1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Solaris build fix for Oracle's LDAP libraries.
8 Bugzilla 1109, patch from Stephen Usher.
10 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
12 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
15 TF/02 Fix a couple more cases where we did not log the error message
16 when unlink() failed. See also change 4.74-TF/03.
18 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
19 lock up or crash if it happened to be inside a call to libc when it
20 got a SIGUSR1 from exiwhat.
22 The SIGUSR1 handler appends the current process status to the process
23 log which is later printed by exiwhat. It used to use the general
24 purpose logging code to do this, but several functions it calls are
27 The new output code in the SIGUSR1 handler is specific to the process
28 log, and simple enough that it's easy to inspect for signal safety.
29 Removing some special cases also simplifies the general logging code.
30 Removing the spurious timestamps from the process log simplifies
33 TF/04 Improved ratelimit ACL condition.
35 The /noupdate option has been deprecated in favour of /readonly which
36 has clearer semantics. The /leaky, /strict, and /readonly update modes
37 are mutually exclusive. The update mode is no longer included in the
38 database key; it just determines when the database is updated. (This
39 means that when you upgrde Exim will forget old rate measurements.)
41 Exim now checks that the per_* options are used with an update mode that
42 makes sense for the current ACL. For example, when Exim is processing a
43 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
44 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
45 must specify per_mail/readonly. If you omit the update mode it defaults to
46 /leaky where that makes sense (as before) or /readonly where required.
48 The /noupdate option is now undocumented but still supported for
49 backwards compatibility. It is equivalent to /readonly except that in
50 ACLs where /readonly is required you may specify /leaky/noupdate or
51 /strict/noupdate which are treated the same as /readonly.
53 A useful new feature is the /count= option. This is a generalization
54 of the per_byte option, so that you can measure the throughput of other
55 aggregate values. For example, the per_byte option is now equivalent
56 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
58 The per_rcpt option has been generalized using the /count= mechanism
59 (though it's more complicated than the per_byte equivalence). When it is
60 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
61 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
62 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
63 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
64 Note that using per_rcpt with a non-readonly update mode in more than
65 one ACL will cause the recipients to be double-counted. (The per_mail
66 and per_byte options don't have this problem.)
68 The handling of very low rates has changed slightly. If the computed rate
69 is less than the event's count (usually one) then this event is the first
70 after a long gap. In this case the rate is set to the same as this event's
71 count, so that the first message of a spam run is counted properly.
73 The major new feature is a mechanism for counting the rate of unique
74 events. The new per_addr option counts the number of different
75 recipients that someone has sent messages to in the last time period. It
76 behaves like per_rcpt if all the recipient addresses are different, but
77 duplicate recipient addresses do not increase the measured rate. Like
78 the /count= option this is a general mechanism, so the per_addr option
79 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
80 example, measure the rate that a client uses different sender addresses
81 with the options per_mail/unique=$sender_address. There are further
82 details in the main documentation.
84 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
86 TF/06 Removed a few PCRE remnants.
88 TF/07 Automatically extract Exim's version number from tags in the git
89 repository when doing development or release builds.
95 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
97 PP/02 Harmonised TLS library version reporting; only show if debugging.
98 Layout now matches that introduced for other libraries in 4.74 PP/03.
100 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
102 PP/04 New "dns_use_edns0" global option.
104 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
107 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
108 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
110 TK/01 Updated PolarSSL code to 0.14.2.
111 Bugzilla 1097. Patch from Andreas Metzler.
113 PP/07 Catch divide-by-zero in ${eval:...}.
116 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
119 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
120 format-string attack -- SECURITY: remote arbitrary code execution.
122 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
123 time unintentionally subject to list matching rules, letting the header
124 cause arbitrary Exim lookups (of items which can occur in lists, *not*
125 arbitrary string expansion). This allowed for information disclosure.
127 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
128 INT_MIN/-1 -- value coerced to INT_MAX.
134 NM/01 Workround for PCRE version dependency in version reporting
137 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
138 This fixes portability to compilers other than gcc, notably
139 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
141 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
142 makefiles for portability to HP-UX and POSIX correctness.
144 PP/01 Permit LOOKUP_foo enabling on the make command-line.
145 Also via indented variable definition in the Makefile.
146 (Debugging by Oliver Heesakkers).
148 PP/02 Restore caching of spamd results with expanded spamd_address.
149 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
151 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
152 Improves build reliability. Fix from: Frank Elsner
154 NM/02 Fix wide character breakage in the rfc2047 coding
155 Fixes bug 1064. Patch from Andrey N. Oktyabrski
157 NM/03 Allow underscore in dnslist lookups
158 Fixes bug 1026. Patch from Graeme Fowler
160 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
161 Code patches from Adam Ciarcinski of NetBSD.
163 NM/04 Fixed exiqgrep to cope with mailq missing size issue
166 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
167 is logged, to avoid truncation. Patch from John Horne.
169 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
170 Patch from Jakob Hirsch.
172 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
173 SQL string expansion failure details.
174 Patch from Andrey Oktyabrski.
176 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
177 Patch from Simon Arlott.
179 PP/09 New lookups functionality failed to compile on old gcc which rejects
180 extern declarations in function scope.
181 Patch from Oliver Fleischmann
183 PP/10 Use sig_atomic_t for flags set from signal handlers.
184 Check getgroups() return and improve debugging.
185 Fixed developed for diagnosis in bug 927 (which turned out to be
188 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
189 Patch from Mark Zealey.
191 PP/12 Bugzilla 1056: Improved spamd server selection.
192 Patch from Mark Zealey.
194 PP/13 Bugzilla 1086: Deal with maildir quota file races.
195 Based on patch from Heiko Schlittermann.
197 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
198 Patch from Uwe Doering, sign-off by Michael Haardt.
200 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
201 variable declaration deep within a block. Bug and patch from
204 PP/15 lookups-Makefile IRIX compatibilty coercion.
206 PP/16 Make DISABLE_DKIM build knob functional.
208 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
209 Patch by Simon Arlott
211 TF/03 Fix valgrind.h portability to C89 compilers that do not support
212 variable argument macros. Our copy now differs from upstream.
218 TF/01 Failure to get a lock on a hints database can have serious
219 consequences so log it to the panic log.
221 TF/02 Log LMTP confirmation messages in the same way as SMTP,
222 controlled using the smtp_confirmation log selector.
224 TF/03 Include the error message when we fail to unlink a spool file.
226 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
227 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
228 for maintaining out-of-tree patches for some time.
230 PP/01 Bugzilla 139: Documentation and portability issues.
231 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
232 Handle per-OS dynamic-module compilation flags.
234 PP/02 Let /dev/null have normal permissions.
235 The 4.73 fixes were a little too stringent and complained about the
236 permissions on /dev/null. Exempt it from some checks.
237 Reported by Andreas M. Kirchwitz.
239 PP/03 Report version information for many libraries, including
240 Exim version information for dynamically loaded libraries. Created
241 version.h, now support a version extension string for distributors
242 who patch heavily. Dynamic module ABI change.
244 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
245 privilege escalation vulnerability whereby the Exim run-time user
246 can cause root to append content of the attacker's choosing to
249 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
252 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
253 If dropping privileges for untrusted macros, we disabled normal logging
254 on the basis that it would fail; for the Exim run-time user, this is not
255 the case, and it resulted in successful deliveries going unlogged.
256 Fixed. Reported by Andreas Metzler.
262 PP/01 Date: & Message-Id: revert to normally being appended to a message,
263 only prepend for the Resent-* case. Fixes regression introduced in
264 Exim 4.70 by NM/22 for Bugzilla 607.
266 PP/02 Include check_rfc2047_length in configure.default because we're seeing
267 increasing numbers of administrators be bitten by this.
269 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
271 PP/03 Bugzilla 994: added openssl_options main configuration option.
273 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
275 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
277 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
279 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
280 without a peer certificate, leading to a segfault because of an
281 assumption that peers always have certificates. Be a little more
282 paranoid. Problem reported by Martin Tscholak.
284 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
285 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
286 NB: ClamAV planning to remove STREAM in "middle of 2010".
287 CL also introduces -bmalware, various -d+acl logging additions and
288 more caution in buffer sizes.
290 PP/09 Implemented reverse_ip expansion operator.
292 PP/10 Bugzilla 937: provide a "debug" ACL control.
294 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
296 PP/12 Bugzilla 973: Implement --version.
298 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
300 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
302 PP/15 Bugzilla 816: support multiple condition rules on Routers.
304 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
305 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
306 ignore trailing whitespace.
308 JJ/02 prevent non-panic DKIM error from being sent to paniclog
310 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
313 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
314 Notification from Dr Andrew Aitchison.
316 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
317 ExtendedDetectionInfo response format.
318 Notification from John Horne.
320 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
323 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
324 XSL and documented dependency on system catalogs, with examples of how
327 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
330 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
331 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
332 configuration file which is writeable by the Exim user or group.
334 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
335 of configuration files to cover files specified with the -C option if
336 they are going to be used with root privileges, not just the default
339 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
340 option (effectively making it always true).
342 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
343 files to be used while preserving root privileges.
345 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
346 that rogue child processes cannot use them.
348 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
349 run-time user, instead of root.
351 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
352 Exim run-time user without dropping privileges.
354 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
355 result string, instead of calling string_vformat() twice with the same
358 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
359 for other users. Others should always drop root privileges if they use
360 -C on the command line, even for a whitelisted configure file.
362 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
364 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
370 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
371 $data_path, and $header_path variables; fixed documentation bugs and
374 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
375 exipick to access non-standard spools, including the "frozen" queue
378 NM/01 Bugzilla 965: Support mysql stored procedures.
379 Patch from Alain Williams
381 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
383 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
384 Patch from Andreas Metzler
386 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
387 Patch from Kirill Miazine
389 NM/05 Bugzilla 671: Added umask to procmail example.
391 JJ/03 installed exipick 20100323.0, fixing doc bug
393 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
394 directory. Notification and patch from Dan Rosenberg.
396 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
398 TK/02 Improve log output when DKIM signing operation fails.
400 MH/01 Treat the transport option dkim_domain as a colon separated
401 list, not as a single string, and sign the message with each element,
402 omitting multiple occurences of the same signer.
404 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
405 Bugzilla 985, 986. Patch by Simon Arlott
407 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
408 Patch by Simon Arlott
410 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
411 MBX locking. Notification from Dan Rosenberg.
417 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
419 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
421 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
423 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
425 NM/04 Bugzilla 915: Flag broken perl installation during build.
431 TK/01 Added patch by Johannes Berg that expands the main option
432 "spamd_address" if it starts with a dollar sign.
434 TK/02 Write list of recipients to X-Envelope-Sender header when building
435 the mbox-format spool file for content scanning (suggested by Jakob
438 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
439 (http://www.dcc-servers.net/) support via dccifd. Activated by
440 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
442 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
443 by Mark Daniel Reidel <mr@df.eu>.
445 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
446 When building exim an external PCRE library is now needed -
447 PCRE is a system library on the majority of modern systems.
448 See entry on PCRE_LIBS in EDITME file.
450 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
451 conversation. Added nologin parameter to request.
452 Patch contributed by Kirill Miazine.
454 TF/01 Do not log submission mode rewrites if they do not change the address.
456 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
458 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
459 log files in place. Contributed by Roberto Lima.
461 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
463 TF/03 Bugzilla 615: When checking the local_parts router precondition
464 after a local_part_suffix or local_part_prefix option, Exim now
465 does not use the address's named list lookup cache, since this
466 contains cached lookups for the whole local part.
468 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
469 Robert Millan. Documentation is in experimental-spec.txt.
471 TF/04 Bugzilla 668: Fix parallel build (make -j).
473 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
475 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
476 Patch provided by Jan Srzednicki.
478 TF/05 Leading white space used to be stripped from $spam_report which
479 wrecked the formatting. Now it is preserved.
481 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
482 that they are available at delivery time.
484 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
486 TF/08 TLS error reporting now respects the incoming_interface and
487 incoming_port log selectors.
489 TF/09 Produce a more useful error message if an SMTP transport's hosts
490 setting expands to an empty string.
