssassin report back
2946 as a large number of 550 error lines. This bug was coincidentally fixed by
2947 increasing the size of one of Exim's internal buffers (big_buffer) that
2948 happened as part of the Exiscan merge. However, to be on the safe side, I
2949 have made the code more robust (and fixed the comments that describe what
2952 53. Now that there can be additional text after "Completed" in log lines (if
2953 the queue_time_overall log selector is set), a one-byte patch to exigrep
2954 was needed to allow it to recognize "Completed" as not the last thing in
2957 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2958 patch that reportedly fixes this has been added. I am not expert enough to
2959 create a test for it. This is what the patch creator wrote:
2961 "I found a little strange behaviour of ldap code when working with
2962 Windows 2003 AD Domain, where users was placed in more than one
2963 Organization Units. When I tried to give exim partial DN, the exit code
2964 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2965 But simultaneously result of request was absolutely normal ldap result,
2966 so I produce this patch..."
2968 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2969 REFERENCE, so I have modified the code to exclude the patch when that macro
2972 55. Some experimental protocols are using DNS PTR records for new purposes. The
2973 keys for these records are domain names, not reversed IP addresses. The
2974 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2975 leaves it alone. Component reversal etc. now happens only for IP addresses.
2978 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2980 57. Double the size of the debug message buffer (to 2048) so that more of very
2981 long debug lines gets shown.
2983 58. The exicyclog utility now does better if the number of log files to keep
2984 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2986 59. Two changes related to the smtp_active_hostname option:
2988 (1) $smtp_active_hostname is now available as a variable.
2989 (2) The default for smtp_banner uses $smtp_active_hostname instead
2990 of $primary_hostname.
2992 60. The host_aton() function is supposed to be passed a string that is known
2993 to be a valid IP address. However, in the case of IPv6 addresses, it was
2994 not checking this. This is a hostage to fortune. Exim now panics and dies
2995 if the condition is not met. A case was found where this could be provoked
2996 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2997 components; fortuitously, this particular loophole had already been fixed
2998 by change 4.50/55 above.
3000 If there are any other similar loopholes, the new check in host_aton()
3001 itself should stop them being exploited. The report I received stated that
3002 data on the command line could provoke the exploit when Exim was running as
3003 exim, but did not say which command line option was involved. All I could
3004 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3005 running as the user.
3008 61. There was a buffer overflow vulnerability in the SPA authentication code
3009 (which came originally from the Samba project). I have added a test to the
3010 spa_base64_to_bits() function which I hope fixes it.
3013 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3014 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3016 63. The daemon start-up calls getloadavg() while still root for those OS that
3017 need the first call to be done as root, but it missed one case: when
3018 deliver_queue_load_max is set with deliver_drop_privilege. This is
3019 necessary for the benefit of the queue runner, because there is no re-exec
3020 when deliver_drop_privilege is set.
3022 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3023 This has been fixed.
3025 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3026 was in use, was not putting the data itself into the right store pool;
3027 consequently, it could be overwritten for a subsequent message in the same
3028 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3031 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3033 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3034 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3035 the function were treating the return as a boolean value, which happened to
3036 work because 0=false and not-0=true, but is not correct code.
3038 68. The host_aton() function was not handling scoped IPv6 addresses (those
3039 with, for example, "%eth0" on the end) correctly.
3041 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3042 negated items (that is, ~something) in unsigned ints. Some compilers
3043 apparently mutter when there is no cast.
3045 70. If an address verification called from an ACL failed, and did not produce a
3046 user-specific message (i.e. there was only a "system" message), nothing was
3047 put in $acl_verify_message. In this situation, it now puts the system
3050 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3051 session; change 4.31/43 added the unwanted input to the log line - except
3052 that it did not do this in the start of session case. It now does.
3054 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3055 This is wrong and can cause the other end to generate a synchronization
3056 error if it is another Exim or anything else that does the synchronization
3057 check. A QUIT command is no longer sent after a timeout.
3059 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3060 during host lookups.
3062 74. The defer_ok option of callout verification was not working if it was used
3063 when verifying addresses in header lines, that is, for this case:
3065 verify = header_sender/callout=defer_ok
3067 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3068 those file descriptors could be used for SMTP connections. If anything
3069 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3070 could be sent to the SMTP client, causing chaos. The daemon now opens
3071 stdin, stdout, and stderr to /dev/null when it puts itself into the
3074 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3075 log by default. The user can override this with suitable Perl magic.
3077 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3078 the log message when discard triggers, was not working for the DATA ACL or
3079 for the non-SMTP ACL.
3081 78. Error message wording change in sieve.c.
3083 79. If smtp_accept_max_per_host was set, the number of connections could be
3084 restricted to fewer than expected, because the daemon was trying to set up
3085 a new connection before checking whether the processes handling previous
3086 connections had finished. The check for completed processes is now done
3087 earlier. On busy systems, this bug wouldn't be noticed because something
3088 else would have woken the daemon, and it would have reaped the completed
3091 80. If a message was submitted locally by a user whose login name contained one
3092 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3093 It caused a spool format error. I have fixed the spool reading code. A
3094 related problem was that the "from" clause in the Received: line became
3095 illegal because of the space(s). It is now covered by ${quote_local_part.
3097 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3100 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3102 83. Patch from Sieve maintainer to fix unterminated string problem in
3103 "vacation" handling.
3105 84. Some minor changes to the Linux configuration files to help with other
3106 OS variants using glibc.
3108 85. One more patch for Sieve to update vacation handling to latest spec.
3111 ----------------------------------------------------
3112 See the note above about the 4.44 and 4.50 releases.
3113 ----------------------------------------------------
3119 1. Change 4.43/35 introduced a bug that caused file counts to be
3120 incorrectly computed when quota_filecount was set in an appendfile
3123 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3124 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3127 3. Give more explanation in the error message when the command for a transport
3128 filter fails to execute.
3130 4. There are several places where Exim runs a non-Exim command in a
3131 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3132 was being done only for the command run by the queryprogram router. It is
3133 now done for all such subprocesses. The other cases are: ${run, transport
3134 filters, and the commands run by the lmtp and pipe transports.
3136 5. Some older OS have a limit of 256 on the maximum number of file
3137 descriptors. Exim was using setrlimit() to set 1000 as a large value
3138 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3139 systems. I've change it so that if it can't get 1000, it tries for 256.
3141 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3142 was an oversight, and furthermore, ever since the addition of extra
3143 controls (e.g. 4.43/32), the checks on when to allow different forms of
3144 "control" were broken. There should now be diagnostics for all cases when a
3145 control that does not make sense is encountered.
3147 7. $recipients is now available in the predata ACL (oversight).
3149 8. Tidy the search cache before the fork to do a delivery from a message
3150 received from the command line. Otherwise the child will trigger a lookup
3151 failure and thereby defer the delivery if it tries to use (for example) a
3152 cached ldap connection that the parent has called unbind on.
3154 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3155 of $address_data from the recipient verification was clobbered by the
3156 sender verification.
3158 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3159 was its contents. (It was OK if the option was not defined at all.)
3161 11. A "Completed" log line is now written for messages that are removed from
3162 the spool by the -Mrm option.
3164 12. $host_address is now set to the target address during the checking of
3165 ignore_target_hosts.
3167 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3168 being passed; this would have caused $sender_host_name to have been used if
3169 matching the list had actually called for a host name (not very likely,
3170 since this list is usually IP addresses). A host name is now passed as
3173 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3174 code that specifies a non-restarting handler (typically sigaction() in
3175 modern systems) in an attempt to fix a rare and obscure crash bug.
3177 15. Narrowed the window for a race in the daemon that could cause it to ignore
3178 SIGCHLD signals. This is not a major problem, because they are used only to
3179 wake it up if nothing else does.
3181 16. A malformed maildirsize file could cause Exim to calculate negative values
3182 for the mailbox size or file count. Odd effects could occur as a result.
3183 The maildirsize information is now recalculated if the size or filecount
3186 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3187 support for a long time. Removed HAVE_SYS_VFS_H.
3189 18. Updated exipick to current release from John Jetmore.
3191 19. Allow an empty sender to be matched against a lookup in an address list.
3192 Previously the only cases considered were a regular expression, or an
3195 20. Exim went into a mad DNS lookup loop when doing a callout where the
3196 host was specified on the transport, if the DNS lookup yielded more than
3199 21. The RFC2047 encoding function was originally intended for short strings
3200 such as real names; it was not keeping to the 75-character limit for
3201 encoded words that the RFC imposes. It now respects the limit, and
3202 generates multiple encoded words if necessary. To be on the safe side, I
3203 have increased the buffer size for the ${rfc2047: expansion operator from
3206 22. Failure to deliver a bounce message always caused it to be frozen, even if
3207 there was an errors_to setting on the router. The errors_to setting is now
3210 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3211 canonical form (fully expanded) before being placed in
3212 $sender_host_address.
3214 24. Updated eximstats to version 1.33
3216 25. Include certificate and key file names in error message when GnuTLS fails
3217 to set them up, because the GnuTLS error message doesn't include the name
3218 of the failing file when there is a problem reading it.
3220 26. Expand error message when OpenSSL has problems setting up cert/key files.
3223 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3224 (this can affect the format of dates).
3226 28. exim_tidydb, when checking for the continued existence of a message for
3227 which it has found a message-specific retry record, was not finding
3228 messages that were in split spool directories. Consequently, it was
3229 deleting retry records that should have stayed in existence.
3231 29. eximstats updated to version 1.35
3232 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3233 1.35 - bugfix such that pie charts by volume are generated correctly
3235 30. The SPA authentication driver was not abandoning authentication and moving
3236 on to the next authenticator when an expansion was forced to fail,
3237 contradicting the general specification for all authenticators. Instead it
3238 was generating a temporary error. It now behaves as specified.
3240 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3241 (the order specifies the preference for clients). The order is now AES256,
3242 AES128, 3DES, ARCFOUR128.
3244 31. Small patch to Sieve code - explicitly set From: when generating an
3247 32. Exim crashed if a remote delivery caused a very long error message to be
3248 recorded - for instance if somebody sent an entire SpamAssassin report back
3249 as a large number of 550 error lines. This bug was coincidentally fixed by
3250 increasing the size of one of Exim's internal buffers (big_buffer) that
3251 happened as part of the Exiscan merge. However, to be on the safe side, I
3252 have made the code more robust (and fixed the comments that describe what
3255 33. Some experimental protocols are using DNS PTR records for new purposes. The
3256 keys for these records are domain names, not reversed IP addresses. The
3257 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3258 leaves it alone. Component reversal etc. now happens only for IP addresses.