492 NM/06 Bugzilla 744: EXPN did not work under TLS.
493 Patch provided by Phil Pennock.
495 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
496 Patch provided by Richard Godbee.
498 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
499 acl_smtp_notquit, added index entry.
501 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
502 Patch provided by Eugene Bujak.
504 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
505 accept(). Patch provided by Maxim Dounin.
507 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
508 Patch provided by Phil Pennock.
510 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
512 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
513 Patch provided by Brad "anomie" Jorsch.
515 NM/14 Bugzilla 622: Added timeout setting to transport filter.
516 Patch provided by Dean Brooks.
518 TK/05 Add native DKIM support (does not depend on external libraries).
520 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
521 Patch provided by Graeme Fowler.
523 NM/16 Bugzilla 851: Documentation example syntax fix.
525 NM/17 Changed NOTICE file to remove references to embedded PCRE.
527 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
530 NM/19 Bugzilla 745: TLS version reporting.
531 Patch provided by Phil Pennock.
533 NM/20 Bugzilla 167: bool: condition support.
534 Patch provided by Phil Pennock.
536 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
537 clients. Patch provided by Phil Pennock.
539 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
540 Patch provided by Brad "anomie" Jorsch.
542 NM/23 Bugzilla 687: Fix misparses in eximstats.
543 Patch provided by Heiko Schlittermann.
545 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
546 Patch provided by Heiko Schlittermann.
548 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
549 plus update to original patch.
551 NM/26 Bugzilla 799: Documentation correction for ratelimit.
553 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
554 Patch provided by David Brownlee.
556 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
558 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
560 NM/30 Bugzilla 888: TLS documentation bugfixes.
562 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
564 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
565 Unlike the original bugzilla I have changed all shell scripts in src tree.
567 NM/33 Bugzilla 898: Transport filter timeout fix.
568 Patch by Todd Rinaldo.
570 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
571 Patch by Serge Demonchaux.
573 NM/35 Bugzilla 39: Base64 decode bug fixes.
574 Patch by Jakob Hirsch.
576 NM/36 Bugzilla 909: Correct connect() call in dcc code.
578 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
580 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
582 NM/39 Bugzilla 911: Fixed MakeLinks build script.
588 TK/01 Add preliminary DKIM support. Currently requires a forked version of
589 ALT-N's libdkim that I have put here:
590 http://duncanthrax.net/exim-experimental/
592 Note to Michael Haardt: I had to rename some vars in sieve.c. They
593 were called 'true' and it seems that C99 defines that as a reserved
594 keyword to be used with 'bool' variable types. That means you could
595 not include C99-style headers which use bools without triggering
596 build errors in sieve.c.
598 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
599 as mailq or other aliases. Changed the --help handling significantly
600 to do whats expected. exim_usage() emits usage/help information.
602 SC/01 Added the -bylocaldomain option to eximstats.
604 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
606 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
608 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
614 PH/01 Another patch from the Sieve maintainer.
616 PH/02 When an IPv6 address is converted to a string for single-key lookup
617 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
618 dots are used instead of colons so that keys in lsearch files need not
619 contain colons. This was done some time before quoting was made available
620 in lsearch files. However, iplsearch files do require colons in IPv6 keys
621 (notated using the quote facility) so as to distinguish them from IPv4
622 keys. This meant that lookups for IP addresses in host lists did not work
623 for iplsearch lookups.
625 This has been fixed by arranging for IPv6 addresses to be expressed with
626 colons if the lookup type is iplsearch. This is not incompatible, because
627 previously such lookups could never work.
629 The situation is now rather anomolous, since one *can* have colons in
630 ordinary lsearch keys. However, making the change in all cases is
631 incompatible and would probably break a number of configurations.
633 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
636 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
637 conversion specification without a maximum field width, thereby enabling
638 a rogue spamd server to cause a buffer overflow. While nobody in their
639 right mind would setup Exim to query an untrusted spamd server, an
640 attacker that gains access to a server running spamd could potentially
641 exploit this vulnerability to run arbitrary code as the Exim user.
643 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
644 $primary_hostname instead of what libspf2 thinks the hosts name is.
646 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
647 a directory entry by the name of the lookup key. Previously, if a
648 symlink pointed to a non-existing file or a file in a directory that
649 Exim lacked permissions to read, a lookup for a key matching that
650 symlink would fail. Now it is enough that a matching directory entry
651 exists, symlink or not. (Bugzilla 503.)
653 PH/03 The body_linecount and body_zerocount variables are now exported in the
656 PH/04 Added the $dnslist_matched variable.
658 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
659 This means they are set thereafter only if the connection becomes
662 PH/06 Added the client_condition to authenticators so that some can be skipped
663 by clients under certain conditions.
665 PH/07 The error message for a badly-placed control=no_multiline_responses left
666 "_responses" off the end of the name.
668 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
670 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
671 (without spaces) instead of just copying the configuration text.
673 PH/10 Added the /noupdate option to the ratelimit ACL condition.
675 PH/11 Added $max_received_linelength.
677 PH/12 Added +ignore_defer and +include_defer to host lists.
679 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
680 way in which PCRE > 7.0 is built.
682 PH/14 Implemented queue_only_load_latch.
684 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
685 MAIL command. The effect was to mangle the value on 64-bit systems.
687 PH/16 Another patch from the Sieve maintainer.
689 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
691 PH/18 If a system quota error occurred while trying to create the file for
692 a maildir delivery, the message "Mailbox is full" was not appended to the
693 bounce if the delivery eventually timed out. Change 4.67/27 below applied
694 only to a quota excession during the actual writing of the file.
696 PH/19 It seems that peer DN values may contain newlines (and other non-printing
697 characters?) which causes problems in log lines. The DN values are now
698 passed through string_printing() before being added to log lines.
700 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
701 and InterBase are left for another time.)
703 PH/21 Added message_body_newlines option.
705 PH/22 Guard against possible overflow in moan_check_errorcopy().
707 PH/23 POSIX allows open() to be a macro; guard against that.
709 PH/24 If the recipient of an error message contained an @ in the local part
710 (suitably quoted, of course), incorrect values were put in $domain and
711 $local_part during the evaluation of errors_copy.
717 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
718 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
721 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
722 issue a MAIL command.
724 PH/02 In an ACL statement such as
726 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
728 if a client was not listed at all, or was listed with a value other than
729 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
730 the condition was not true (as it should be), so access was not denied.
731 The bug was that the ! inversion was incorrectly passed on to the second
732 item. This has been fixed.
734 PH/03 Added additional dnslists conditions == and =& which are different from
735 = and & when the dns lookup returns more than one IP address.
737 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
738 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
740 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
741 FSYNC, which compiles an option called disable_fsync that allows for
742 bypassing fsync(). The documentation is heavily laced with warnings.
744 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
746 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
747 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
748 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
749 including adding "make clean"; (3) Added -fPIC when compiling the test
750 dynamically loaded module, to get rid of a warning.
752 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
753 message fails, move_frozen_messages = true and ignore_bounce_errors_after
754 = 0s. The bug is otherwise harmless.
756 PH/07 There was a bug in the dovecot authenticator such that the value of
757 $auth1 could be overwritten, and so not correctly preserved, after a
758 successful authentication. This usually meant that the value preserved by
759 the server_setid option was incorrect.
761 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
763 PH/09 Installed PCRE release 7.0.
765 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
766 run for batched SMTP input. It is now run at the start of every message
767 in the batch. While fixing this I discovered that the process information
768 (output by running exiwhat) was not always getting set for -bs and -bS
769 input. This is fixed, and it now also says "batched" for BSMTP.
771 PH/11 Added control=no_pipelining.
773 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
774 patch, slightly modified), and move the expansion of helo_data till after
775 the connection is made in the smtp transport (so it can use these
778 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
780 PH/14 Added log_selector = +pid.
782 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
784 PH/16 Add ${if forany and ${if forall.
786 PH/17 Added dsn_from option to vary the From: line in DSNs.
788 PH/18 Flush SMTP output before performing a callout, unless control =
789 no_callout_flush is set.
791 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
792 was true (the default) a successful delivery failed to delete the retry
793 item, thus causing premature timeout of the address. The bug is now
796 PH/20 Added hosts_avoid_pipelining to the smtp transport.
798 PH/21 Long custom messages for fakedefer and fakereject are now split up
799 into multiline reponses in the same way that messages for "deny" and
800 other ACL rejections are.
802 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
803 with slight modification.
805 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
806 draft, changing the syntax and factoring some duplicate code".
808 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
809 for deliveries of the second and subsequent messages over the same SMTP
812 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
813 ${reduce, with only minor "tidies".
815 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
817 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
818 expansion side effects.
820 PH/27 When a message times out after an over-quota error from an Exim-imposed
821 quota, the bounce message says "mailbox is full". This message was not
822 being given when it was a system quota that was exceeded. It now should
825 MH/03 Made $recipients available in local_scan(). local_scan() already has
826 better access to the recipient list through recipients_list[], but
827 $recipients can be useful in postmaster-provided expansion strings.
829 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
830 in the case of a MAIL command with additional options following the
831 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
832 were accidentally chopped off.
834 PH/29 SMTP synchronization checks are implemented when a command is read -
835 there is a check that no more input is waiting when there shouldn't be
836 any. However, for some commands, a delay in an ACL can mean that it is
837 some time before the response is written. In this time, more input might
838 arrive, invalidly. So now there are extra checks after an ACL has run for
839 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
840 pipelining has not been advertised.
842 PH/30 MH's patch to allow iscntrl() characters to be list separators.
844 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
845 returned in the SMTP response when smtp_return_error_details was false.
848 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
849 instead of the C I/O that was originally supplied, because problems were
852 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
853 Exim which did not show up earlier: it was assuming that a call to
854 SSL_CTX_set_info_callback() might give an error value. In fact, there is
855 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
856 was a macro that became an assignment, so it seemed to work. This has
857 changed to a proper function call with a void return, hence the compile
858 error. Exim's code has been fixed.
860 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
863 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
865 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
867 PH/37 If a message is not accepted after it has had an id assigned (e.g.
868 because it turns out to be too big or there is a timeout) there is no
869 "Completed" line in the log. When some messages of this type were
870 selected by exigrep, they were listed as "not completed". Others were
871 picked up by some special patterns. I have improved the selection
872 criteria to be more general.
874 PH/38 The host_find_failed option in the manualroute router can now be set
875 to "ignore", to completely ignore a host whose IP address cannot be
876 found. If all hosts are ignored, the behaviour is controlled by the new
877 host_all_ignored option.
879 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
880 homing or because of multiple MX records with /mx) generated more than
881 one IP address, and the following item turned out to be the local host,
882 all the secondary addresses of the first item were incorrectly removed
883 from the list, along with the local host and any following hosts (which
884 is what is supposed to happen).
886 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
887 whoever called Exim into the -H file. In the case of the daemon it was
888 behaving confusingly. When first started, it used values for whoever
889 started the daemon, but after a SIGHUP it used the Exim user (because it
890 calls itself on a restart). I have changed the code so that it now always
893 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
894 message are rejected with the same error (e.g. no authentication or bad
895 sender address), and a DATA command is nevertheless sent (as can happen
896 with PIPELINING or a stupid MUA), the error message that was given to the
897 RCPT commands is included in the rejection of the DATA command. This is
898 intended to be helpful for MUAs that show only the final error to their
901 PH/42 Another patch from the Sieve maintainer.
903 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
904 Eximstats - Fixed some broken HTML links and added missing column headers
906 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
907 columns for Rejects, Temp Rejects, Ham, and Spam rows.
909 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
911 PH/43 Yet another patch from the Sieve maintainer.
913 PH/44 I found a way to check for a TCP/IP connection going away before sending
914 the response to the final '.' that terminates a message, but only in the
915 case where the client has not sent further data following the '.'
916 (unfortunately, this is allowed). However, in many cases there won't be
917 any further data because there won't be any more messages to send. A call
918 to select() can be used: if it shows that the input is "ready", there is
919 either input waiting, or the socket has been closed. An attempt to read
920 the next input character can distinguish the two cases. Previously, Exim
921 would have sent an OK response which the client would never have see.
922 This could lead to message repetition. This fix should cure that, at
923 least in a lot of common cases.
925 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
926 advertised in response to EHLO.
932 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
933 fixed by 4.65/MH/01 (is this a record?) are fixed:
935 (i) An empty string was always treated as zero by the numeric comparison
936 operators. This behaviour has been restored.
938 (ii) It is documented that the numeric comparison operators always treat
939 their arguments as decimal numbers. This was broken in that numbers
940 starting with 0 were being interpreted as octal.
942 While fixing these problems I realized that there was another issue that
943 hadn't been noticed. Values of message_size_limit (both the global option
944 and the transport option) were treated as octal if they started with 0.
945 The documentation was vague. These values are now always treated as
946 decimal, and I will make that clear in the documentation.
952 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
953 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
956 MH/01 Don't check that the operands of numeric comparison operators are
957 integers when their expansion is in "skipping" mode (fixes bug
958 introduced by 4.64-PH/07).
960 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
961 child addresses, Exim now panics and dies. Previously, because the count
962 is held in a short int, deliveries were likely to be lost. As such a
963 large number of recipients for a single message is ridiculous
964 (performance will be very, very poor), I have chosen to impose a limit
965 rather than extend the field.
971 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
972 leftover -K file (the existence of which was triggered by #402).
973 While we were at it, introduced process PID as part of the -K
974 filename. This should rule out race conditions when creating
977 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
978 processing considerably. Previous code took too long for large mails,
979 triggering a timeout which in turn triggers #401.
981 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
982 in the DK code in transports.c. sendfile() is not really portable,
983 hence the _LINUX specificness.
985 TF/01 In the add_headers option to the mail command in an Exim filter,
986 there was a bug that Exim would claim a syntax error in any
987 header after the first one which had an odd number of characters
990 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
991 callout verification, Exim cached a "reject" for the entire domain. This
992 is correct for most verifications, but it is not correct for a recipient
993 verification with use_sender or use_postmaster set, because in that case
994 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
995 case of MAIL FROM:<> rejection from other early rejections (e.g.
996 rejection of HELO). When verifying a recipient using a non-null MAIL
997 address, the cache is ignored if it shows MAIL FROM:<> rejection.
998 Whatever the result of the callout, the value of the domain cache is
999 left unchanged (for any other kind of callout, getting as far as trying
1000 RCPT means that the domain itself is ok).
1002 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1005 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1006 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1007 session, the daemon ignores ECONNECT errors and logs others; it now
1008 ignores EPIPE as well.
1010 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1011 (quoted-printable decoding).
1013 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1014 later the small subsequent patch to fix an introduced bug.
1016 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1018 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1020 PH/08 An error is now given if message_size_limit is specified negative.
1022 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1023 to be given (somewhat) arbitrary names.
1025 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1028 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1029 miscellaneous code fixes
1031 PH/10 Added the log_reject_target ACL modifier to specify where to log
1034 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1035 hostname. This is wrong, because it relates to the incoming message (and
1036 probably the interface on which it is arriving) and not to the outgoing
1037 callout (which could be using a different interface). This has been
1038 changed to use the value of the helo_data option from the smtp transport
1039 instead - this is what is used when a message is actually being sent. If
1040 there is no remote transport (possible with a router that sets up host
1041 addresses), $smtp_active_hostname is used.
1043 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1044 tweaks were necessary in order to get it to work (see also 21 below):
1045 (a) The code assumed that strncpy() returns a negative number on buffer
1046 overflow, which isn't the case. Replaced with Exim's string_format()
1048 (b) There were several signed/unsigned issues. I just did the minimum
1049 hacking in of casts. There is scope for a larger refactoring.
1050 (c) The code used strcasecmp() which is not a standard C function.
1051 Replaced with Exim's strcmpic() function.
1052 (d) The code set only $1; it now sets $auth1 as well.
1053 (e) A simple test gave the error "authentication client didn't specify
1054 service in request". It would seem that Dovecot has changed its
1055 interface. Fortunately there's a specification; I followed it and
1056 changed what the client sends and it appears to be working now.
1058 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1061 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1062 address A is aliased to B and C, where B exists and C does not. Without
1063 -v the output is "A verified" because verification stops after a
1064 successful redirection if more than one address is generated. However,
1065 with -v the child addresses are also verified. Exim was outputting "A
1066 failed to verify" and then showing the successful verification for C,
1067 with its parentage. It now outputs "B failed to verify", showing B's
1068 parentage before showing the successful verification of C.
1070 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1071 look up a TXT record in a specific list after matching in a combined
1074 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1075 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1076 they consult the DNS. I had assumed they would set it the way they
1077 wanted; and indeed my experiments on Linux seem to show that in some
1078 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1079 To be on the safe side, however, I have now made the interface to
1080 host_find_byname() similar to host_find_bydns(), with an argument
1081 containing the DNS resolver options. The host_find_byname() function now
1082 sets these options at its start, just as host_find_bydns() does. The smtp
1083 transport options dns_qualify_single and dns_search_parents are passed to
1084 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1085 of host_find_byname() use the default settings of RES_DEFNAMES
1086 (qualify_single) but not RES_DNSRCH (search_parents).
1088 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1089 spool_read_header() do less string testing, by means of a preliminary
1090 switch on the second character of optional "-foo" lines. (This is
1091 overdue, caused by the large number of possibilities that now exist.
1092 Originally there were few.) While I was there, I also converted the
1093 str(n)cmp tests so they don't re-test the leading "-" and the first
1094 character, in the hope this might squeeze out yet more improvement.
1096 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1097 flag allowing group syntax was set by the header_syntax check but not
1098 turned off, possible causing trouble later; (2) The flag was not being
1099 set at all for the header_verify test, causing "group"-style headers to
1100 be rejected. I have now set it in this case, and also caused header_
1101 verify to ignore an empty address taken from a group. While doing this, I
1102 came across some other cases where the code for allowing group syntax
1103 while scanning a header line wasn't quite right (mostly, not resetting
1104 the flag correctly in the right place). These bugs could have caused
1105 trouble for malformed header lines. I hope it is now all correct.
1107 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1108 with the "reply" argument non-NULL. The code, however (which originally
1109 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1110 but it didn't always do it. This confused somebody who was copying the
1111 code for some other use. I have removed all the tests.
1113 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1114 feature that was used to support insecure browsers during the U.S. crypto
1115 embargo. It requires special client support, and Exim is probably the
1116 only MTA that supported it -- and would never use it because real RSA is
1117 always available. This code has been removed, because it had the bad
1118 effect of slowing Exim down by computing (never used) parameters for the
1119 RSA_EXPORT functionality.
1121 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1122 authenticator to fail if there's a tab character in the incoming data
1123 (there should never be unless someone is messing about, as it's supposed
1124 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1125 if the connection is using TLS or if the remote IP is the same as the
1126 local IP, and the "valid-client-cert option" if a client certificate has
1129 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1130 authenticators. This can be used for authorization after authentication
1131 succeeds. (In the case of plaintext, it servers for both authentication
1134 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1135 if any retry times were supplied.
1137 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1138 connection, where there is no client IP address to check. In this
1139 situation, the verify now always succeeds.
1141 PH/25 Applied John Jetmore's -Mset patch.
1143 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1145 PH/27 In a string expansion for a processed (not raw) header when multiple
1146 headers of the same name were present, leading whitespace was being
1147 removed from all of them, but trailing whitespace was being removed only
1148 from the last one. Now trailing whitespace is removed from each header
1149 before concatenation. Completely empty headers in a concatenation (as
1150 before) are ignored.
1152 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1153 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1155 PH/29 [Removed. This was a change that I later backed out, and forgot to
1156 correct the ChangeLog entry (that I had efficiently created) before
1157 committing the later change.]
1159 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1160 address errors (4xx response to RCPT) over the same connection as other
1161 messages routed to the same hosts. Such deliveries are always "forced",
1162 so retry times are not inspected. This resulted in far too many retries
1163 for the affected addresses. The effect occurred only when there were more
1164 hosts than the hosts_max_try setting in the smtp transport when it had
1165 the 4xx errors. Those hosts that it had tried were not added to the list
1166 of hosts for which the message was waiting, so if all were tried, there
1167 was no problem. Two fixes have been applied:
1169 (i) If there are any address or message errors in an SMTP delivery, none
1170 of the hosts (tried or untried) are now added to the list of hosts
1171 for which the message is waiting, so the message should not be a
1172 candidate for sending over the same connection that was used for a
1173 successful delivery of some other message. This seems entirely
1174 reasonable: after all the message is NOT "waiting for some host".
1175 This is so "obvious" that I'm not sure why it wasn't done
1176 previously. Hope I haven't missed anything, but it can't do any
1177 harm, as the worst effect is to miss an optimization.
1179 (ii) If, despite (i), such a delivery is accidentally attempted, the
1180 routing retry time is respected, so at least it doesn't keep
1181 hammering the server.
1183 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1184 in ${readsocket because some servers need this prod.
1186 PH/32 Added some extra debug output when updating a wait-xxx database.
1188 PH/33 The hint "could be header name not terminated by colon", which has been
1189 given for certain expansion errors for a long time, was not being given
1190 for the ${if def:h_colon_omitted{... case.
1192 PH/34 The spec says: "With one important exception, whenever a domain list is
1193 being scanned, $domain contains the subject domain." There was at least
1194 one case where this was not true.
1196 PH/35 The error "getsockname() failed: connection reset by peer" was being
1197 written to the panic log as well as the main log, but it isn't really
1198 panic-worthy as it just means the connection died rather early on. I have
1199 removed the panic log writing for the ECONNRESET error when getsockname()
1202 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1203 runs only) independently of the message's sender address. This meant
1204 that, if the 4xx error was in fact related to the sender, a different
1205 message to the same recipient with a different sender could confuse
1206 things. In particualar, this can happen when sending to a greylisting
1207 server, but other circumstances could also provoke similar problems.
1208 I have changed the default so that the retry time for these errors is now
1209 based a combination of the sender and recipient addresses. This change
1210 can be overridden by setting address_retry_include_sender=false in the
1213 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1214 remote server are returned as part of bounce messages. This was not
1215 happening for LMTP over a pipe (the lmtp transport), but now it is the
1216 same for both kinds of LMTP.
1218 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1219 in header lines that were in fact CNAMEs. This is no longer the case.
1221 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1222 and queue runs started by the daemon processed all messages. This has
1223 been fixed so that -R and -S can now usefully be given with -q<time>.
1225 PH/40 Import PCRE release 6.7 (fixes some bugs).
1227 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1229 PH/42 Give an error if -q is specified more than once.
1231 PH/43 Renamed the variables $interface_address and $interface_port as
1232 $received_ip_address and $received_port, to make it clear that these
1233 values apply to message reception, and not to the outgoing interface when
1234 a message is delivered. (The old names remain recognized, of course.)
1236 PH/44 There was no timeout on the connect() call when using a Unix domain
1237 socket in the ${readsocket expansion. There now is.
1239 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1240 be meaningful with "accept".
1242 SC/01 Eximstats V1.43
1243 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1245 SC/02 Eximstats V1.44
1246 Use a glob alias rather than an array ref in the generated
1247 parser. This improves both readability and performance.
1249 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1250 Collect SpamAssassin and rejection statistics.
1251 Don't display local sender or destination tables unless
1252 there is data to show.
1253 Added average volumes into the top table text output.
1255 SC/04 Eximstats V1.46
1256 Collect data on the number of addresses (recipients)
1257 as well as the number of messages.
1259 SC/05 Eximstats V1.47
1260 Added 'Message too big' to the list of mail rejection
1261 reasons (thanks to Marco Gaiarin).
1263 SC/06 Eximstats V1.48
1264 Mainlog lines which have GMT offsets and are too short to
1265 have a flag are now skipped.
1267 SC/07 Eximstats V1.49 (Alain Williams)
1268 Added the -emptyok flag.
1270 SC/08 Eximstats V1.50
1271 Fixes for obtaining the IP address from reject messages.
1273 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1274 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1275 whitesspace changes from 4.64-PH/27
1277 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1280 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1281 are found, allow negative numbers in numeric criteria)
1283 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1285 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1286 to match changes made in 4.64-PH/43
1288 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1290 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1291 "domain" is actually an IP address, and force a failure. This locks out
1292 those revolvers/nameservers that support "A-for-A" lookups, in
1293 contravention of the specifications.