3261 34. The host_aton() function is supposed to be passed a string that is known
3262 to be a valid IP address. However, in the case of IPv6 addresses, it was
3263 not checking this. This is a hostage to fortune. Exim now panics and dies
3264 if the condition is not met. A case was found where this could be provoked
3265 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3266 components; fortuitously, this particular loophole had already been fixed
3267 by change 4.50/55 or 4.44/33 above.
3269 If there are any other similar loopholes, the new check in host_aton()
3270 itself should stop them being exploited. The report I received stated that
3271 data on the command line could provoke the exploit when Exim was running as
3272 exim, but did not say which command line option was involved. All I could
3273 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3274 running as the user.
3277 35. There was a buffer overflow vulnerability in the SPA authentication code
3278 (which came originally from the Samba project). I have added a test to the
3279 spa_base64_to_bits() function which I hope fixes it.
3282 36. The daemon start-up calls getloadavg() while still root for those OS that
3283 need the first call to be done as root, but it missed one case: when
3284 deliver_queue_load_max is set with deliver_drop_privilege. This is
3285 necessary for the benefit of the queue runner, because there is no re-exec
3286 when deliver_drop_privilege is set.
3288 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3289 was in use, was not putting the data itself into the right store pool;
3290 consequently, it could be overwritten for a subsequent message in the same
3291 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3294 38. Sometimes the final signoff response after QUIT could fail to get
3295 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3296 < 0 before doing a fflush(). This bug looks as though it goes back to the
3297 introduction of TLS in release 3.20, but "sometimes" must have been rare
3298 because the tests only now provoked it.
3304 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3305 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3306 Now it returns an integer. A number of calls were still expecting a T/F
3307 return. Fortuitously, in all cases, the tests worked in OK situations,
3308 which is the norm. However, things would have gone wrong on any write
3309 failures on the smtp file descriptor. This function is used when sending
3310 messages over SMTP and also when doing verify callouts.
3312 2. When Exim is called to do synchronous delivery of a locally submitted
3313 message (the -odf or -odi options), it no longer closes stderr before doing
3316 3. Implemented the mua_wrapper option.
3318 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3320 5. Implemented the functions header_remove(), header_testname(),
3321 header_add_at_position(), and receive_remove_recipient(), and exported them
3324 6. If an ACL "warn" statement specified the addition of headers, Exim already
3325 inserted X-ACL-Warn: at the start if there was no header name. However, it
3326 was not making this test for the second and subsequent header lines if
3327 there were newlines in the string. This meant that an invalid header could
3328 be inserted if Exim was badly configured.
3330 7. Allow an ACL "warn" statement to add header lines at the start or after all
3331 the Received: headers, as well as at the end.
3333 8. Added the rcpt_4xx retry error code.
3335 9. Added postmaster_mailfrom=xxx to callout verification option.
3337 10. Added mailfrom=xxxx to the callout verification option, for verify=
3340 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3341 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3343 12. Inserted some casts to stop certain compilers warning when using pointer
3344 differences as field lengths or precisions in printf-type calls (mostly
3345 affecting debugging statements).
3347 13. Added optional readline() support for -be (dynamically loaded).
3349 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3350 same clock tick as a message's arrival, so that its received time was the
3351 same as the "first fail" time on the retry record, and that message
3352 remained on the queue past the ultimate address timeout, every queue runner
3353 would try a delivery (because it was past the ultimate address timeout) but
3354 after another failure, the ultimate address timeout, which should have then
3355 bounced the address, did not kick in. This was a "< instead of <=" error;
3356 in most cases the first failure would have been in the next clock tick
3357 after the received time, and all would be well.
3359 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3360 being recognized when the domain list was tested by the match_domain
3361 condition in an expansion string.
3363 16. Added the ${str2b64: operator.
3365 17. Exim was always calling setrlimit() to set a large limit for the number of
3366 processes, without checking whether the existing limit was already
3367 adequate. (It did check for the limit on file descriptors.) Furthermore,
3368 errors from getrlimit() and setrlimit() were being ignored. Now they are
3369 logged to the main and panic logs, but Exim does carry on, to try to do its
3370 job under whatever limits there are.
3372 18. Imported PCRE 5.0.
3374 19. Trivial typo in log message " temporarily refused connection" (the leading
3377 20. If the log selector return_path_on_delivery was set and an address was
3378 redirected to /dev/null, the delivery process crashed because it assumed
3379 that a return path would always be set for a "successful" delivery. In this
3380 case, the whole delivery is bypassed as an optimization, and therefore no
3383 21. Internal re-arrangement: the function for sending a challenge and reading
3384 a response while authentication was assuming a zero-terminated challenge
3385 string. It's now changed to take a pointer and a length, to allow for
3386 binary data in such strings.
3388 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3390 23. Exim was not respecting finduser_retries when seeking the login of the
3391 uid under which it was called; it was always trying 10 times. (The default
3392 setting of finduser_retries is zero.) Also, it was sleeping after the final
3393 failure, which is pointless.
3395 24. Implemented tls_on_connect_ports.
3397 25. Implemented acl_smtp_predata.
3399 26. If the domain in control=submission is set empty, Exim assumes that the
3400 authenticated id is a complete email address when it generates From: or
3401 Sender: header lines.
3403 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3404 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3405 chown and chgrp in /bin and hostname in /usr/bin.
3407 28. Exim was keeping the "process log" file open after each use, just as it
3408 does for the main log. This opens the possibility of it remaining open for
3409 long periods when the USR1 signal hits a daemon. Occasional processlog
3410 errors were reported, that could have been caused by this. Anyway, it seems
3411 much more sensible not to leave this file open at all, so that is what now
3414 29. The long-running daemon process does not normally write to the log once it
3415 has entered its main loop, and it closes the log before doing so. This is
3416 so that log files can straightforwardly be renamed and moved. However,
3417 there are a couple of unusual error situations where the daemon does write
3418 log entries, and I had neglected to close the log afterwards.
3420 30. The text of an SMTP error response that was received during a remote
3421 delivery was being truncated at 512 bytes. This is too short for some of
3422 the long messages that one sometimes sees. I've increased the limit to
3425 31. It is now possible to make retry rules that apply only when a message has a
3426 specific sender, in particular, an empty sender.
3428 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3429 it possible to be selective about when SMTP synchronization is enforced.
3431 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3433 32. Implemented hosts_connection_nolog.
3435 33. Added an ACL for QUIT.
3437 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3440 35. Added mailbox_size and mailbox_filecount to appendfile.
3442 36. Added control = no_multiline_responses to ACLs.
3444 37. There was a bug in the logic of the code that waits for the clock to tick
3445 in the case where the clock went backwards by a substantial amount such
3446 that the microsecond fraction of "now" was more than the microsecond
3447 fraction of "then" (but the whole seconds number was less).
3449 38. Added support for the libradius Radius client library this is found on
3450 FreeBSD (previously only the radiusclient library was supported).
3456 1. When certain lookups returned multiple values in the form name=value, the
3457 quoting of the values was not always being done properly. Specifically:
3458 (a) If the value started with a double quote, but contained no whitespace,
3460 (b) If the value contained whitespace other than a space character (i.e.
3461 tabs or newlines or carriage returns) it was not quoted.
3462 This fix has been applied to the mysql and pgsql lookups by writing a
3463 separate quoting function and calling it from the lookup code. The fix
3464 should probably also be applied to nisplus, ibase and oracle lookups, but
3465 since I cannot test any of those, I have not disturbed their existing code.
3467 2. A hit in the callout cache for a specific address caused a log line with no
3468 reason for rejecting RCPT. Now it says "Previous (cached) callout
3469 verification failure".
3471 3. There was an off-by-one bug in the queryprogram router. An over-long
3472 return line was truncated at 256 instead of 255 characters, thereby
3473 overflowing its buffer with the terminating zero. As well as fixing this, I
3474 have increased the buffer size to 1024 (and made a note to document this).
3476 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3477 when Exim is waiting for an SMTP response from a remote server, Exim
3478 restarts its select() call on the socket, thereby resetting its timeout.
3479 This is not a problem when such interrupts are rare. Somebody set up a cron
3480 job to run exiwhat every 2 minutes, which is less than the normal select()
3481 timeout (5 or 10 minutes). This meant that the select() timeout never
3482 kicked in because it was always reset. I have fixed this by comparing the
3483 time when an interrupt arrives with the time at the start of the first call
3484 to select(). If more time than the timeout has elapsed, the interrupt is
3485 treated as a timeout.
3487 5. Some internal re-factoring in preparation for the addition of Sieve
3488 extensions (by MH). In particular, the "personal" test is moved to a
3489 separate function, and given an option for scanning Cc: and Bcc: (which is
3490 not set for Exim filters).
3492 6. When Exim created an email address using the login of the caller as the
3493 local part (e.g. when creating a From: or Sender: header line), it was not
3494 quoting the local part when it contained special characters such as @.
3496 7. Installed new OpenBSD configuration files.
3498 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3499 try to make them clearer.
3501 9. Callout options, other than the timeout value, were being ignored when
3502 verifying sender addresses in header lines. For example, when using
3504 verify = header_sender/callout=no_cache
3506 the cache was (incorrectly) being used.
3508 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3509 only the Cygwin environment.
3511 11. When return_path_on_delivery was set as a log selector, if different remote
3512 addresses in the same message used different return paths and parallel
3513 remote delivery occurred, the wrong values would sometimes be logged.
3514 (Whenever a remote delivery process finished, the return path value from
3515 the most recently started remote delivery process was logged.)
3517 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3518 lines when AUTH and/or TLS are in use. This is the "received protocol"
3519 field. Exim used to use "asmtp" for authenticated SMTP, without any
3520 indication (in the protocol name) for TLS use. Now it follows the RFC and
3521 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3522 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3523 names appear in log lines as well as in Received: header lines.
3525 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3526 extensions, and comparison tests, and to fix some bugs.
3528 14. Changes to the "personal" filter test:
3530 (1) The test was buggy in that it was just doing the equivalent of
3531 "contains" tests on header lines. For example, if a user's address was
3532 anne@some.where, the "personal" test would incorrectly be true for
3534 To: susanne@some.where
3536 This test is now done by extracting each address from the header in turn,
3537 and checking the entire address. Other tests that are part of "personal"
3538 are now done using regular expressions (for example, to check local parts
3539 of addresses in From: header lines).