1295 PH/48 When a host name was looked up from an IP address, and the subsequent
1296 forward lookup of the name timed out, the host name was left in
1297 $sender_host_name, contrary to the specification.
1299 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1300 restricted to single-key lookups, Exim was not diagnosing an error if
1301 * or *@ was used with a query-style lookup.
1303 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1305 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1306 long ago, but noone interested enough thought of it. Let's just say that
1307 the "1.1" means that there are some new functions that weren't there at
1308 some point in the past.
1310 PH/51 Error processing for expansion failure of helo_data from an smtp
1311 transport during callout processing was broken.
1313 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1314 tested/used via the -bh/-bhc/-bs options.
1316 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1317 bug, fixed in subsequent PCRE releases).
1319 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1320 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1322 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1328 SC/01 Use a glob alias rather than an array ref in eximstats generated
1329 parser. This improves both readability and performance.
1331 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1332 Don't display local sender or destination tables in eximstats unless
1333 there is data to show.
1334 Added average volumes into the eximstats top table text output.
1336 SC/03 Collect data on the number of addresses (recipients) as well
1337 as the number of messages in eximstats.
1339 TF/01 Correct an error in the documentation for the redirect router. Exim
1340 does (usually) call initgroups() when daemonizing.
1342 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1343 with consistent privilege compared to when running as a daemon.
1345 TF/03 Note in the spec that $authenticated_id is not set for local
1346 submissions from trusted users.
1348 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1349 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1351 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1352 by adding some example configuration directives to the default
1353 configuration file. A little bit of work is required to uncomment the
1354 directives and define how usernames and passwords are checked, but
1355 there is now a framework to start from.
1357 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1358 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1359 without this. I don't know how relevant this is to other LDAP libraries.
1361 PH/02 Add the verb name to the "unknown ACL verb" error.
1363 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1365 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1367 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1368 directory not expanded when it should be if an expanded home directory
1369 was set for the address (which is overridden by the transport).
1371 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1374 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1375 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1376 because it is too late at that time, and has no effect.
1378 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1379 security issue with \' (bugzilla #107). I could not use the
1380 PQescapeStringConn() function, because it needs a PGconn value as one of
1383 PH/08 When testing addresses using -bt, indicate those final addresses that
1384 are duplicates that would not cause an additional delivery. At least one
1385 person was confused, thinking that -bt output corresponded to deliveries.
1386 (Suppressing duplicates isn't a good idea as you lose the information
1387 about possibly different redirections that led to the duplicates.)
1389 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1390 systems where poll() doesn't work, in particular OS X.
1392 PH/10 Added more information to debugging output for retry time not reached.
1394 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1395 operations in malware.c.
1397 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1400 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1401 syslog_duplication set false, log lines that would normally be written
1402 both the the main log and to the reject log were not written to syslog at
1405 PH/14 In the default configuration, change the use of "message" in ACL warn
1406 statements to "add_header".
1408 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1409 not followed by a command (e.g. "seen endif").
1411 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1412 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1415 PH/17 Added extra conditions to the default value of delay_warning_condition
1419 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1420 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1421 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1424 The Auto-Submitted: and various List- headers are standardised, whereas I
1425 don't think Precedence: ever was.
1427 PH/18 Refactored debugging code in route_finduser() to show more information,
1428 in particular, the error code if getpwnam() issues one.
1430 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1431 This is apparently needed in addition to the PH/07 change above to avoid
1432 any possible encoding problems.
1434 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1435 but not after initializing Perl.
1437 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1438 output them only if debugging. By default they are written stderr,
1439 apparently, which is not desirable.
1441 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1444 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1447 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1449 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1450 authenticated or an ident call has been made. Suppress the default
1451 values for $authenticated_id and $authenticated_sender (but permit -oMai
1452 and -oMas) when testing with -bh.
1454 PH/24 Re-jigged the order of the tests in the default configuration so that the
1455 tests for valid domains and recipients precede the DNS black list and CSA
1456 tests, on the grounds that those ones are more expensive.
1458 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1459 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1460 This bug exists in every version of Exim that I still have, right back to
1463 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1464 However, an attempt to turn on case-sensitivity in a regex key by
1465 including (?-i) didn't work because the subject string was already
1466 lowercased, and the effects were non-intuitive. It turns out that a
1467 one-line patch can be used to allow (?-i) to work as expected.
1473 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1474 other effects) broke the use of negated acl sub-conditions.
1476 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1479 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1480 "Deny" causes Exim to reject the incoming connection with a 554 error.
1481 Unfortunately, if there is a major crisis, such as a disk failure,
1482 tcp-wrappers gives "deny", whereas what one would like would be some
1483 kind of temporary error. A kludge has been added to help with this.
1484 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1485 554 error is used if errno is still zero or contains ENOENT (which occurs
1486 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1489 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1491 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1492 errors. Otherwise a message that provokes a temporary error (when other
1493 messages do not) can cause a whole host to time out.
1495 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1496 addresses were routed directly to files or pipes from a redirect router.
1497 File deliveries just didn't batch; pipe deliveries might have suffered
1500 PH/06 A failure to get a lock for a hints database would erroneously always say
1501 "Failed to get write lock", even when it was really a read lock.
1503 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1504 of 0600. This has been changed to use the value of the lockfile_mode
1505 option (which defaults to 0600).
1507 PH/08 Applied small patch from the Sieve maintainer.
1509 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1510 folder from quota calculations, a direct delivery into this folder messed
1511 up the contents of the maildirsize file. This was because the regex was
1512 used only to exclude .Trash (or whatever) when the size of the mailbox
1513 was calculated. There was no check that a delivery was happening into an
1514 excluded directory. This bug has been fixed by ignoring all quota
1515 processing for deliveries into excluded directories.
1517 PH/10 Added the maildirfolder_create_regex option to appendfile.
1523 PH/01 The code for finding all the local interface addresses on a FreeBSD
1524 system running IPv6 was broken. This may well have applied to all BSD
1525 systems, as well as to others that have similar system calls. The broken
1526 code found IPv4 interfaces correctly, but gave incorrect values for the
1527 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1528 that it would not match correctly against @[] and not recognize the IPv6
1531 PH/02 The ipliteral router was not recognizing addresses of the form user@
1532 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1534 PH/03 Added disable_ipv6.
1536 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1537 decoded form, because it is most often used to construct To: headers
1538 lines in autoreplies, and the decoded form may well be syntactically
1539 invalid. However, $reply_address has leading white space removed, and all
1540 newlines turned into spaces so that the autoreply transport does not
1543 PH/05 If group was specified without a user on a router, and no group or user
1544 was specified on a transport, the group from the router was ignored.
1546 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1547 for visible compile-time settings that can be used to change these
1548 numbers, for those that want even more. Backwards compatibility with old
1549 spool files has been maintained. However, going back to a previous Exim
1550 release will lost any variables that are in spool files.
1552 PH/07 Two small changes when running in the test harness: increase delay when
1553 passing a TCP/IP connection to a new process, in case the original
1554 process has to generate a bounce, and remove special handling of
1555 127.0.0.2 (sic), which is no longer necessary.
1557 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1558 be the same on different OS.
1560 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1563 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1564 whether --show-vars was specified or not
1566 JJ/02 exipick: Added support for new ACL variable spool format introduced
1569 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1570 syntactically invalid From: or Reply-to: line, and a filter used this to
1571 generate an autoreply, and therefore failed to obtain an address for the
1572 autoreply, Exim could try to deliver to a non-existent relative file
1573 name, causing unrelated and misleading errors. What now happens is that
1574 it logs this as a hard delivery error, but does not attempt to create a
1577 PH/11 The exinext utility has a -C option for testing purposes, but although
1578 the given file was scanned by exinext itself; it wasn't being passed on
1579 when Exim was called.
1581 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1582 an end-of-file indication when reading a command response.
1584 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1585 compiled. In many other places in Exim, IPv6 addresses are always
1586 recognized, so I have changed this. It also means that IPv4 domain
1587 literals of the form [IPV4:n.n.n.n] are now always recognized.
1589 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1590 used if the router is not running as root, for example, when verifying at
1591 ACL time, or when using -bh. The debugging output from this situation was
1592 non-existent - all you got was a failure to exec. I have made two
1595 (a) Failures to set uid/gid, the current directory, or a process leader
1596 in a subprocess such as that created by queryprogram now generate
1597 suitable debugging ouput when -d is set.
1599 (b) The queryprogram router detects when it is not running as root,
1600 outputs suitable debugging information if -d is set, and then runs
1601 the subprocess without attempting to change uid/gid.
1603 PH/15 Minor change to Makefile for building test_host (undocumented testing
1606 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1607 additional section of a DNS packet that returns MX or SRV records.
1608 Instead, it always explicitly searches for A/AAAA records. This avoids
1609 major problems that occur when a DNS server includes only records of one
1610 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1611 fixed another bug: if SRV records were looked up and the corresponding
1612 address records were *not* found in the additional section, the port
1613 values from the SRV records were lost.
1615 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1616 using the correct key (the original address) when searching the retry
1617 rules in order to find which one to use for generating the retry hint.
1619 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1620 adding the default one. Similarly, if it contains a Reply-To: header, the
1621 errors_reply_to option, if set, is not used.
1623 PH/19 When calculating a retry time, Exim used to measure the "time since
1624 failure" by looking at the "first failed" field in the retry record. Now
1625 it does not use this if it is later than than the arrival time of the
1626 message. Instead it uses the arrival time. This makes for better
1627 behaviour in cases where some deliveries succeed, thus re-setting the
1628 "first failed" field. An example is a quota failure for a huge message
1629 when small messages continue to be delivered. Without this change, the
1630 "time since failure" will always be short, possible causing more frequent
1631 delivery attempts for the huge message than are intended.
1632 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1634 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1635 $1, $2, $3) because the numerical variables can be reset during some
1636 expansion items (e.g. "match"), thereby losing the authentication data.
1638 PH/21 Make -bV show the size of off_t variables so that the test suite can
1639 decide whether to run tests for quotas > 2G.
1641 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1642 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1643 filecount value is greater than 2G or if a quota value is greater than 2G
1644 on a system where the size of off_t is not greater than 4, a panic error
1647 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1648 never match. The debug and -bh output now contains an explicit error
1649 message indicating a malformed IPv4 address or mask.
1651 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1652 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1653 PH/23 above applies.
1655 PH/25 Do not write to syslog when running in the test harness. The only
1656 occasion when this arises is a failure to open the main or panic logs
1657 (for which there is an explicit test).
1659 PH/26 Added the /no_tell option to "control=freeze".
1661 PH/27 If a host name lookup failed very early in a connection, for example, if
1662 the IP address matched host_lookup and the reverse lookup yielded a name
1663 that did not have a forward lookup, an error message of the form "no IP
1664 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1665 could be logged. Now it outputs the IP address instead of "NULL".
1667 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1668 allows the sender and the authenticated sender to be set when
1669 submitting a message from within Exim. Since child_open_exim() is
1670 documented for local_scan(), the new function should be too.
1672 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1673 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1674 results in an empty string is now treated as unset.
1676 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1678 PH/31 Added sender_verify_fail logging option.
1680 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1681 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1682 tidied the source and removed it altogether.
1684 PH/33 When a queue run was abandoned because the load average was too high, a
1685 log line was always written; now it is written only if the queue_run log
1686 selector is set. In addition, the log line for abandonment now contains
1687 information about the queue run such as the pid. This is always present
1688 in "start" and "stop" lines but was omitted from the "abandon" line.
1690 PH/34 Omit spaces between a header name and the colon in the error message that
1691 is given when verify = headers_syntax fails (if there are lots of them,
1692 the message gets confusing).
1694 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1695 names, as there are now some PTR records that contain slashes. This check
1696 is only to protect against broken name servers that fall over on strange
1697 characters, so the fact that it applies to all lookups doesn't matter.
1699 PH/36 Now that the new test suite is complete, we can remove some of the
1700 special code in Exim that was needed for the old test suite. For example,
1701 sorting DNS records because real resolvers return them in an arbitrary
1702 order. The new test suite's fake resolver always returns records in the
1705 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1706 bounces) except when queue_only is set, to avoid logging races between
1707 the different processes.