3541 (2) The list of non-personal local parts in From: addresses has been
3542 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3543 taken from the Sieve specification recommendations.
3545 (3) If the message contains any header line starting with "List-" it is
3546 treated as non-personal.
3548 (4) The test for "circular" in the Subject: header line has been removed
3549 because it now seems ill-conceived.
3551 15. Minor typos in src/EDITME comments corrected.
3553 16. Installed latest exipick from John Jetmore.
3555 17. If headers_add on a router specified a text string that was too long for
3556 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3557 of string_sprintf() is now avoided.
3559 18. $message_body_size was not set (it was always zero) when running the DATA
3560 ACL and the local_scan() function.
3562 19. For the "mail" command in an Exim filter, no default was being set for
3563 the once_repeat time, causing a random time value to be used if "once" was
3564 specified. (If the value happened to be <= 0, no repeat happened.) The
3565 default is now 0s, meaning "never repeat". The "vacation" command was OK
3566 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3567 (I found it when inspecting the code).
3569 20. There is now an overall timeout for performing a callout verification. It
3570 defaults to 4 times the callout timeout, which applies to individual SMTP
3571 commands during the callout. The overall timeout applies when there is more
3572 than one host that can be tried. The timeout is checked before trying the
3573 next host. This prevents very long delays if there are a large number of
3574 hosts and all are timing out (e.g. when the network connections are timing
3575 out). The value of the overall timeout can be changed by specifying an
3576 additional sub-option for "callout", called "maxwait". For example:
3578 verify = sender/callout=5s,maxwait=20s
3580 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3581 to the end before writing, but this should make it even safer).
3583 22. Exim was forgetting that it had advertised PIPELINING for the second and
3584 subsequent messages on an SMTP connection. It was also not resetting its
3585 memory on STARTTLS and an internal HELO.
3587 23. When Exim logs an SMTP synchronization error within a session, it now
3588 records whether PIPELINING has been advertised or not.
3590 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3591 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3592 rather than long int.
3594 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3596 26. Added the never_mail option to autoreply.
3602 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3603 crash if the getsockname() call failed; this can happen if a connection is
3604 closed very soon after it is established. The problem was simply in the
3605 order in which certain operations were done, causing Exim to try to write
3606 to the SMTP stream before it had set up the file descriptor. The bug has
3607 been fixed by making things happen in the correct order.
3613 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3614 before the connection was closed, thus losing the rejection response.
3616 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3617 some early Solaris releases, but causes trouble in current releases where
3618 socklen_t is defined.
3620 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3623 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3626 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3627 that releasing the top of it at the end releases what was used for sub-
3628 expansions (unless the block got too big). However, discard this block if
3629 the first thing is a variable or header, so that we can use its block when
3630 it is dynamic (useful for very large $message_headers, for example).
3632 6. Lookups now cache *every* query, not just the most recent. A new, separate
3633 store pool is used for this. It can be recovered when all lookup caches are
3634 flushed. Lookups now release memory at the end of their result strings.
3635 This has involved some general refactoring of the lookup sources.
3637 7. Some code has been added to the store_xxx() functions to reduce the amount
3638 of flapping under certain conditions.
3640 8. log_incoming_interface used to affect only the <= reception log lines. Now
3641 it causes the local interface and port to be added to several more SMTP log
3642 lines, for example "SMTP connection from", and rejection lines.
3644 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3646 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3648 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3649 could be overwritten at the end of the current message (or the start of a
3650 new message if it was set in a HELO ACL). The value is now preserved for
3651 the duration of the SMTP connection.
3653 12. If a transport had a headers_rewrite setting, and a matching header line
3654 contained an unqualified address, that address was qualified, even if it
3655 did not match any rewriting rules. The underlying bug was that the values
3656 of the flags that permit the existence of unqualified sender and recipient
3657 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3658 non-local messages, and by -bnq for local messages) were not being
3659 preserved with the message after it was received.
3661 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3662 "next input=" as part of the text comprising the host identity instead of
3663 the correct text. The code was using the same buffer for two different
3664 strings. However, depending on which order the printing function evaluated
3665 its arguments, the bug did not always show up. Under Linux, for example, my
3666 test suite worked just fine.
3668 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3669 doesn't work with some older versions of Perl. It has been changed to "my",
3670 which in any case is probably the better facility to use.
3672 15. A really picky compiler found some instances of statements for creating
3673 error messages that either had too many or two few arguments for the format
3676 16. The size of the buffer for calls to the DNS resolver has been increased
3677 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3678 for addresses that have a lot of PTR records. This alleviates a problem; it
3679 does not fully solve it.
3681 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3682 buffer now truncates the list and logs the incident, which is the same
3683 action as happens when Exim is looking up a host name and its aliases.
3684 Previously in this situation something unpredictable would happen;
3685 sometimes it was "internal error: store_reset failed".
3687 18. If a server dropped the connection unexpectedly when an Exim client was
3688 using GnuTLS and trying to read a response, the client delivery process
3689 crashed while trying to generate an error log message.
3691 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3692 string, for example:
3694 warn message = H1: something\nH2: something
3696 the text was added as a single header line from Exim's point of view
3697 though it ended up OK in the delivered message. However, searching for the
3698 second and subsequent header lines using $h_h2: did not work. This has been
3699 fixed. Similarly, if a system filter added multiple headers in this way,
3700 the routers could not see them.
3702 20. Expanded the error message when iplsearch is called with an invalid key to
3703 suggest using net-iplsearch in a host list.
3705 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3706 ACLs are no longer actually imposed (and a message to that effect is
3709 22. If a "gecos" field in a passwd entry contained escaped characters, in
3710 particular, if it contained a \" sequence, Exim got it wrong when building
3711 a From: or a Sender: header from that name. A second bug also caused
3712 incorrect handling when an unquoted " was present following a character
3713 that needed quoting.
3715 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3716 was not being matched caselessly.
3718 24. Arranged for all hyphens in the exim.8 source to be escaped with
3721 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3722 properly. Recipient callout cache records were still being keyed to include
3723 the sender, even when use_sender was set false. This led to far more
3724 callouts that were necessary. The sender is no longer included in the key
3725 when use_sender is false.
3727 26. Added "control = submission" modifier to ACLs.
3729 27. Added the ${base62d: operator to decode base 62 numbers.
3731 28. dnsdb lookups can now access SRV records.
3733 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3734 the configuration file.
3736 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3737 (-v) mode. This makes the output for a verbose queue run more intelligible.
3739 31. Added a use_postmaster feature to recipient callouts.
3741 32. Added the $body_zerocount variable, containing the number of binary zero
3742 bytes in the message body.
3744 33. The time of last modification of the "new" subdirectory is now used as the
3745 "mailbox time last read" when there is a quota error for a maildir
3748 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3750 35. Added +ignore_unknown as a special item in host lists.
3752 36. Code for decoding IPv6 addresses in host lists is now included, even if
3753 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3754 address was recognized as an IP address, but was then not correctly decoded
3755 into binary, causing unexpected and incorrect effects when compared with
3762 1. Very minor rewording of debugging text in manualroute to say "list of
3763 hosts" instead of "hostlist".
3765 2. If verify=header_syntax was set, and a header line with an unqualified
3766 address (no domain) and a large number of spaces between the end of the
3767 name and the colon was received, the reception process suffered a buffer
3768 overflow, and (when I tested it) crashed. This was caused by some obsolete
3769 code that should have been removed. The fix is to remove it!
3771 3. When running in the test harness, delay a bit after writing a bounce
3772 message to get a bit more predictability in the log output.
3774 4. Added a call to search_tidyup() just before forking a reception process. In
3775 theory, someone could use a lookup in the expansion of smtp_accept_max_
3776 per_host which, without the tidyup, could leave open a database connection.
3778 5. Added the variables $recipient_data and $sender_data which get set from a
3779 lookup success in an ACL "recipients" or "senders" condition, or a router
3780 "senders" option, similar to $domain_data and $local_part_data.
3782 6. Moved the writing of debug_print from before to after the "senders" test
3785 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3786 problems for message scanning, either using a data ACL, or using
3787 local_scan() because the Received: header was not generated till after they
3788 were called (in order to set the time as the time of reception completion).
3789 I have revised the way this works. The header is now generated after the
3790 body is received, but before the ACL or local_scan() are called. After they
3791 are run, the timestamp in the header is updated.
3797 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3798 before starting a queue runner without re-exec. This happened only when
3799 deliver_drop_privilege was set or when the Exim user was set to root. The
3800 effect of the bug was that timeouts during subsequent deliveries caused
3801 crashes instead of being properly handled. The handler is now left at its
3802 default (and expected) setting.
3804 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3805 message, again when deliver_drop_privilege is set or Exim is run as root.
3806 The bug described in (1) was not present in this case, but the tidying up
3807 of the other signals was missing. I have made the two cases consistent.
3809 3. The ignore_target_hosts setting on a manualroute router was being ignored
3810 for hosts that were looked up using the /MX notation.
3812 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3815 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3816 operated on the sender address. After changing the $sender_address to <>
3817 for the sender address verify, Exim was re-instated it as the original
3818 (before rewriting) address, but remembering that it had rewritten it, so it
3819 wasn't rewriting it again. This bug also had the effect of breaking the
3820 sender address verification caching when the sender address was rewritten.
3822 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3823 This has been changed so that if the ip literal address matches
3824 ignore_target_hosts, the router declines.
3826 7. Added expansion conditions match_domain, match_address, and match_local_
3827 part (NOT match_host).
3829 8. The placeholder for the Received: header didn't have a length field set.
3831 9. Added code to Exim itself and to exim_lock to test for a specific race
3832 condition that could lead to file corruption when using MBX delivery. The
3833 issue is with the lockfile that is created in /tmp. If this file is removed
3834 after a process has opened it but before that process has acquired a lock,
3835 there is the potential for a second process to recreate the file and also
3836 acquire a lock. This could lead to two Exim processes writing to the file
3837 at the same time. The added code performs the same test as UW imapd; it
3838 checks after acquiring the lock that its file descriptor still refers to
3839 the same named file.
3841 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3842 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3845 11. Added the smtp_active_hostname option. If used, this will typically be made
3846 to depend on the incoming interface address. Because $interface_address is
3847 not set up until the daemon has forked a reception process, error responses
3848 that can happen earlier (such as "too many connections") no longer contain
3851 12. If an expansion in a condition on a "warn" statement fails because a lookup
3852 defers, the "warn" statement is abandoned, and the next ACL statement is
3853 processed. Previously this caused the whole ACL to be aborted.