1709 PH/38 Panic-die if .include specifies a non-absolute path.
1711 PH/39 A tweak to the "H" retry rule from its user.
1713 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1714 a label. They prevented compilation on older perls.
1716 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1717 a warning to be raised on newish perls.
1719 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1720 on queue. Changes to match documented behaviour of showing count of
1721 messages matching specified criteria.
1723 PH/40 Changed the default ident timeout from 30s to 5s.
1725 PH/41 Added support for the use of login_cap features, on those BSD systems
1726 that have them, for controlling the resources used by pipe deliveries.
1728 PH/42 The content-scanning code uses fopen() to create files in which to put
1729 message data. Previously it was not paying any attention to the mode of
1730 the files. Exim runs with umask(0) because the rest of the code creates
1731 files with open(), and sets the required mode explicitly. Thus, these
1732 files were ending up world-writeable. This was not a big issue, because,
1733 being within the spool directory, they were not world-accessible. I have
1734 created a function called modefopen, which takes an additional mode
1735 argument. It sets umask(777), creates the file, chmods it to the required
1736 mode, then resets the umask. All the relevant calls to fopen() in the
1737 content scanning code have been changed to use this function.
1739 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1740 to 24 hours. This avoids potential overflow problems when processing G
1741 and H retry rules. I suspect nobody ever tinkers with this value.
1743 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1745 PH/45 When the plaintext authenticator is running as a client, the server's
1746 challenges are checked to ensure they are valid base64 strings. By
1747 default, the authentication attempt is cancelled if an invalid string is
1748 received. Setting client_ignore_invalid_base64 true ignores these errors.
1749 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1750 they are received. Thus, the responses can be made to depend on the
1751 challenges. If an invalid string is ignored, an empty string is placed in
1754 PH/46 Messages that are created by the autoreply transport now contains a
1755 References: header, in accordance with RFCs 2822 and 3834.
1757 PH/47 Added authenticated_sender_force to the smtp transport.
1759 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1761 PH/49 Installed latest patch from the Sieve maintainer.
1763 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1764 was also set, the appendfile transport was unnecessarily scanning a
1765 directory of message files (e.g. for maildir delivery) to find the count
1766 of files (along with the size), even though it did not need this
1767 information. It now does the scan only if it needs to find either the
1768 size of the count of files.
1770 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1772 PH/52 Two bugs concerned with error handling when the smtp transport is
1775 (i) Exim was not creating retry information for temporary errors given
1776 for individual recipients after the DATA command when the smtp transport
1777 was used in LMTP mode. This meant that they could be retried too
1778 frequently, and not timed out correctly.
1780 (ii) Exim was setting the flag that allows error details to be returned
1781 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1782 recipients that were returned after the DATA command.
1784 PH/53 This is related to PH/52, but is more general: for any failing address,
1785 when detailed error information was permitted to be returned to the
1786 sender, but the error was temporary, then after the final timeout, only
1787 "retry timeout exceeded" was returned. Now it returns the full error as
1788 well as "retry timeout exceeded".
1790 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1791 do this, and (what is worse) MTAs that accept it.
1793 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1794 will now be deprecated.
1796 PH/56 New os.c-cygwin from the Cygwin maintainer.
1798 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1799 formats (previously only available in exim formats via -bpr, -bpru,
1800 and -bpra. Now also available in native and exiqgrep formats)
1802 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1803 with very large, slow to parse queues
1805 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1807 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1809 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1810 responses to authentication challenges, though it was showing the
1811 challenges; (ii) I've removed the CR characters from the debug output for
1814 PH/58 Allow for the insertion of a newline as well as a space when a string
1815 is turned into more than one encoded-word during RFC 2047 encoding. The
1816 Sieve code now uses this.
1818 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1819 data_4xx, lost_connection, tls_required.
1821 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1822 message was being sent as an SMTP response.
1824 PH/61 Add -l and -k options to exicyclog.
1826 PH/62 When verifying, if an address was redirected to one new address, so that
1827 verification continued, and the new address failed or deferred after
1828 having set something in $address_data, the value of $address_data was not
1829 passed back to the ACL. This was different to the case when no
1830 redirection occurred. The value is now passed back in both cases.
1832 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1833 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1834 use login_cap.h, so on its own it isn't the distinguishing feature. The
1835 new name refers directly to the setclassresources() function.
1837 PH/65 Added configuration files for NetBSD3.
1839 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1841 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1842 is preferred over IPv4.
1844 PH/68 The bounce_return_message and bounce_return_body options were not being
1845 honoured for bounces generated during the reception of non-SMTP messages.
1846 In particular, this applied to messages rejected by the ACL. This bug has
1847 been fixed. However, if bounce_return_message is true and bounce_return_
1848 body is false, the headers that are returned for a non-SMTP message
1849 include only those that have been read before the error was detected.
1850 (In the case of an ACL rejection, they have all been read.)
1852 PH/69 The HTML version of the specification is now built in a directory called
1853 spec_html instead of spec.html, because the latter looks like a path with
1854 a MIME-type, and this confuses some software.
1856 PH/70 Catch two compiler warnings in sieve.c.
1858 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1859 function verify_get_ident() calls ip_connect() to connect a socket, but
1860 if the "connect()" function timed out, ip_connect() used to close the
1861 socket. However, verify_get_ident() also closes the socket later, and in
1862 between Exim writes to the log, which may get opened at this point. When
1863 the socket was closed in ip_connect(), the log could get the same file
1864 descriptor number as the socket. This naturally causes chaos. The fix is
1865 not to close the socket in ip_connect(); the socket should be closed by
1866 the function that creates it. There was only one place in the code where
1867 this was missing, in the iplookup router, which I don't think anybody now
1868 uses, but I've fixed it anyway.
1870 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1871 well as to direct DNS lookups. Otherwise the handling of names in host
1872 lists is inconsistent and therefore confusing.
1878 PH/01 Two changes to the default runtime configuration:
1880 (1) Move the checks for relay_from_hosts and authenticated clients from
1881 after to before the (commented out) DNS black list checks.
1883 (2) Add control=submission to the relay_from_hosts and authenticated
1884 clients checks, on the grounds that messages accepted by these
1885 statements are most likely to be submissions.
1887 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1889 (1) Generate an error if the third argument for the ${prvs expansion is
1892 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1895 (3) Reset the variables that are obtained from the first argument of
1896 ${prvscheck and used in the second argument before leaving the code,
1897 because their memory is reclaimed, so using them afterwards may do
1900 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1901 one (it's much easier than Tom thought :-).
1903 (5) Because of (4), we can now allow for the use of $prvscheck_result
1904 inside the third argument.
1906 PH/03 For some reason, the default setting of PATH when running a command from
1907 a pipe transport was just "/usr/bin". I have changed it to
1910 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1911 anything to be listed in the output from -bV.
1913 PH/05 When a filter generated an autoreply, the entire To: header line was
1914 quoted in the delivery log line, like this:
1916 => >A.N.Other <ano@some.domain> <original@ddress> ...
1918 This has been changed so that it extracts the operative address. There
1919 may be more than one such address. If so, they are comma-separated, like
1922 => >ano@some.domain,ona@other.domain <original@ddress> ...
1924 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1925 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1926 not being looked up in the rDNS to get a host name, Exim was showing the
1927 IP address twice in Received: lines, even though the IP addresses were
1928 identical. For example:
1930 Received: from [1.2.3.4] (helo=[1.2.3.4])
1932 However, if the real host name was known, it was omitting the HELO data
1933 if it matched the actual IP address. This has been tidied up so that it
1934 doesn't show the same IP address twice.
1936 PH/07 When both +timestamp and +memory debugging was on, the value given by
1937 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1938 called by the debug printing, thereby overwriting the timestamp buffer.
1939 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1941 PH/08 When the original message was included in an autoreply transport, it
1942 always said "this is a copy of the message, including all the headers",
1943 even if body_only or headers_only was set. It now gives an appropriate
1946 PH/09 Applied a patch from the Sieve maintainer which:
1948 o fixes some comments
1949 o adds the (disabled) notify extension core
1950 o adds some debug output for the result of if/elsif tests
1951 o points to the current vacation draft in the documentation
1952 and documents the missing references header update
1956 o fixes a bug in processing the envelope test (when testing
1957 multiple envelope elements, the last element determinted the
1960 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1961 Electronic Mail") by including:
1963 Auto-submitted: auto-generated
1965 in the messages that it generates (bounce messages and others, such as
1966 warnings). In the case of bounce messages for non-SMTP mesages, there was
1967 also a typo: it was using "Auto_submitted" (underscore instead of
1968 hyphen). Since every message generated by Exim is necessarily in response
1969 to another message, thes have all been changed to:
1971 Auto-Submitted: auto-replied
1973 in accordance with these statements in the RFC:
1975 The auto-replied keyword:
1977 - SHOULD be used on messages sent in direct response to another
1978 message by an automatic process,
1980 - MUST NOT be used on manually-generated messages,
1982 - MAY be used on Delivery Status Notifications (DSNs) and Message
1983 Disposition Notifications (MDNs),
1985 - MUST NOT be used on messages generated by automatic or periodic
1986 processes, except for messages which are automatic responses to
1989 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1990 to the default Received: header definition.
1992 PH/12 Added log selector acl_warn_skipped (default on).
1994 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1995 variables because (a) they are in the wrong storage pool and (b) even if
1996 they were copied, it wouldn't work properly because of the caching.
1998 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1999 checking when decoding. Apparently there are clients that generate
2000 overlong encoded strings. Why am I not surprised?
2002 PH/15 If the first argument of "${if match_address" was not empty, but did not
2003 contain an "@" character, Exim crashed. Now it writes a panic log message
2004 and treats the condition as false.
2006 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2008 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2009 extension "envelope-auth". The code is finished and in agreement with
2010 other implementations, but there is no documentation so far and in fact,
2011 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2012 not changing the active code.
2014 Print executed "if" and "elsif" statements when debugging is used. This
2015 helps a great deal to understand what a filter does.
2017 Document more things not specified clearly in RFC3028. I had all this
2018 sorted out, when out of a sudden new issues came to my mind. Oops."
2020 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2023 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2024 canonical form (as documented). However, after a host name lookup from
2025 the IP address, check_host() was doing a simple string comparison with
2026 addresses acquired from the DNS when checking that the found name did
2027 have the original IP as one of its addresses. Since any found IPv6
2028 addresses are likely to be in abbreviated form, the comparison could
2029 fail. Luckily, there already exists a function for doing the comparison
2030 by converting both addresses to binary, so now that is used instead of
2031 the text comparison.
2033 PH/20 There was another similar case to PH/19, when a complete host name was
2034 given in a host list; looking up its IP address could give an abbreviated
2035 form, whereas the current host's name might or might not be abbreviated.
2036 The same fix has been applied.
2042 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2043 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2046 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2047 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2049 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2051 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2052 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2053 Also, for "Auto-Submitted", treat anything other than "no" as
2054 non-personal, in accordance with RFC 3834. (Previously it treated
2055 anything starting "auto-" as non-personal.)
2057 TF/01 The control=submission/name=... option had a problem with syntax
2058 errors if the name included a slash character. The /name= option
2059 now slurps the rest of the string, so it can include any characters
2060 but it must come last in the list of options (after /sender_retain
2063 PH/05 Some modifications to the interface to the fake nameserver for the new
2071 TK/01 Added the "success_on_redirect" address verification option. See
2072 NewStuff for rationale and an example.
2074 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2076 PH/02 Patch to exigrep to allow it to work on syslog lines.
2078 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2079 fread() to skip over the body file's header line, because in Cygwin the
2080 header line is locked and is inaccessible.
2082 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2083 co-exist for some time) to make it clear that it is the Exim ID that is
2084 referenced, not the Message-ID: header line.
2086 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2087 string_format() function, because snprintf() does not exist on all
2090 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2091 ${stat: expansion item.
2093 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2094 protocol synchronization error", to keep the pedants happy.
2096 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2097 well as for IRIX systems, when gcc is being used. See the host.c source
2100 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2102 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2105 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2106 transport and to the smtp transport in LMTP mode.