3855 13. Added the iplsearch lookup type.
3857 14. Added ident_timeout as a log selector.
3859 15. Added tls_certificate_verified as a log selector.
3861 16. Added a global option tls_require_ciphers (compare the smtp transport
3862 option of the same name). This controls incoming TLS connections.
3864 17. I finally figured out how to make tls_require_ciphers do a similar thing
3865 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3866 before starting the TLS session.
3868 18. Tabs are now shown as \t in -bP output.
3870 19. If the log selector return_path_on_delivery was set, Exim crashed when
3871 bouncing a message because it had too many Received: header lines.
3873 20. If two routers both had headers_remove settings, and the first one included
3874 a superfluous trailing colon, the final name in the first list and the
3875 first name in the second list were incorrectly joined into one item (with a
3876 colon in the middle).
3882 1. Added -C and -D options to the exinext utility, mainly to make it easier
3883 to include in the automated testing, but these could be helpful when
3884 multiple configurations are in use.
3886 2. The exinext utility was not formatting the output nicely when there was
3887 an alternate port involved in the retry record key, nor when there was a
3888 message id as well (for retries that were specific to a specific message
3889 and a specific host). It was also confused by IPv6 addresses, because of
3890 the additional colons they contain. I have fixed the IPv4 problem, and
3891 patched it up to do a reasonable job for IPv6.
3893 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3894 delivery, the log line now contains "pipelined" if PIPELINING was used.
3896 4. An SMTP transport process used to panic and die if the bind() call to set
3897 an explicit outgoing interface failed. This has been changed; it is now
3898 treated in the same way as a connect() failure.
3900 5. A reference to $sender_host_name in the part of a conditional expansion
3901 that was being skipped was still causing a DNS lookup. This no longer
3904 6. The def: expansion condition was not recognizing references to header lines
3905 that used bh_ and bheader_.
3907 7. Added the _cache feature to named lists.
3909 8. The code for checking quota_filecount in the appendfile transport was
3910 allowing one more file than it should have been.
3912 9. For compatibility with Sendmail, the command line option
3920 and sets the incoming protocol and host name (for trusted callers). The
3921 host name and its colon can be omitted when only the protocol is to be set.
3922 Note the Exim already has two private options, -pd and -ps, that refer to
3923 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3924 "s", but I don't think that's a major issue.
3926 10. A number of refactoring changes to the code, none of which should affect
3929 (a) The number of logging options was getting close to filling up the
3930 32-bit word that was used as a bit map. I have split them into two classes:
3931 those that are passed in the argument to log_write(), and those that are
3932 only ever tested independently outside of that function. These are now in
3933 separate 32-bit words, so there is plenty of room for expansion again.
3934 There is no change in the user interface or the logging behaviour.
3936 (b) When building, for example, log lines, the code previously used a
3937 macro that called string_cat() twice, in order to add two strings. This is
3938 not really sufficiently general. Furthermore, there was one instance where
3939 it was actually wrong because one of the argument was used twice, and in
3940 one call a function was used. (As it happened, calling the function twice
3941 did not affect the overall behaviour.) The macro has been replaced by a
3942 function that can join an arbitrary number of extra strings onto a growing
3945 (c) The code for expansion conditions now uses a table and a binary chop
3946 instead of a serial search (which was left over from when there were very
3947 few conditions). Also, it now recognizes conditions like "pam" even when
3948 the relevant support is not compiled in: a suitably worded error message is
3949 given if an attempt is made to use such a condition.
3951 11. Added ${time_interval:xxxxx}.
3953 12. A bug was causing one of the ddress fields not to be passed back correctly
3954 from remote delivery subprocesses. The field in question was not being
3955 subsequently used, so this caused to problems in practice.
3957 13. Added new log selectors queue_time and deliver_time.
3959 14. Might have fixed a bug in maildirsizefile handling that threw up
3960 "unexpected character" debug warnings, and recalculated the data
3961 unnecessarily. In any case, I expanded the warning message to give more
3964 15. Added the message "Restricted characters in address" to the statements in
3965 the default ACL that block characters like @ and % in local parts.
3967 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3968 Three changes have been made:
3970 (a) There was a serious bug; a negative response to MAIL caused the whole
3971 recipient domain to be cached as invalid, thereby blocking all messages
3972 to all local parts at the same domain, from all senders. This bug has
3973 been fixed. The domain is no longer cached after a negative response to
3974 MAIL if the sender used is not empty.
3976 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3979 (c) A new callout option, "use_sender" has been added for people who want
3980 the modified behaviour.
3986 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3989 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3990 indeed breaks things for older releases.
3992 3. Added additional logging to the case where there is a problem reading data
3993 from a filter that is running in a subprocess using a pipe, in order to
3994 try to track down a specific problem.
3996 4. Testing facility fudge: when running in the test harness and attempting
3997 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3998 getting "No route to host". Convert this to a timeout.
4000 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4003 6. Some OS don't have socklen_t but use size_t instead. This affects the
4004 fifth argument of getsockopt() amongst other things. This is now
4005 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4006 can be set for individual OS. I have set it for SunOS5, OSF1, and
4007 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4008 some earlier ones do not.
4010 7. Change 4.30/15 was not doing the test caselessly.
4012 8. The standard form for an IPv6 address literal was being rejected by address
4013 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4014 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4015 this, as well as the form without the "IPv6" on the front (but only when
4016 address literals are enabled, of course).
4018 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4020 10. Exim crashed if a message with an empty sender address specified by -f
4021 encountered a router with an errors_to setting. This could be provoked only
4022 by a command such as
4026 where an empty string was supplied; "<>" did not hit this bug.
4028 11. Installed PCRE release 4.5.
4030 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4031 remained set. It is now erased.
4033 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4034 times from message ids (which are base 36 rather than the normal 62).
4036 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4037 were being counted as actual protocol errors, and logged if the log
4038 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4039 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4040 and DATA following a set of rejected RCPTs do not count as protocol errors.
4041 In other words, Exim assumes they were pipelined, though this may not
4042 actually be the case. Of course, in all cases the client gets an
4043 appropriate error code.
4045 15. If a lookup fails in an ACL condition, a message about the failure may
4046 be available; it is used if testing the ACL cannot continue, because most
4047 such messages specify what the cause of the deferral is. However, some
4048 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4049 that caused an old message to be retained and used if a later statement
4050 caused a defer, replacing the real cause of the deferral.
4052 16. If an IP address had so many PTR records that the DNS lookup buffer
4053 was not large enough to hold them, Exim could crash while trying to process
4054 the truncated data. It now detects and logs this case.
4056 17. Further to 4.21/58, another change has been made: if (and only if) the
4057 first line of a message (the first header line) ends with CRLF, a bare LF
4058 in a subsequent header line has a space inserted after it, so as not to
4059 terminate the header.
4061 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4062 unnecessarily, used atoi() instead of strtol(), and didn't check the
4063 termination when getting file sizes from file names by regex.
4065 19. Completely re-implemented the support for maildirsize files, in the light
4066 of a number of problems with the previous contributed implementation
4067 (4.30/29). In particular:
4069 . If the quota is zero, the maildirsize file is maintained, but no quota is
4072 . If the maildir directory does not exist, it is created before any attempt
4073 to write a maildirsize file.
4075 . The quota value in the file is just a cache; if the quota is changed in
4076 the transport, the new value overrides.
4078 . A regular expression is available for excluding directories from the
4081 20. The autoreply transport checks the characters in options that define the
4082 message's headers; it allows continued headers, but it was checking with
4083 isspace() after an embedded newline instead of explicitly looking for a
4086 21. If all the "regular" hosts to which an address was routed had passed their
4087 expiry times, and had not reached their retry times, the address was
4088 bounced, even if fallback hosts were defined. Now Exim should go on to try
4091 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4092 equivalent code in the SMTP transport. Some hosts send humungous responses
4093 to HELO/EHLO, more than 1024 it seems.
4095 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4096 gives compiler warnings in some environments. I've now done it "properly",
4099 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4100 (because of problems with the built-in one) was declared to return uschar *
4101 instead of char *, causing compiler failure.
4103 25. Fixed a file descriptor leak when processing alias/forward files.
4105 26. Fixed a minor format string issue in dbfn.c.
4107 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4109 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4110 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4113 29. When a sender address is verified, it is cached, to save repeating the test
4114 when there is more than one recipient in a message. However, when the
4115 verification involves a callout, it is possible for different callout
4116 options to be set for different recipients. It is too complicated to keep
4117 track of this in the cache, so now Exim always runs a verification when a
4118 callout is required, relying on the callout cache for the optimization.
4119 The overhead is duplication of the address routing, but this should not be
4122 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4123 to be verified with callout=postmaster, and the main callout worked but the
4124 postmaster check failed, the verification correctly failed. However, if a
4125 subsequent RCPT command asked for sender verification *without* the
4126 postmaster check, incorrect caching caused this verification also to fail,
4129 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4130 it was not caching the DNS options (qualify_single, search_parents) that
4131 were used when the lookup failed. A subsequent lookup with different
4132 options therefore always gave the same answer, though there were cases
4133 where it should not have. (Example: a "domains = !$mx_any" option on a
4134 dnslookup router: the "domains" option is always processed without any
4135 widening, but the router might have qualify_single set.) Now Exim uses the
4136 cached value only when the same options are set.
4138 32. Added John Jetmore's "exipick" utility to the distribution.
4140 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4141 than a timeout (e.g. a certificate is required, and is not provided), an
4142 Exim server now closes the connection immediately. Previously it waited for
4143 the client to close - but if the client is SSL, it seems that they each
4144 wait for each other, leading to a delay before one of them times out.
4146 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4147 maintained 0.8.x compatibility because I don't think many are using it, and
4148 it is clearly obsolete.
4150 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4153 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4154 client certificate was expired. A simple patch fixes this, though I don't
4155 understand the full logic of why the verify callback is called multiple
4158 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4159 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4160 which causes problems with some clients (such as the Certicom SSL Plus
4161 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4162 disables the coutermeasure allowing Eudora to connect."
4164 38. Exim was not checking that a write() to a log file succeeded. This could
4165 lead to Bad Things if a log got too big, in particular if it hit a file
4166 size limit. Exim now panics and dies if it cannot write to a log file, just
4167 as it does if it cannot open a log file.