2108 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2110 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2111 run a filter in a subprocess. This could lead to confusion in subsequent
2112 lookups in the parent process. There should also be a search_tidyup() at
2113 the end of the subprocess.
2115 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2116 only if the host matched helo_try_verify_hosts, which caused the
2117 verification to occur when the EHLO/HELO command was issued. The ACL just
2118 tested the remembered result. Now, if a previous verification attempt has
2119 not happened, "verify = helo" does it there and then.
2121 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2123 TK/03 Fix log output including CR from clamd.
2125 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2126 exist provoked a memory error which could cause a segfault.
2128 PH/15 Installed PCRE 6.2
2130 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2132 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2133 of the problem. Specifically, suggested +O2 rather than +O1 for the
2136 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2138 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2139 record, the port was not correctly logged when the outgoing_port log
2140 selector was set (it logged the transort's default port).
2142 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2143 fallback_hosts, and "hosts" in the smtp transport.
2145 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2146 host errors such as "Connection refused".
2148 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2149 authentication with radiusclient 0.4.9:
2151 - Error returned from rc_read_config was caught wrongly
2152 - Username/password not passed on to radius server due to wrong length.
2154 The presumption is that some radiusclient API changes for 4.51/PH/17
2155 were not taken care of correctly. The code is still untested by me (my
2156 Linux distribution still has 0.3.2 of radiusclient), but it was
2157 contributed by a Radius user.
2159 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2160 expanding the "port" option of the smtp transport.
2162 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2163 while reading a MIME header. Thanks to Tom Hughes for a patch.
2165 PH/24 Include config.h inside local_scan.h so that configuration settings are
2168 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2169 that in an ACL for RCPT (for example), you can examine exactly what was
2172 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2173 commands, but it was not correctly comparing the address with the actual
2174 client host address. Thus, it would show the EHLO address in Received:
2175 header lines when this was not necessary.
2177 PH/27 Added the % operator to ${eval:}.
2179 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2180 it should be ignoring failures (because with -C, for example, it has lost
2181 privilege). It wasn't ignoring creation failures other than "already
2184 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2187 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2188 because an input line was too long, either on its own, or by virtue of
2189 too many continuations, the temporary file was not being removed, and the
2190 return code was incorrect.
2192 PH/31 Missing "BOOL" in function definition in filtertest.c.
2194 PH/32 Applied Sieve patches from the maintainer.
2196 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2198 PH/33 Added "verify = not_blind".
2200 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2201 Local/Makefile (with some defaults set). These are used in built scripts
2202 such as exicyclog, but they have never been used in the exim_install
2203 script (though there are many overriding facilities there). I have
2204 arranged that the exim_install script now takes note of these two
2207 PH/35 Installed configuration files for Dragonfly.
2209 PH/36 When a locally submitted message by a trusted user did not contain a
2210 From: header, and the sender address was obtained from -f or from an SMTP
2211 MAIL command, and the trusted user did not use -F to supply a sender
2212 name, $originator_name was incorrectly used when constructing a From:
2213 header. Furthermore, $originator_name was used for submission mode
2214 messages from external hosts without From: headers in a similar way,
2215 which is clearly wrong.
2217 PH/37 Added control=suppress_local_fixups.
2219 PH/38 When log_selector = +received_sender was set, and the addition of the
2220 sender made the log line's construction buffer exactly full, or one byte
2221 less than full, an overflow happened when the terminating "\n" was
2224 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2225 when the result of a list match is failure because a DNS lookup failed.
2227 PH/40 RM_COMMAND is now used in the building process.
2229 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2230 the "build-* directories that it finds.
2232 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2233 address in a domain literal was a prefix of an interface address.
2235 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2236 when verifying a sender address, unless rewrite_headers is false.
2238 PH/44 Wrote a long comment about why errors_to addresses are verified as
2239 recipients, not senders.
2241 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2242 the ratelimit ACL was added.
2244 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2246 PH/46 Added extra information about PostgreSQL errors to the error string.
2248 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2249 avoiding the need to install special zones in a real server. This is
2250 backwards compatible; if it can't find the fake resolver, it drops back.
2251 Thus, both old and new test suites can be run.
2253 TF/02 Added util/ratelimit.pl
2255 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2256 clock is set back in time.
2258 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2259 Candler <B.Candler@pobox.com>.
2261 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2262 OK for addresses that are the result of redirections.
2264 PH/48 A number of further additions for the benefit of the new test suite,
2265 including a fake gethostbyname() that interfaces to the fake DNS resolver
2268 TF/06 The fix for widen_domains has also been applied to qualify_single and
2269 search_parents which are the other dnslookup options that can cause
2272 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2275 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2277 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2278 service" if it encountered an MX record with an empty target hostname.
2279 The message is now "an MX or SRV record indicated no SMTP service".
2281 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2282 if the DNS of the sending site is misconfigured. This is quite a
2283 common situation. This change restores the behaviour of treating a
2284 helo verification defer as a failure.
2286 PH/51 If self=fail was set on a router, the bounce message did not include the
2287 actual error message.
2293 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2295 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2296 command itself ended in error, the underlying message about the transport
2297 filter timeout was being overwritten with the pipe command error. Now the
2298 underlying error message should be appended to the second error message.
2300 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2302 PH/02 Reduce the amount of output that "make" produces by default. Full output
2303 can still be requested.
2305 PH/03 The warning log line about a condition test deferring for a "warn" verb
2306 was being output only once per connection, rather than after each
2307 occurrence (because it was using the same function as for successful
2308 "warn" verbs). This seems wrong, so I have changed it.
2310 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2311 it should not have, which might have caused a crash in the right
2312 circumstances, but probably never did.
2314 PH/04 Installed a modified version of Tony Finch's patch to make submission
2315 mode fix the return path as well as the Sender: header line, and to
2316 add a /name= option so that you can make the user's friendly name appear
2319 TF/03 Added the control = fakedefer ACL modifier.
2321 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2322 Mark Lowes for thorough testing.
2324 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2326 TK/03 Merged latest SRS patch from Miles Wilton.
2328 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2329 with the definition in sysexits.h (which is #included earlier).
2330 Fortunately, Exim does not actually use EX_OK. The code used to try to
2331 preserve the sysexits.h value, by assumimg that macro definitions were
2332 scanned for macro replacements. I have been disabused of this notion,
2333 so now the code just undefines EX_OK before #including unistd.h.
2335 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2336 in the smtp transport. When a block could not be written in a single
2337 write() function, the timeout was being re-applied to each part-write.
2338 This seems wrong - if the receiver was accepting one byte at a time it
2339 would take for ever. The timeout is now adjusted when this happens. It
2340 doesn't have to be particularly precise.
2342 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2343 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2345 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2346 without a domain if the check to <postmaster@domain> fails.
2348 SC/01 Eximstats: added -xls and the ability to specify output files
2349 (patch written by Frank Heydlauf).
2351 SC/02 Eximstats: use FileHandles for outputing results.
2353 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2355 SC/04 Eximstats: fixed display of large numbers with -nvr option
2357 SC/05 Eximstats: fixed merging of reports with empty tables.
2359 SC/06 Eximstats: added the -include_original_destination flag
2361 SC/07 Eximstats: removed tabs and trailing whitespace.
2363 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2365 TK/06 MBOX spool code: Add real "From " MBOX separator line
2366 so the .eml file is really in mbox format (even though
2367 most programs do not really care). Patch from Alex Miller.
2369 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2370 The latter is generated from $received_to and is only set if the
2371 message has one envelope recipient. SA can use these headers,
2372 obviously out-of-the-box. Patch from Alex Miller.
2374 PH/08 The ${def test on a variable was returning false if the variable's
2375 value was "0", contrary to what the specification has always said!
2376 The result should be true unless the variable is empty.
2378 PH/09 The syntax error of a character other than { following "${if
2379 def:variable_name" (after optional whitespace) was not being diagnosed.
2380 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2381 accidental colon was present, for example, could give incorrect results.
2383 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2384 result is used (not including appendfile, where other changes are about
2387 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2388 This involved changing a lot of size variables from int to off_t. It
2389 should work with maildirs and everything.
2391 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2392 spamd dying while we are connected to it.
2394 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2397 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2398 function for BDB 4.3.
2400 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2402 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2403 the sources for PCRE have been re-arranged and more files are now
2406 PH/15 The code I had for printing potentially long long variables in PH/11
2407 above was not the best (it lost precision). The length of off_t variables
2408 is now inspected at build time, and an appropriate printing format (%ld
2409 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2410 to be "long long int" or "long int". This is needed for the internal
2411 formatting function string_vformat().
2413 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2414 the configuration file to be ":syslog", then the script "guesses" where
2415 the logs files are, rather than using the compiled in default. In our
2416 case the guess is not the same as the compiled default, so the script
2417 suddenly stopped working when I started to use syslog. The patch checks
2418 to see if log_file_path is "". If so, it attempts to read it from exim
2419 with no configuration file to get the compiled in version, before it
2420 falls back to the previous guessing code."
2422 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2423 implementing BATV in an Exim configuration. See NewStuff for the gory
2426 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2427 Makefile that are specific to HP-UX.
2429 PH/18 If the "use_postmaster" option was set for a recipient callout together
2430 with the "random" option, the postmaster address was used as the MAIL
2431 FROM address for the random test, but not for the subsequent recipient
2432 test. It is now used for both.
2434 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2435 patch removes a few documentation additions to RFC 3028, because the
2436 latest draft now contains them. It adds the new en;ascii-case comparator
2437 and a new error check for 8bit text in MIME parts. Comparator and
2438 require names are now matched exactly. I enabled the subaddress
2439 extension, but it is not well tested yet (read: it works for me)."
2441 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2442 rework some of the code of TK/09 above to avoid the hardwired use of
2443 "%lld" and "long long". Replaced the call to snprintf() with a call to
2446 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2447 records point to non-existent hosts", "retry timeout exceeded", and
2448 "retry time not reached for any host after a long failure period".
2450 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2451 experimental DomainKeys support:
2453 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2454 (2) On an error such as an illegally used "control", the wrong name for
2455 the control was given.
2457 These problems did NOT occur unless DomainKeys support was compiled.
2459 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2461 PH/24 Added ${if match_ip condition.
2463 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2464 databases so that it will be absolutely obvious if a crash occurs in the
2465 DB library. This is a regular occurrence (often caused by mis-matched
2468 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2469 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2470 picked up on a user's system that detects such things. There doesn't seem
2471 to be a gcc warning option for this - only an attribute that has to be
2472 put on the function's prototype. It seems that in Fedora Core 4 they have
2473 set this on a number of new functions. No doubt there will be more in due
2476 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2477 specify a transport. However, if an address that was verified by such a
2478 router was the subject of a callout, Exim crashed because it tried to
2479 read the rcpt_include_affixes from the non-existent transport. Now it
2480 just assumes that the setting of that option is false. This bug was
2481 introduced by 4.51/PH/31.
2483 PH/28 Changed -d+all to exclude +memory, because that information is very
2484 rarely of interest, but it makes the output a lot bigger. People tend to
2485 do -d+all out of habit.
2487 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2488 code in os-type was giving problems when libc.so lives in lib64, like on
2491 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2492 aren't the modern standard, and it seems that some systems' include files
2493 don't always have them. Exim was already checking for some of the newer
2494 ones like T_AAAA, and defining it itself. I've added checks for all the
2495 record types that Exim uses.
2497 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2498 not automatically generating a new one, as it is supposed to. This
2499 prevented TLS from working. If the file did exist, but contained invalid
2500 data, a new version was generated, as expected. It was only the case of a
2501 non-existent file that was broken.
2503 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2504 with a change in libdomainkeys > 0.64.
2506 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2507 from DNS. If the selector record carries the flag, it now has
2508 precedence over the domain-wide flag.
2510 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2512 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2513 the use of a port name that isn't defined in /etc/services) occurred, the
2514 message was deferred as in a normal delivery, and thus remained on the
2515 spool, instead of being failed because of the mua_wrapper setting. This
2516 is now fixed, and I tidied up some of the mua_wrapper messages at the
2519 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2520 the messages in a hash of arrays rather than using individual hashes.