4169 39. Modified OS/Makefile-Linux so that it now contains
4171 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4173 The two -D definitions ensure that Exim is compiled with large file
4174 support, which makes it possible to handle log files that are bigger than
4177 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4178 instance) a domain was checked against a named list that involved a lookup,
4179 causing $domain_data to be set, then another domain was checked against the
4180 same list, then the first domain was re-checked, the value of $domain_data
4181 after the final check could be wrong. In particular, if the second check
4182 failed, it could be set empty. This bug probably also applied to
4185 41. The strip_trailing_dot option was not being applied to the address given
4186 with the -f command-line option.
4188 42. The code for reading a message's header from the spool was incrementing
4189 $received_count, but never initializing it. This meant that the value was
4190 incorrect (doubled) while delivering a message in the same process in which
4191 it was received. In the most common configuration of Exim, this never
4192 happens - a fresh exec is done - but it can happen when
4193 deliver_drop_privilege is set.
4195 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4196 it now includes up to 150 characters of the unexpected data in the log
4199 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4200 and building data strings. The size of both of these buffers was 10 000
4201 bytes - far larger than anybody would *ever* want, thought I. Needless to
4202 say, somebody hit the limit. I have increased the maximum line length to
4203 20 000 and the maximum data length of concatenated lines to 100 000. I have
4204 also fixed two bugs, because there was no checking on these buffers. Tsk,
4205 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4206 buffer is too small.
4208 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4209 lsearch lookups. Now it does.
4211 46. When parsing a route_list item in a manualroute router, a fixed-length
4212 buffer was used for the list of hosts. I made this 1024 bytes long,
4213 thinking that nobody would ever have a list of hosts that long. Wrong.
4214 Somebody had a whole pile of complicated expansion conditions, and the
4215 string was silently truncated, leading to an expansion error. It turns out
4216 that it is easier to change to an unlimited length (owing to other changes
4217 that have happened since this code was originally written) than to build
4218 structure for giving a limitation error. The length of the item that
4219 expands into the list of hosts is now unlimited.
4221 47. The lsearch lookup could not handle data where the length of text line was
4222 more than 4095 characters. Such lines were truncated, leading to shortened
4223 data being returned. It should now handle lines of any length.
4225 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4226 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4229 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4230 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4231 no longer happens in comment lines. A list of replacements is now placed
4232 at the head of all of the source files, except those whose only change is
4233 to replace PERL_COMMAND in the very first #! line.
4235 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4236 messages on the queue, with a bottom-up merge sort, using code contributed
4237 by Michael Haardt. This should make operations like -bp somewhat faster on
4238 large queues. It won't affect queue runners, except when queue_run_in_order
4241 51. Installed eximstats 1.31 in the distribution.
4243 52. Added support for SRV lookups to the dnslookup router.
4245 53. If an ACL referred to $message_body or $message_body_end, the value was not
4246 reset for any messages that followed in the same SMTP session.
4248 54. The store-handling optimization for building very long strings was not
4249 differentiating between the different store pools. I don't think this
4250 actually made any difference in practice, but I've tidied it.
4252 55. While running the routers to verify a sender address, $sender_address
4253 was still set to the sender address. This is wrong, because when routing to
4254 send a bounce to the sender, it would be empty. Therefore, I have changed
4255 it so that, while verifying a sender address, $sender_address is set to <>.
4256 (There is no change to what happens when verifying a recipient address.)
4258 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4259 target A or AAAA records (if not already returned) without resetting the
4260 qualify_single or search_parents options of the DNS resolver. These are
4261 inappropriate in this case because the targets of MX and SRV records must
4262 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4263 target that, when qualified, matched something in the local domain. These
4264 two options are now turned off when doing these lookups.
4266 57. It seems that at least some releases of Reiserfs (which does not have the
4267 concept of a fixed number of inodes) returns zero and not -1 for the
4268 number of available inodes. This interacted badly with check_spool_inodes,
4269 which assumed that -1 was the "no such thing" setting. What I have done is
4270 to check that the total number of inodes is greater than zero before doing
4271 the test of how many are available.
4273 58. When a "warn" ACL statement has a log_message modifier, the message is
4274 remembered, and not repeated. This is to avoid a lot of repetition when a
4275 message has many recipients that cause the same warning to be written.
4276 Howewer, Exim was preserving the list of already written lines for an
4277 entire SMTP session, which doesn't seem right. The memory is now reset if a
4278 new message is started.
4280 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4281 debugging output unless log_rewrite was also set.
4283 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4284 of (int)(handle) when we know that handle contains (void *)(-1).
4286 61. The Exim daemon panic-logs an error return when it closes the incoming
4287 connection. However "connection reset by peer" seems to be common, and
4288 isn't really an error worthy of noting specially, so that particular error
4291 62. When Exim is trying to find all the local interfaces, it used to panic and
4292 die if the ioctl to get the interface flags failed. However, it seems that
4293 on at least one OS (Solaris 9) it is possible to have an interface that is
4294 included in the list of interfaces, but for which you get a failure error
4295 for this call. This happens when the interface is not "plumbed" into a
4296 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4297 failure of the "get flags" call assumes that the interface is down.
4299 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4300 makes life easier for people who are doing arithmetic on fields extracted
4301 from dates, where you often get leading zeros that should not be
4302 interpreted as octal.
4304 64. Added qualify_domain to the redirect router, to override the global
4307 65. If a pathologically long header line contained very many addresses (the
4308 report of this problem mentioned 10 000) and each of them was rewritten,
4309 Exim could use up a very large amount of memory. (It kept on making new
4310 copies of the header line as it rewrote, and never released the old ones.)
4311 At the expense of a bit more processing, the header rewriting function has
4312 been changed so that it no longer eats memory in this way.
4314 66. The generation of the Received: header has been moved from the time that a
4315 message starts to be received, to the time that it finishes. The timestamp
4316 in the Received: header should now be very close to that of the <= log
4317 line. There are two side-effects of this change:
4319 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4320 logged header lines no longer include the local Received: line, because
4321 it has not yet been created. The same applies to a copy of the message
4322 that is returned to a non-SMTP sender when a message is rejected.
4324 (b) When a filter file is tested using -bf, no additional Received: header
4325 is added to the test message. After some thought, I decided that this
4328 This change does not affect the value of $received_for. It is still set
4329 after address rewriting, but before local_scan() is called.
4331 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4333 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4334 gave an unhelpful panic error message, and a defer error. I have managed to
4335 change this behaviour so that it now rejects any supplied certificate,
4336 which seems right, as the list of acceptable certificates is empty.
4338 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4339 gave an unhelpful defer error. I have not managed to make this reject any
4340 supplied certificates, but the error message it gives is "no certificate
4341 supplied", which is not helpful.
4343 70. exigrep's output now also includes lines that are not associated with any
4344 message, but which match the given pattern. Implemented by a patch from
4345 Martin Sluka, which also tidied up the Perl a bit.
4347 71. Recipient callout verification, like sender verification, was using <> in
4348 the MAIL FROM command. This isn't really the right thing, since the actual
4349 sender may affect whether the remote host accepts the recipient or not. I
4350 have changed it to use the actual sender in the callout; this means that
4351 the cache record is now keyed on a recipient/sender pair, not just the
4352 recipient address. There doesn't seem to be a real danger of callout loops,
4353 since a callout by the remote host to check the sender would use <>.
4354 [SEE ABOVE: changed after hitting problems.]
4356 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4357 temporary errors. However, in the case of such a code being given after
4358 the end of a data transmission (i.e. after ".") Exim was failing to write
4359 a retry record for the message. (Yes, there was some broken host that was
4360 actually sending 8xx at this point.)
4362 73. An unknown lookup type in a host list could cause Exim to panic-die when
4363 the list was checked. (An example that provoked this was putting <; in the
4364 middle of a list instead of at the start.) If this happened during a DATA
4365 ACL check, a -D file could be left lying around. This kind of configuration
4366 error no longer causes Exim to die; instead it causes a defer errror. The
4367 incident is still logged to the main and panic logs.
4369 74. Buglet left over from Exim 3 conversion. The message "too many messages
4370 in one connection" was written to the rejectlog but not the mainlog, except
4371 when address rewriting (yes!) was being logged.
4373 75. Added write_rejectlog option.
4375 76. When a system filter was run not as root (that is, when system_filter_user
4376 was set), the values of the $n variables were not being returned to the
4377 main process; thus, they were not subsequently available in the $sn
4380 77. Added +return_path_on_delivery log selector.
4382 78. A connection timeout was being treated differently from recipients deferred
4383 when testing hosts_max_try with a message that was older than the host's
4384 retry timeout. (The host should not be counted, thus allowing all hosts to
4385 be tried at least once before bouncing.) This may have been the cause of an
4386 occasionally reported bug whereby a message would remain on the queue
4387 longer than the retry timeout, but would be bounced if a delivery was
4388 forced. I say "may" because I never totally pinned down the problem;
4389 setting up timeout/retry tests is difficult. See also the next item.
4391 79. The ultimate address timeout was not being applied to errors that involved
4392 a combination of host plus message (for example, a timeout on a MAIL
4393 command). When an address resolved to a number of possible hosts, and they
4394 were not all tried for each delivery (e.g. because of hosts_max_try), a
4395 message could remain on the queue longer than the retry timeout.
4397 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4400 81. Fixed an obscure SMTP outgoing bug which required at least the following
4401 conditions: (a) there was another message waiting for the same server;
4402 (b) the server returned 5xx to all RCPT commands in the first message so
4403 that the message was not completed; (c) the server dropped the connection
4404 or gave a negative response to the RSET that Exim sends to abort the
4405 transaction. The observed case was a dropped connection after DATA that had
4406 been sent in pipelining mode. That is, the server had advertised PIPELINING
4407 but was not implementing it correctly. The effect of the bug was incorrect
4408 behaviour, such as trying another host, and this could lead to a crash.
4414 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4415 and daemon.c were passed as pointers to ints; they should have been
4416 pointers to socklen_t variables (which are typically unsigned ints).
4418 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4421 3. Fixed a really odd bug that affected only the testing scheme; patching a
4422 certain fixed string in the binary changed the value of another string that
4423 happened to be identical to the end of the original first string.
4425 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4426 name", it returns that address as the IP address. On some operating
4427 systems (e.g. Solaris), it also passes back the IP address string as the
4428 "host name". However, on others (e.g. Linux), it passes back an empty
4429 string. Exim wasn't checking for this, and was changing the host name to an
4430 empty string, assuming it had been canonicized.