2521 This is a bit cleaner and results in dramatic memory savings, albeit
2522 at a slight CPU cost.
2524 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2525 as requested by Marc Sherman.
2527 SC/10 Eximstats: added histograms for user specified patterns as requested
2530 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2532 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2533 fopen() in the content-scanning modules that did not already have it.
2539 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2540 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2542 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2544 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2546 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2547 to test Sieve filters that use "vacation".
2549 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2550 that changes the way the GnuTLS parameters are stored in the cache file.
2551 The new format can be generated externally. For backward compatibility,
2552 if the data in the cache doesn't make sense, Exim assumes it has read an
2553 old-format file, and it generates new data and writes a new file. This
2554 means that you can't go back to an older release without removing the
2557 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2558 work if there are any delivery delays because "one_time" forces the
2559 parent to be marked "delivered", so its unseen clone is never tried
2560 again. For this reason, Exim now forbids the simultaneous setting of
2563 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2564 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2565 entry if you want to know the details. The fix, however, neglected to
2566 consider the case when local delivery batching is involved. The test for
2567 "previously delivered" was not happening when checking to see if an
2568 address could be batched with a previous (undelivered) one; under
2569 certain circumstances this could lead to multiple deliveries to the same
2572 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2573 in its include files, and this causes problems building Exim.
2575 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2576 header_syntax) but Exim was just ignoring anything given after a slash.
2577 In particular, this caused confusion with an attempt to use "verify =
2578 reverse_host_lookup/defer_ok". An error is now given when options are
2579 supplied for verify items that do not have them. (Maybe reverse_host_
2580 lookup should have a defer_ok option, but that's a different point.)
2582 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2583 defined by RFC 821) to 2048, because there were problems with some AUTH
2584 commands, and RFC 1869 says the size should be increased for extended
2585 SMTP commands that take arguments.
2587 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2590 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2591 "unknown" error; now it says that the functionality isn't in the binary.
2593 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2594 an address' error message when a string expansion fails (syntax or
2595 whatever). Otherwise the password may appear in the log. Following change
2596 PH/42 below, there is no longer a chance of it appearing in a bounce
2599 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2601 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2602 of its addresses was ever tried. (Bugzilla bug #2.)
2604 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2605 the result incorrectly in the debug output. (It correctly added a newline
2606 to what was transported.)
2608 TF/01 Added $received_time.
2610 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2611 commented out examples of how to interface to a virus scanner and to
2612 SpamAssassin. Also added commented examples of av_scanner and
2613 spamd_address settings.
2615 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2616 and controls are allowed in which ACLs. There were a couple of minor
2617 errors. Some of the entries in the conditions table (which is a table of
2618 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2619 negation of where the condition IS allowed.
2621 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2623 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2624 header file does not have a version number, so I've had to invent a new
2625 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2626 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2627 radiusclient), but it was contributed by a Radius user.
2629 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2630 files or directories, for OpenSSL.
2632 PH/19 When an Exim process that is running as root has to create an Exim log
2633 file, it does so in a subprocess that runs as exim:exim so as to get the
2634 ownership right at creation (otherwise, other Exim processes might see
2635 the file with the wrong ownership). There was no test for failure of this
2636 fork() call, which would lead to the process getting stuck as it waited
2637 for a non-existent subprocess. Forks do occasionally fail when resources
2638 run out. I reviewed all the other calls to fork(); they all seem to check
2641 PH/20 When checking for unexpected SMTP input at connect time (before writing
2642 the banner), Exim was not dealing correctly with a non-positive return
2643 from the read() function. If the client had disconnected by this time,
2644 the result was a log entry for a synchronization error with an empty
2645 string after "input=" when read() returned zero. If read() returned -1
2646 (an event I could not check), uninitialized data bytes were printed.
2647 There were reports of junk text (parts of files, etc) appearing after
2650 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2652 PH/22 Added support for macro redefinition, and (re)definition in between
2653 driver and ACL definitions.
2655 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2656 forgetting to use the resulting value; it was using the unexpanded value.
2658 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2659 hadn't been configured. The fix is from Juergen Kreileder, who
2660 understands it better than I do:
2662 "Here's what I see happening with three configured cyrus_sasl
2663 authenticators configured (plain, login, cram-md5):
2665 On startup auth_cyrus_sasl_init() gets called for each of these.
2666 This means three calls to sasl_listmech() without a specified mech_list.
2667 => SASL tests which mechs of all available mechs actually work
2668 => three warnings about OTP not working
2669 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2671 With the patch, sasl_listmech() also gets called three times. But now
2672 SASL's mech_list option is set to the server_mech specified in the the
2673 authenticator. Or in other words, the answer from sasl_listmech()
2674 gets limited to just the mech you're testing for (which is different
2676 => the return list contains just 'plain' or 'login', 'cram-md5' or
2677 nothing depending on the value of ob->server_mech.
2679 I've just tested the patch: Authentication still works fine,
2680 unavailable mechs specified in the exim configuration are still
2681 caught, and the auth.log warnings about OTP are gone."
2683 PH/25 When debugging is enabled, the contents of the command line are added
2684 to the debugging output, even when log_selector=+arguments is not
2687 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2688 answer is "GNU", and only if the return is "GNU/something" is the answer
2691 PH/27 $acl_verify_message is now set immediately after the failure of a
2692 verification in an ACL, and so is available in subsequent modifiers. In
2693 particular, the message can be preserved by coding like this:
2695 warn !verify = sender
2696 set acl_m0 = $acl_verify_message
2698 Previously, $acl_verify_message was set only while expanding "message"
2699 and "log_message" when a very denied access.
2701 PH/28 Modified OS/os.c-Linux with
2703 -#ifndef OS_LOAD_AVERAGE
2704 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2706 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2707 nomenclature these days.)
2709 PH/29 Installed patch from the Sieve maintainer that adds the options
2710 sieve_useraddress and sieve_subaddress to the redirect router.
2712 PH/30 In these circumstances:
2713 . Two addresses routed to the same list of hosts;
2714 . First host does not offer TLS;
2715 . First host accepts first address;
2716 . First host gives temporary error to second address;
2717 . Second host offers TLS and a TLS session is established;
2718 . Second host accepts second address.
2719 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2720 and peerdn, if requested) that were in fact used only for the second
2723 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2724 attention to any local part prefix or suffix that was matched by the
2725 router that accepted the address. It now behaves in the same way as it
2726 does for delivery: the affixes are removed from the local part unless
2727 rcpt_include_affixes is set on the transport.
2729 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2730 timeout during the DATA phase of an incoming message.
2732 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2733 applied a patch sanctioned by the Sieve maintainer.
2735 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2736 the uid or gid is negative. A case of a negative gid caused this to be
2737 noticed. The fix allows for either to be negative.
2739 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2740 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2743 PH/36 The change PH/12 above was broken. Fixed it.
2745 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2746 the grounds that routing the same address twice would always produce the
2747 same answer. This might have been true once, but it is certainly no
2748 longer true now. Routing a child address may depend on the previous
2749 routing that produced that child. Some complicated redirection strategies
2750 went wrong when messages had multiple recipients, and made Exim's
2751 behaviour dependent on the order in which the addresses were given.
2753 I have moved the duplicate checking until after the routing is complete.
2754 Exim scans the addresses that are assigned to local and remote
2755 transports, and removes any duplicates. This means that more work will be
2756 done, as duplicates will always all be routed, but duplicates are
2757 presumably rare, so I don't expect this is of any significance.
2759 For deliveries to pipes, files, and autoreplies, the duplicate checking
2760 still happens during the routing process, since they are not going to be
2763 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2764 It corrects a timeout issue with spamd. This is Ian's comment: "The
2765 background is that sometimes spamd either never reads data from a
2766 connection it has accepted, or it never writes response data. The exiscan
2767 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2768 blindly assumes that writes won't block so it may never time out."
2770 PH/39 Allow G after quota size as well as K and M.
2772 PH/40 The value set for $authenticated_id in an authenticator may not contain
2773 binary zeroes or newlines because the value is written to log lines and
2774 to spool files. There was no check on this. Now the value is run through
2775 the string_printing() function so that such characters are converted to
2776 printable escape sequences.
2778 PH/41 $message_linecount is a new variable that contains the total number of
2779 lines in the message. Compare $body_linecount, which is the count for the
2782 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2783 bounce and delay warning messages, except in certain special cases, which
2786 (a) An SMTP error message from a remote host;
2787 (b) A message specified in a :fail: redirection;
2788 (c) A message specified in a "fail" command in a system filter;
2789 (d) A message specified in a FAIL return from the queryprogram router;
2790 (e) A message specified by the cannot_route_message router option.
2792 In these cases only, Exim does include the error details in bounce and
2793 warning messages. There are also a few cases where bland messages such
2794 as "unrouteable address" or "local delivery error" are given.
2796 PH/43 $value is now also set for the "else" part of a ${run expansion.
2798 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2799 being worked on, but at least Exim now implements the latest version to
2802 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2803 process to complete was treated as a delivery failure, a timeout while
2804 writing the message to the pipe was logged, but erroneously treated as a
2805 successful delivery. Such timeouts include transport filter timeouts. For
2806 consistency with the overall process timeout, these timeouts are now
2807 treated as errors, giving rise to delivery failures by default. However,
2808 there is now a new Boolean option for the pipe transport called
2809 timeout_defer, which, if set TRUE, converts the failures into defers for
2810 both kinds of timeout. A transport filter timeout is now identified in
2813 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2814 systems where "make" and "gmake" are different, calling "gmake" at top
2815 level broke things. I've arranged for the value of $(MAKE) to be passed
2816 from the Makefile to this script so that it can call the same version of
2820 A note about Exim versions 4.44 and 4.50
2821 ----------------------------------------
2823 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2824 changes of various kinds. As a consequence, a big documentation update was
2825 needed. This delayed the release for rather longer than seemed good, especially
2826 in the light of a couple of (minor) security issues. Therefore, the changes
2827 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2828 release. So 4.44 and 4.50 are in effect two different branches that both start
2831 I have left the 4.50 change log unchanged; it contains all the changes since
2832 4.43. The change log for 4.44 is below; many of its items are identical to
2833 those for 4.50. This seems to be the most sensible way to preserve the
2834 historical information.
2840 1. Minor wording change to the doc/README.SIEVE file.
2842 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2843 computation of the current number of files was incorrect.
2845 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2846 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2849 4. Give more explanation in the error message when the command for a transport
2850 filter fails to execute.
2852 5. There are several places where Exim runs a non-Exim command in a
2853 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2854 was being done only for the command run by the queryprogram router. It is
2855 now done for all such subprocesses. The other cases are: ${run, transport
2856 filters, and the commands run by the lmtp and pipe transports.
2858 6. Added CONFIGURE_GROUP build-time option.
2860 7. Some older OS have a limit of 256 on the maximum number of file
2861 descriptors. Exim was using setrlimit() to set 1000 as a large value
2862 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2863 systems. I've change it so that if it can't get 1000, it tries for 256.
2865 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2866 was an oversight, and furthermore, ever since the addition of extra
2867 controls (e.g. 4.43/32), the checks on when to allow different forms of
2868 "control" were broken. There should now be diagnostics for all cases when a
2869 control that does not make sense is encountered.
2871 9. Added the /retain_sender option to "control=submission".
2873 10. $recipients is now available in the predata ACL (oversight).
2875 11. Tidy the search cache before the fork to do a delivery from a message
2876 received from the command line. Otherwise the child will trigger a lookup
2877 failure and thereby defer the delivery if it tries to use (for example) a
2878 cached ldap connection that the parent has called unbind on.
2880 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2881 of $address_data from the recipient verification was clobbered by the
2882 sender verification.
2884 13. The value of address_data from a sender verification is now available in
2885 $sender_address_data in subsequent conditions in the ACL statement.
2887 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2889 15. Added a new option "connect=<time>" to callout options, to set a different
2892 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2893 was its contents. (It was OK if the option was not defined at all.)
2895 17. A "Completed" log line is now written for messages that are removed from
2896 the spool by the -Mrm option.
2898 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2899 information about exactly what failed.
2901 19. Added -dd to debug only the daemon process.