4432 5. Although rare, it is permitted to have more than one PTR record for a given
4433 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4434 all the names associated with an address, because they do in Solaris.
4435 However, it seems that they do not in Linux for data that comes from the
4436 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4437 I found this out when I moved to a new Linux workstation and tried to run
4438 the Exim test suite.
4440 To get round this problem I have changed the code so that it now does its
4441 own call to the DNS to look up PTR records when searching for a host name.
4442 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4443 addresses that are only in /etc/hosts are still found.
4445 This behaviour is, however, controlled by an option called host_lookup_
4446 order, which defaults to "bydns:byaddr". If people want to use the other
4447 order, or indeed, just use one or the other means of lookup, they can
4448 specify it in this variable.
4450 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4451 some operating systems, this comes back from gethostbyaddr() as an empty
4452 string, and this is what Exim used to test for. However, it seems that in
4453 other systems, "." is yielded. Exim now tests for this case too.
4455 7. The values of check_spool_space and check_log_space are now held internally
4456 as a number of kilobytes instead of an absolute number of bytes. If a
4457 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4458 kilobyte. This means that much larger values can be stored.
4460 8. Exim monitor: an attempt to get the action menu when not actually pointing
4461 at a message produces an empty menu entitled "No message selected". This
4462 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4463 no entries in it ("Shell widget menu has zero width and/or height"). So I
4464 have added a single, blank menu entry in this case.
4466 9. Added ${quote_local_part.
4468 10. MIME decoding is now applied to the contents of Subject: header lines when
4471 11. Now that a reference to $sender_host_address automatically causes a reverse
4472 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4473 host lookup before query-style lookups in lists that might use this
4474 variable. This has therefore been abolished, and the "net-" prefix is no
4475 longer necessary for query-style lookups.
4477 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4478 have been a typo for LFLAGS, so it has been changed.
4480 13. The install script calls Exim with "-C /dev/null" in order to find the
4481 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4482 to be output. Howeve, since Exim outputs its version number before the
4483 error, it didn't break the script. It just looked ugly. I fixed this by
4484 always allowing "-C /dev/null" if the caller is root.
4486 14. Ignore overlarge ACL variable number when reading spool file - insurance
4487 against a later release with more variables having written the file.
4489 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4490 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4491 this, as well as the form without the "IPv6" on the front.
4493 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4494 OS/Makefile-Darwin file.
4496 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4497 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4499 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4500 variable. However, some OS use ints for inodes. I've added cast to long int
4501 to get rid of the compiler warning.
4503 19. I had forgotten to lock out "/../" in configuration file names when
4504 ALT_CONFIG_PREFIX was set.
4506 20. Routers used for verification do not need to specify transports. However,
4507 if such a router generated a host list, and callout was configured, Exim
4508 crashed, because it could not find a port number from the (non-existent)
4509 transport. It now assumes port 25 in this circumstance.
4511 21. Added the -t option to exigrep.
4513 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4514 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4515 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4516 them is set, LOOKUP_LSEARCH is forced.
4518 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4520 24. Added sender and host information to the "rejected by local_scan()" log
4521 line; previously there was no indication of these.
4523 25. Added .include_if_exists.
4525 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4526 Linux. It turns out that not all file systems support this. Apparently some
4527 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4528 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4529 not supported on the file descriptor, is now ignored when Exim is trying to
4530 sync a directory. This applies only to Linux.
4532 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4534 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4535 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4538 29. Added support for maildirsize files from supplied patch (modified a bit).
4540 30. The use of :fail: followed by an empty string could lead Exim to respond to
4541 sender verification failures with (e.g.):
4543 550 Verification failed for <xxx>
4544 550 Sender verify failed
4546 where the first response line was missing the '-' that indicates it is not
4547 the final line of the response.
4549 31. The loop for finding the name of the user that called Exim had a hardwired
4550 limit of 10; it now uses the value of finduser_retries, which is used for
4551 all other user lookups.
4553 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4556 33. Exim was neglecting to zero errno before one call of strtol() when
4557 expanding a string and expecting an integer value. On some systems this
4558 resulted in spurious "integer overflow" errors. Also, it was casting the
4559 result into an int without checking.
4561 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4562 did not work. The code looks as if it has *never* worked, though it appears
4563 to have been documented since at least releast 1.62. I have made it work.
4565 35. The "timeout_DNS" error in retry rules, also documented since at least
4566 1.62, also never worked. As it isn't clear exactly what this means, and
4567 clearly it isn't a major issue, I have abolished the feature by treating it
4568 as "timeout", and writing a warning to the main and panic logs.
4570 36. The display of retry rules for -brt wasn't always showing the error code
4573 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4574 timeout_connect_A, timeout_connect_MX.
4576 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4577 to the empty sender.
4579 39. The daemon was not analysing the content of -oX till after it had closed
4580 stderr and disconnected from the controlling terminal. This meant that any
4581 syntax errors were only noted on the panic log, and the return code from
4582 the command was 0. By re-arranging the code a little, I've made the
4583 decoding happen first, so such errors now appear on stderr, and the return
4584 code is 1. However, the actual setting up of the sockets still happens in
4585 the disconnected process, so errors there are still only recorded on the
4588 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4589 connections (as happens on some IP stacks) was logged at start up time as
4590 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4591 it from "IPv6 and IPv4", which means that two separate sockets are being
4594 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4595 says whether AF_INET or AF_INET6 was passed as an argument.
4597 42. Exiwhat output was messed up when time zones were included in log
4600 43. Exiwhat now gives more information about the daemon's listening ports,
4601 and whether -tls-on-connect was used.
4603 44. The "port" option of the smtp transport is now expanded.
4605 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4606 silently ignored. Now an error message is written to the main and panic
4609 46. There's a new ACL modifier called "logwrite" which writes to a log file
4610 as soon as it is encountered.
4612 47. Added $local_user_uid and $local_user_gid at routing time.
4614 48. Exim crashed when trying to verify a sender address that was being
4617 49. Exim was recognizing only a space character after ".include". It now also
4618 recognizes a tab character.
4620 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4621 extracting the relevant information from the specification. The man page no
4622 longer contains scrambled data for the -d option, and I've added a section
4623 at the front about calling Exim under different names.
4625 51. Added "extra_headers" argument to the "mail" command in filter files.
4627 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4630 53. Installed eximstats 1.29.
4632 54. Added transport_filter_timeout as a generic transport option.
4634 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4635 Cc: header lines. This was required by RFC 822, but it not required by RFC
4638 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4639 incoming messages that did not have them. Now it does so only if the
4640 message originates locally, that is, if there is no associated remote host
4641 address. When Resent- header lines are present, this applies to the Resent-
4642 lines rather than the non-Resent- lines.
4644 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4645 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4647 58. Messages for configuration errors now include the name of the main
4648 configuration file - useful now that there may be more than one file in a
4649 list (.included file names were always shown).
4651 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4652 for those rare installations that do not start the daemon as root or run it
4653 setuid root. I've cut out the call to initgroups() if the daemon is not
4656 60. The Exim user and group can now be bound into the binary as text strings
4657 that are looked up at the start of Exim's processing.
4659 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4661 62. Added $mailstore_basename variable.
4663 63. Installed patch to sieve.c from Michael Haardt.
4665 64. When Exim failed to open the panic log after failing to open the main log,
4666 the original message it was trying to log was written to stderr and debug
4667 output, but if they were not available (the usual case in production), it
4668 was lost. Now it is written to syslog before the two lines that record the
4669 failures to open the logs.
4671 65. Users' Exim filters run in subprocesses under the user's uid. It is
4672 possible for a "deliver" command or an alias in a "personal" command to
4673 provoke an address rewrite. If logging of address rewriting is configured,
4674 this fails because the process is not running as root or exim. There may be
4675 a better way of dealing with this, but for the moment (because 4.30 needs
4676 to be released), I have disabled address rewrite logging when running a
4677 filter in a non-root, non-exim process.
4683 1. The buildconfig auxiliary program wasn't quoting the value set for
4684 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4685 not defined. This bug was masked in 4.22 by the effect that was fixed in
4688 2. Some messages that were rejected after a message id was allocated were
4689 shown as "incomplete" by exigrep. It no longer does this for messages that
4690 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4692 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4693 have allow_domain_literals set, the ID did not get logged in the <= line.
4694 Domain literals are now always recognized in Message-ID: header lines.
4696 4. The first argument for a ${extract expansion item is the key name or field
4697 number. Leading and trailing spaces in this item were not being ignored,
4698 causing some misleading effects.
4700 5. When deliver_drop_privilege was set, single queue runner processes started
4701 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4702 same command in the process it spins off) were not dropping privilege.
4704 6. When the daemon running as "exim" started a queue runner, it always
4705 re-executed Exim in the spun-off process. This is a waste of effort when
4706 deliver_drop_privilege is set. The new process now just calls the
4707 queue-runner function directly.
4713 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4716 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4717 ignored. Though the use of -oP was forcing the writing of a pid file, it
4718 was always written to the default place.
4720 3. If the message "no IP address found for host xxxx" is generated during
4721 incoming verification, it is now followed by identification of the incoming
4722 connection (so you can more easily find what provoked it).
4724 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4726 5. Added some features to "harden" Exim a bit more against certain attacks:
4728 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4729 be put in Local/Makefile. This is like the never_users runtime option,
4730 but it cannot be overridden. The default setting is "root".
4732 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4733 prefix string with which any file named in a -C command line option
4736 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4737 is retained for -C and -D only if the caller of Exim is root. Without
4738 it, the exim user may also use -C and -D and retain privilege.
4740 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4741 command line option is disabled.
4743 6. Macro names set by the -D option must start with an upper case letter, just
4744 like macro names defined in the configuration file.
4746 7. Added "dereference=" facility to LDAP.
4748 8. Two instances of the typo "uknown" in the source files are fixed.
4750 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4751 the Configure-Makefile script screwed up while processing it.
4753 10. Incorporated PCRE 4.4.
4755 11. The SMTP synchronization check was not operating right at the start of an
4756 SMTP session. For example, it could not catch a HELO sent before the client
4757 waited for the greeting. There is now a check for outstanding input at the
4758 point when the greeting is written. Because of the duplex, asynchronous
4759 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4760 way, but not yet received, when the check is performed.
4762 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4763 on TCP/IP sockets, because this apparently causes some broken clients to
4766 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4767 unchanged) from the Cygwin maintainer.
4769 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4770 was supported instead of testing for mbx. Effectively a typo.
4772 15. The spa authenticator server code was not checking that the input it
4773 received was valid base64.
4775 16. The debug output line for the "set" modifier in ACLs was not showing the
4776 name of the variable that was being set.
4778 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4780 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4781 lookup on its own. Something else was needed to trigger the lookup. For
4782 example, a match in host_lookup or the need for a host name in a host list.
4783 Now, if $sender_host_name is referenced and the host name has not yet been
4784 looked up, a lookup is performed. If the lookup fails, the variable remains
4785 empty, and $host_lookup_failed is set to "1".
4787 19. Added "eqi" as a case-independent comparison operator.
4789 20. The saslauthd authentication condition could segfault if neither service
4790 nor realm was specified.
4792 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4793 error message that was logged was misleading, and incoming SMTP
4794 connections were dropped. The message is now more accurate, and temporary
4795 errors are given to SMTP connections.
4797 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4798 (except RSET) with a 421 error, until QUIT is received. However, it was
4799 failing to send a response to QUIT.
4801 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4802 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4803 ACL is now run after the helo_try_verify_hosts code.
4805 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4806 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4807 case-independent, so other case variants are also recognized). Apparently
4808 some systems use these upper case variants.
4810 25. If more than two messages were waiting for the same host, and a transport
4811 filter was specified for the transport, Exim sent two messages over the
4812 same TCP/IP connection, and then failed with "socket operation on non-
4813 socket" when it tried to send the third.
4815 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4817 27. The extern definition of crypt16() in expand.c was not being excluded when
4818 the OS had its own crypt16() function.
4820 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4821 as a preferred synonym for return_size_limit, both as an option and as an
4824 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4826 30. Changed the default configuration ACL to relax the local part checking rule
4827 for addresses that are not in any local domains. For these addresses,
4828 slashes and pipe symbols are allowed within local parts, but the sequence
4829 /../ is explicitly forbidden.
4831 31. SPA server authentication was not clearing the challenge buffer before
4834 32. log_message in a "warn" ACL statement was writing to the reject log as
4835 well as to the main log, which contradicts the documentation and doesn't
4836 seem right (because no rejection is happening). So I have stopped it.
4838 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4839 However, I am unable to do any testing of this.
4841 34. Fixed an infelicity in the appendfile transport. When checking directories
4842 for a mailbox, to see if any needed to be created, it was accidentally
4843 using path names with one or more superfluous leading slashes; tracing
4844 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4846 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4847 contents are added to the log line that is written for every discarded
4848 recipient. (Previously a log_message setting was ignored.)
4850 36. The ${quote: operator now quotes the string if it is empty.
4852 37. The install script runs exim in order to find its version number. If for
4853 some reason other than non-existence or emptiness, which it checks, it
4854 could not run './exim', it was installing it with an empty version number,
4855 i.e. as "exim-". This error state is now caught, and the installation is
4858 38. An argument was missing from the function that creates an error message
4859 when Exim fails to connect to the socket for saslauthd authentication.
4860 This could cause Exim to crash, or give a corrupted message.
4862 39. Added isip, isip4, and isip6 to ${if conditions.
4864 40. The ACL variables $acl_xx are now saved with the message, and can be
4865 accessed later in routers, transports, and filters.
4867 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4868 strings in the file are not string-expanded.
4870 42. If a MAIL command specified a SIZE value that was too large to fit into an
4871 int variable, the check against message_size_limit failed. Such values are
4872 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4873 day this will have to be increased, but I don't think I want to be around
4874 when emails are that large.
4881 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4882 iconv() is not standard in FreeBSD.
4884 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4885 IPv6 enabled. The observed symptom was a segmentation fault on return from
4886 the function os_common_find_running_interfaces() in src/os.c.
4888 3. In the check_special_case() function in daemon.c I had used "errno" as an
4889 argument name, which causes warnings on some systems. This was basically a
4890 typo, since it was named "eno" in the comments!
4892 4. The code that waits for the clock to tick (at a resolution of some fraction
4893 of a second) so as to ensure message-id uniqueness was always waiting for
4894 at least one whole tick, when it could have waited for less. [This is
4895 almost certainly not relevant at current processor speeds, where it is
4896 unlikely to ever wait at all. But we try to future-proof.]
4898 5. The function that sleeps for a time interval that includes fractions of a
4899 second contained a race. It did not block SIGALRM between setting the
4900 timer, and suspending (a couple of lines later). If the interval was short
4901 and the sigsuspend() was delayed until after it had expired, the suspension
4902 never ended. On busy systems this could lead to processes getting stuck for
4905 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4906 process, before it forks any delivery processes. The open lookup caching
4907 mechanism meant that the open file or database connection was passed into
4908 the delivery process. The problem was that delivery processes always tidy
4909 up cached lookup data. This could cause a problem for the next delivery
4910 process started by the queue runner, because the external queue runner
4911 process does not know about the closure. So the next delivery process
4912 still has data in the lookup cache. In the case of a file lookup, there was
4913 no problem because closing a file descriptor in a subprocess doesn't affect
4914 the parent. However, if the lookup was caching a connection to a database,
4915 the connection was closed, and the second delivery process was likely to
4916 see errors such as "PGSQL: query failed: server closed the connection
4917 unexpectedly". The problem has been fixed by closing all cached lookups
4918 in a queue runner before running a delivery process.
4920 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4921 seem to have the "const" qualifier which it has on other OS. I've
4924 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4925 *of the same type* (client or server) with the same public name that an
4926 error should be diagnosed.
4928 9. When Exim looked up a host name for an IP address, but failed to find the
4929 original IP address when looking up the host name (a safety check), it
4930 output the message "<ip address> does not match any IP for NULL", which was
4931 confusing, to say the least. The bug was that the host name should have
4932 appeared instead of "NULL".
4934 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4935 user that is built into the binary, and the -C or -D options is used, root
4936 privilege is dropped before the configuration file is read. In addition,
4937 logging is switched to stderr instead of the normal log files. If the
4938 configuration then re-defines the Exim user, the unprivileged environment
4939 is probably not what is expected, so Exim logs a panic warning message (but
4942 However, if deliver_drop_privilege is set, the unprivileged state may well
4943 be exactly what is intended, so the warning has been cut out in that case,
4944 and Exim is allowed to try to write to its normal log files.
4950 1. smtp_return_error_details was not giving details for temporary sender
4951 or receiver verification errors.
4953 2. Diagnose a configuration error if two authenticators have the same public
4956 3. Exim used not to create the message log file for a message until the first
4957 delivery attempt. This could be confusing when incoming messages were held
4958 for policy or load reasons. The message log file is now created at the time
4959 the message is received, and an initial "Received" line is written to it.
4961 4. The automatically generated man page for command line options had a minor
4962 bug that caused no ill effects; however, a more serious problem was that
4963 the procedure for building the man page automatically didn't always
4964 operate. Consequently, release 4.20 contains an out-of-date version. This
4965 shouldn't happen again.
4967 5. When building Exim with embedded Perl support, the script that builds the
4968 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4969 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4971 6. The freeze_tell option was not being used for messages that were frozen on
4972 arrival, either by an ACL or by local_scan().
4974 7. Added the smtp_incomplete_transaction log selector.
4976 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4977 was accepting AUTH without a new EHLO.
4979 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4980 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4983 10. Logging of TCP/IP connections (when configured) now happens in the main
4984 daemon process instead of the child process, so that the TCP/IP connection
4985 count is more accurate (but it can never be perfect).
4987 11. The use of "drop" in a nested ACL was not being handled correctly in the
4988 outer ACL. Now, if condition failure induced by the nested "drop" causes
4989 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4990 or "require"), the connection is dropped.
4992 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4993 that yield "discard" can now be used with an "accept" or a "discard" verb,
4994 but an error is generated for any others (because I can't see a useful way
4995 to define what should happen).
4997 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4998 are now processed in the same way as lines in the Exim configuration file.
4999 In particular, continuation lines are supported.
5001 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5003 15. Added -ti meaning -t -i.
5005 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5006 domains, and warn by logging if others are found.
5008 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5009 returned from GIFCONF when Exim is trying to find the list of interfaces on
5010 a host. The code in os.c has been modified to copy each ifreq to an aligned
5011 structure in all cases.
5013 Also, in some cases, the returned ifreq's were being copied to a 'struct
5014 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5015 means the last couple of bytes of an IPv6 address could be chopped if the
5016 ifreq contained only a normal sockaddr (14 bytes storage).
5018 18. Named domain lists were not supported in the hosts_treat_as_local option.
5019 An entry such as +xxxx was not recognized, and was treated as a literal
5022 19. Ensure that header lines added by a DATA ACL are included in the reject log
5023 if the ACL subsequently rejects the message.
5025 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5026 MD5 (which is deprecated).
5028 21. When testing a filter file using -bf, Exim was writing a message when it
5029 took the sender from a "From " line in the message, but it was not doing so
5030 when it took $return_path from a Return-Path: header line. It now does.
5032 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5033 with a valid header line field name (a series of printing characters
5034 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5036 23. Changed "disc" in the source to "disk" to conform to the documentation and
5037 the book and for uniformity.
5039 24. Ignore Sendmail's -Ooption=value command line item.
5041 25. When execve() failed while trying to run a command in a pipe transport,
5042 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5043 could be confused with a return value of 69 from the command itself. This
5044 has been changed to 127, the value the shell returns if it is asked to run
5045 a non-existent command. The wording for the related log line suggests a
5046 non-existent command as the problem.
5048 26. If received_header_text expands to an empty string, do not add a Received:
5049 header line to the message. (Well, it adds a token one on the spool, but
5050 marks it "old" so that it doesn't get used or transmitted.)
5052 27. Installed eximstats 1.28 (addition of -nt option).
5054 28. There was no check for failure on the call to getsockname() in the daemon
5055 code. This can fail if there is a shortage of resources on the system, with
5056 ENOMEM, for example. A temporary error is now given on failure.
5058 29. Contrary to the C standard, it seems that in some environments, the
5059 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5060 program. Exim now does this explicitly; it affects the formatting of
5061 timestamps using strftime().
5063 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5064 complaints. I've now initialized all the variables, to avoid this.
5066 32. Header lines added by a system filter were not being "seen" during
5067 transport-time rewrites.
5069 33. The info_callback() function passed to OpenSSL is set up with type void
5070 (*)(SSL *, int, int), as described somewhere. However, when calling the
5071 function (actually a macro) that sets it up, the type void(*)() is
5072 expected. I've put in a cast to prevent warnings from picky compilers.