2903 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2904 handles timeouts, both on the server side and network timeouts. Renamed the
2905 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2907 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2908 It is now set to "smtps".
2910 22. $host_address is now set to the target address during the checking of
2911 ignore_target_hosts.
2913 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2914 being passed; this would have caused $sender_host_name to have been used if
2915 matching the list had actually called for a host name (not very likely,
2916 since this list is usually IP addresses). A host name is now passed as
2919 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2920 code that specifies a non-restarting handler (typically sigaction() in
2921 modern systems) in an attempt to fix a rare and obscure crash bug.
2923 25. Narrowed the window for a race in the daemon that could cause it to ignore
2924 SIGCHLD signals. This is not a major problem, because they are used only to
2925 wake it up if nothing else does.
2927 26. A malformed maildirsize file could cause Exim to calculate negative values
2928 for the mailbox size or file count. Odd effects could occur as a result.
2929 The maildirsize information is now recalculated if the size or filecount
2932 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2933 support for a long time. Removed HAVE_SYS_VFS_H.
2935 28. Installed the latest version of exipick from John Jetmore.
2937 29. In an address list, if the pattern was not a regular expression, an empty
2938 subject address (from a bounce message) matched only if the pattern was an
2939 empty string. Non-empty patterns were not even tested. This was the wrong
2940 because it is perfectly reasonable to use an empty address as part of a
2941 database query. An empty address is now tested by patterns that are
2942 lookups. However, all the other forms of pattern expect the subject to
2943 contain a local part and a domain, and therefore, for them, an empty
2944 address still always fails if the pattern is not itself empty.
2946 30. Exim went into a mad DNS loop when attempting to do a callout where the
2947 host was specified on an smtp transport, and looking it up yielded more
2948 than one IP address.
2950 31. Re-factored the code for checking spool and log partition space into a
2951 function that finds that data and another that does the check. The former
2952 is then used to implement four new variables: $spool_space, $log_space,
2953 $spool_inodes, and $log_inodes.
2955 32. The RFC2047 encoding function was originally intended for short strings
2956 such as real names; it was not keeping to the 75-character limit for
2957 encoded words that the RFC imposes. It now respects the limit, and
2958 generates multiple encoded words if necessary. To be on the safe side, I
2959 have increased the buffer size for the ${rfc2047: expansion operator from
2962 33. It is now permitted to omit both strings after an "if" condition; if the
2963 condition is true, the result is "true". As before, when the second string
2964 is omitted, a false condition yields an empty string. This makes it less
2965 cumbersome to write custom ACL and router conditions.
2967 34. Failure to deliver a bounce message always caused it to be frozen, even if
2968 there was an errors_to setting on the router. The errors_to setting is now
2971 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2972 canonical form (fully expanded) before being placed in
2973 $sender_host_address.
2975 36. The table in the code that translates DNS record types into text (T_A to
2976 "A" for instance) was missing entries for NS and CNAME. It is just possible
2977 that this could have caused confusion if both these types were looked up
2978 for the same domain, because the text type is used as part of Exim's
2979 per-process caching. But the chance of anyone hitting this buglet seems
2982 37. The dnsdb lookup has been extended in a number of ways.
2984 (1) There is a new type, "zns", which walks up the domain tree until it
2985 finds some nameserver records. It should be used with care.
2987 (2) There is a new type, "mxh", which is like "mx" except that it returns
2988 just the host names, not the priorities.
2990 (3) It is now possible to give a list of domains (or IP addresses) to be
2991 looked up. The behaviour when one of the lookups defers can be
2992 controlled by a keyword.
2994 (4) It is now possible to specify the separator character for use when
2995 multiple records are returned.
2997 38. The dnslists ACL condition has been extended: it is now possible to supply
2998 a list of IP addresses and/or domains to be looked up in a particular DNS
3001 39. Added log_selector=+queue_time_overall.
3003 40. When running the queue in the test harness, wait just a tad after forking a
3004 delivery process, to get repeatability of debugging output.
3006 41. Include certificate and key file names in error message when GnuTLS fails
3007 to set them up, because the GnuTLS error message doesn't include the name
3008 of the failing file when there is a problem reading it.
3010 42. Allow both -bf and -bF in the same test run.
3012 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3014 44. The "Exiscan patch" is now merged into the mainline Exim source.
3016 45. Sometimes the final signoff response after QUIT could fail to get
3017 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3018 < 0 before doing a fflush(). This bug looks as though it goes back to the
3019 introduction of TLS in release 3.20, but "sometimes" must have been rare
3020 because the tests only now provoked it.
3022 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3023 (this can affect the format of dates).
3025 47. exim_tidydb, when checking for the continued existence of a message for
3026 which it has found a message-specific retry record, was not finding
3027 messages that were in split spool directories. Consequently, it was
3028 deleting retry records that should have stayed in existence.
3030 48. Steve fixed some bugs in eximstats.
3032 49. The SPA authentication driver was not abandoning authentication and moving
3033 on to the next authenticator when an expansion was forced to fail,
3034 contradicting the general specification for all authenticators. Instead it
3035 was generating a temporary error. It now behaves as specified.
3037 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3038 (the order specifies the preference for clients). The order is now AES256,
3039 AES128, 3DES, ARCFOUR128.
3041 51. Small patch to Sieve code - explicitly set From: when generating an
3044 52. Exim crashed if a remote delivery caused a very long error message to be
3045 recorded - for instance if somebody sent an entire SpamAssassin report back
3046 as a large number of 550 error lines. This bug was coincidentally fixed by
3047 increasing the size of one of Exim's internal buffers (big_buffer) that
3048 happened as part of the Exiscan merge. However, to be on the safe side, I
3049 have made the code more robust (and fixed the comments that describe what
3052 53. Now that there can be additional text after "Completed" in log lines (if
3053 the queue_time_overall log selector is set), a one-byte patch to exigrep
3054 was needed to allow it to recognize "Completed" as not the last thing in
3057 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3058 patch that reportedly fixes this has been added. I am not expert enough to
3059 create a test for it. This is what the patch creator wrote:
3061 "I found a little strange behaviour of ldap code when working with
3062 Windows 2003 AD Domain, where users was placed in more than one
3063 Organization Units. When I tried to give exim partial DN, the exit code
3064 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3065 But simultaneously result of request was absolutely normal ldap result,
3066 so I produce this patch..."
3068 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3069 REFERENCE, so I have modified the code to exclude the patch when that macro
3072 55. Some experimental protocols are using DNS PTR records for new purposes. The
3073 keys for these records are domain names, not reversed IP addresses. The
3074 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3075 leaves it alone. Component reversal etc. now happens only for IP addresses.
3078 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3080 57. Double the size of the debug message buffer (to 2048) so that more of very
3081 long debug lines gets shown.
3083 58. The exicyclog utility now does better if the number of log files to keep
3084 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3086 59. Two changes related to the smtp_active_hostname option:
3088 (1) $smtp_active_hostname is now available as a variable.
3089 (2) The default for smtp_banner uses $smtp_active_hostname instead
3090 of $primary_hostname.
3092 60. The host_aton() function is supposed to be passed a string that is known
3093 to be a valid IP address. However, in the case of IPv6 addresses, it was
3094 not checking this. This is a hostage to fortune. Exim now panics and dies
3095 if the condition is not met. A case was found where this could be provoked
3096 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3097 components; fortuitously, this particular loophole had already been fixed
3098 by change 4.50/55 above.
3100 If there are any other similar loopholes, the new check in host_aton()
3101 itself should stop them being exploited. The report I received stated that
3102 data on the command line could provoke the exploit when Exim was running as
3103 exim, but did not say which command line option was involved. All I could
3104 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3105 running as the user.
3108 61. There was a buffer overflow vulnerability in the SPA authentication code
3109 (which came originally from the Samba project). I have added a test to the
3110 spa_base64_to_bits() function which I hope fixes it.
3113 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3114 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3116 63. The daemon start-up calls getloadavg() while still root for those OS that
3117 need the first call to be done as root, but it missed one case: when
3118 deliver_queue_load_max is set with deliver_drop_privilege. This is
3119 necessary for the benefit of the queue runner, because there is no re-exec
3120 when deliver_drop_privilege is set.
3122 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3123 This has been fixed.
3125 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3126 was in use, was not putting the data itself into the right store pool;
3127 consequently, it could be overwritten for a subsequent message in the same
3128 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3131 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3133 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3134 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3135 the function were treating the return as a boolean value, which happened to
3136 work because 0=false and not-0=true, but is not correct code.
3138 68. The host_aton() function was not handling scoped IPv6 addresses (those
3139 with, for example, "%eth0" on the end) correctly.
3141 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3142 negated items (that is, ~something) in unsigned ints. Some compilers
3143 apparently mutter when there is no cast.
3145 70. If an address verification called from an ACL failed, and did not produce a
3146 user-specific message (i.e. there was only a "system" message), nothing was
3147 put in $acl_verify_message. In this situation, it now puts the system
3150 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3151 session; change 4.31/43 added the unwanted input to the log line - except
3152 that it did not do this in the start of session case. It now does.
3154 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3155 This is wrong and can cause the other end to generate a synchronization
3156 error if it is another Exim or anything else that does the synchronization
3157 check. A QUIT command is no longer sent after a timeout.
3159 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3160 during host lookups.
3162 74. The defer_ok option of callout verification was not working if it was used
3163 when verifying addresses in header lines, that is, for this case:
3165 verify = header_sender/callout=defer_ok
3167 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3168 those file descriptors could be used for SMTP connections. If anything
3169 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3170 could be sent to the SMTP client, causing chaos. The daemon now opens
3171 stdin, stdout, and stderr to /dev/null when it puts itself into the
3174 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3175 log by default. The user can override this with suitable Perl magic.
3177 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3178 the log message when discard triggers, was not working for the DATA ACL or
3179 for the non-SMTP ACL.
3181 78. Error message wording change in sieve.c.
3183 79. If smtp_accept_max_per_host was set, the number of connections could be
3184 restricted to fewer than expected, because the daemon was trying to set up
3185 a new connection before checking whether the processes handling previous
3186 connections had finished. The check for completed processes is now done
3187 earlier. On busy systems, this bug wouldn't be noticed because something
3188 else would have woken the daemon, and it would have reaped the completed
3191 80. If a message was submitted locally by a user whose login name contained one
3192 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3193 It caused a spool format error. I have fixed the spool reading code. A
3194 related problem was that the "from" clause in the Received: line became
3195 illegal because of the space(s). It is now covered by ${quote_local_part.
3197 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3200 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3202 83. Patch from Sieve maintainer to fix unterminated string problem in
3203 "vacation" handling.
3205 84. Some minor changes to the Linux configuration files to help with other
3206 OS variants using glibc.
3208 85. One more patch for Sieve to update vacation handling to latest spec.
3211 ----------------------------------------------------
3212 See the note above about the 4.44 and 4.50 releases.
3213 ----------------------------------------------------
3219 1. Change 4.43/35 introduced a bug that caused file counts to be
3220 incorrectly computed when quota_filecount was set in an appendfile
3223 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3224 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3227 3. Give more explanation in the error message when the command for a transport
3228 filter fails to execute.
3230 4. There are several places where Exim runs a non-Exim command in a
3231 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3232 was being done only for the command run by the queryprogram router. It is
3233 now done for all such subprocesses. The other cases are: ${run, transport
3234 filters, and the commands run by the lmtp and pipe transports.
3236 5. Some older OS have a limit of 256 on the maximum number of file
3237 descriptors. Exim was using setrlimit() to set 1000 as a large value
3238 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3239 systems. I've change it so that if it can't get 1000, it tries for 256.
3241 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3242 was an oversight, and furthermore, ever since the addition of extra
3243 controls (e.g. 4.43/32), the checks on when to allow different forms of
3244 "control" were broken. There should now be diagnostics for all cases when a
3245 control that does not make sense is encountered.
3247 7. $recipients is now available in the predata ACL (oversight).
3249 8. Tidy the search cache before the fork to do a delivery from a message
3250 received from the command line. Otherwise the child will trigger a lookup
3251 failure and thereby defer the delivery if it tries to use (for example) a
3252 cached ldap connection that the parent has called unbind on.
3254 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3