5074 34. If a DNS black list lookup found a CNAME record, but there were no A
5075 records associated with the domain it pointed at, Exim crashed.
5077 35. If a DNS black list lookup returned more than one A record, Exim ignored
5078 all but the first. It now scans all returned addresses if a particular IP
5079 value is being sought. In this situation, the contents of the
5080 $dnslist_value variable are a list of all the addresses, separated by a
5083 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5084 to accept a host name and all its aliases if the forward lookup for any of
5085 them yielded the IP address of the incoming connection. Now it accepts only
5086 those names whose forward lookup yields the correct IP address. Any other
5087 names are discarded. This closes a loophole whereby a rogue DNS
5088 administrator could create reverse DNS records to break through a
5089 wildcarded host restriction in an ACL.
5091 37. If a user filter or a system filter that ran in a subprocess used any of
5092 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5093 the wrong values were passed to the pipe command ($thisaddress had the
5094 value of $0, $0 had the value of $1, etc). This bug was introduced by
5095 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5097 38. Improved the line breaking for long SMTP error messages from ACLs.
5098 Previously, if there was no break point between 40 and 75 characters, Exim
5099 left the rest of the message alone. Two changes have been made: (a) I've
5100 reduced the minimum length to 35 characters; (b) if it can't find a break
5101 point between 35 and 75 characters, it looks ahead and uses the first one
5102 that it finds. This may give the occasional overlong line, but at least the
5103 remaining text gets split now.
5105 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5106 file descriptors might be low, and that setting 1000 would always raise it.
5107 It turns out that in some environments, the limit is already over 1000 and
5108 that lowering it causes trouble. So now Exim takes care not to decrease it.
5110 40. When delivering a message, the value of $return_path is set to $sender_
5111 address at the start of routing (routers may change the value). By an
5112 oversight, this default was not being set up when an address was tested by
5113 -bt or -bv, which affected the outcome if any router or filter referred to
5116 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5117 log only when "message" or "log_message" are set. However, if one of the
5118 conditions was an address verification, or a call to a nested ACL, the
5119 messages generated by the underlying test were being passed through. This
5120 no longer happens. The underlying message is available in $acl_verify_
5121 message for both "message" and "log_message" expansions, so it can be
5122 passed through if needed.
5124 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5125 new expansion $bh_ to give the encoded byte string without charset
5126 translation. Translation happens only if iconv() is available; HAVE_ICONV
5127 indicates this at build time. HEADERS_CHARSET gives the charset to
5128 translate to; headers_charset can change it in the configuration, and
5129 "headers charset" can change it in an individual filter file.
5131 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5132 that creates RFC 2047 encoded "words" labels them as that charset instead
5133 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5134 expansion operator; (ii) when Exim creates a From: line for a local
5135 message; (iii) when a header line is rewritten to include a "phrase" part.
5137 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5138 buggy, causing it to skip the first lines of messages whose message ID
5139 ended in 'D'. This would not have bitten before Exim release 4.14, because
5140 message IDs were unlikely to end in 'D' before then. The effect was to have
5141 incorrect size information for certain domains.
5143 45. #include "config.h" was missing at the start of the crypt16.c module. This
5144 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5147 46. If there was a timeout during a "random" callout check, Exim treated it as
5148 a failure of the random address, and carried on sending RSET and the real
5149 address. If the delay was just some slowness somewhere, the response to the
5150 original RCPT would be taken as a response to RSET and so on, causing
5151 mayhem of various kinds.
5153 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5154 when I implemented it. It didn't allow for the fact that some option values
5155 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5156 the right test for positive values.
5158 48. Domain names in DNS records are case-independent. Exim always looks them up
5159 in lower case. Some resolvers return domain names in exactly the case they
5160 appear in the zone file, that is, they may contain uppercase letters. Not
5161 all resolvers do this - some return always lower case. Exim was treating a
5162 change of case by a resolver as a change of domain, similar to a widening
5163 of a domain abbreviation. This triggered its re-routing code and so it was
5164 trying to route what was effectively the same domain again. This normally
5165 caused routing to fail (because the router wouldn't handle the domain
5166 twice). Now Exim checks for this case specially, and just changes the
5167 casing of the domain that it ultimately uses when it transmits the message
5170 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5173 50. If a filter generated a file delivery with a non-absolute name (possible if
5174 no home directory exists for the router), the forbid_file option was not
5177 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5178 the existing equality matching.
5180 52. Exim was using ints instead of ino_t variables in some places where it was
5181 dealing with inode numbers.
5183 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5184 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5185 called TMPDIR, and if it finds it is different, it changes its value.
5187 54. The smtp_printf() function is now made available to local_scan() so
5188 additional output lines can be written before returning. There is also an
5189 smtp_fflush() function to enable the detection of a dropped connection.
5190 The variables smtp_input and smtp_batched_input are exported to
5193 55. Changed the default runtime configuration: the message "Unknown user"
5194 has been removed from the ACL, and instead placed on the localuser router,
5195 using the cannot_route_message feature. This means that any verification
5196 failures that generate their own messages won't get overridden. Similarly,
5197 the "Unrouteable address" message that was in the ACL for unverifiable
5198 relay addresses has also been removed.
5200 56. Added hosts_avoid_esmtp to the smtp transport.
5202 57. The exicyclog script was not checking for the esoteric option
5203 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5204 will work only if exicyclog is run under the appropriate euid.
5206 58. Following a discussion on the list, the rules by which Exim recognises line
5207 endings on incoming messages have been changed. The -dropcr and drop_cr
5208 options are now no-ops, retained only for backwards compatibility. The
5209 following line terminators are recognized: LF CRLF CR. However, special
5210 processing applies to CR:
5212 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5213 nor a local message in the state where . is a terminator.
5215 (ii) If a bare CR is encountered in a header line, an extra space is added
5216 after the line terminator so as not to end the header. The reasoning
5217 behind this is that bare CRs in header lines are most likely either
5218 to be mistakes, or people trying to play silly games.
5220 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5221 was being incorrectly given as 18 bytes larger than it should have been.
5222 This is a VOB (very old bug).
5224 60. This may never have affected anything current, but just in case it has:
5225 When the local host is found other than at the start of a list of hosts,
5226 the local host, those with the same MX, and any that follow, are discarded.
5227 When the list in question was part of a longer list of hosts, the following
5228 hosts (not currently being processed) were also being discarded. This no
5229 longer happens. I'm not sure if this situation could ever has previously
5232 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5235 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5236 header. This is something that is recommended in a new Internet Draft, and
5237 is something that is documented as being done by Sendmail. There are two
5238 possible values. For messages generated by the autoreply transport, Exim
5241 Auto-Submitted: auto-replied
5243 whereas for all other generated messages (e.g. bounces) it adds
5245 Auto-Submitted: auto-generated
5247 63. The "personal" condition in filters now includes a test for the
5248 Auto-Submitted: header. If it contains the string "auto-" the message it
5249 not considered personal.
5251 64. Added rcpt_include_affixes as a generic transport option.
5253 65. Added queue_only_override (default true).
5255 66. Added the syslog_duplication option.
5257 67. If what should have been the first header line of a message consisted of
5258 a space followed by a colon, Exim was mis-interpreting it as a header line.
5259 It isn't of course - it is syntactically invalid and should therefore be
5260 treated as the start of the message body. The misbehaviour could have
5261 caused a number of strange effects, including loss of data in subsequent
5262 header lines, and spool format errors.
5264 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5265 client host had authenticated. This control can now be exercised by an ACL
5266 for more flexibility.
5268 69. By default, callouts do not happen when testing with -bh. There is now a
5269 variant, -bhc, which does actually run the callout code, including
5270 consulting and updating the callout cache.
5272 70. Added support for saslauthd authentication, courtesy of Alexander
5275 71. If statvfs() failed on the spool or log directories while checking their
5276 size for availability, Exim confusingly gave the error "space shortage".
5277 Furthermore, in debugging mode it crashed with a floating point exception.
5278 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5279 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5280 serious problem, Exim now writes to the main and panic logs when this
5281 happens, with details of the failure. It then refuses to accept the
5282 incoming message, giving the message "spool directory problem" or "log
5283 directory problem" with a 421 code for SMTP messages.
5285 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5286 0, 1, and 2 exist, because some OS complain for execs without them (see
5287 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5288 descriptors. However, the code omitted to check that the open succeeded,
5289 causing mysterious errors if for some reason the permissions on /dev/null
5290 got screwed. Now Exim writes a message to the main and panic logs, and
5291 bombs out if it can't open /dev/null.
5293 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5294 interact so that it is all more flexible. It is supposed to remain
5295 backwards compatible. Also added extra_local_interfaces.
5297 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5298 to bomb out with an assertion failure - to the client this appears as a
5299 connection drop. This problem occurs in the part of the code that was taken
5300 from the Samba project. Fortunately, the assertion is in a very simple
5301 function, so I have fixed this by reproducing the function inline in the
5302 one place where it is called, and arranging for authentication to fail
5303 instead of killing the process with assert().
5305 75. The SPA client code was not working when the server requested OEM rather
5306 than Unicode encoding.
5308 76. Added code to make require_files with a specific uid setting more usable in
5309 the case where statting the file as root fails - usually a non-root-mounted
5310 NFS file system. When this happens and the failure is EACCES, Exim now
5311 forks a subprocess and does the per-uid checking as the relevant uid.
5313 77. Added process_log_path.
5315 78. If log_file_path was not explicitly set, a setting of check_log_space or
5316 check_log_inodes was ignored.
5318 79. If a space check for the spool or log partitions fails, the incident is now
5319 logged. Of course, in the latter case the data may get lost...
5321 80. Added the %p formatting code to string_format() so that it can be used to
5322 print addresses in debug_print(). Adjusted all the address printing in the
5323 debugging in store.c to use %p rather than %d.
5325 81. There was a concern that a line of code in smtp_in.c could overflow a
5326 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5327 initially expressed, the concern was not well-founded, because trailing
5328 spaces are removed early. However, if the trailing spaces were followed by
5329 a NULL, they did not get removed, so the overflow was possible. Two fixes
5332 (a) I re-wrote the offending code in a cleaner fashion.
5333 (b) If an incoming SMTP command contains a NULL character, it is rejected
5336 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5337 runs initgroups(), so that if the Exim user is in any additional groups,
5338 they will be used during message reception.
5344 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob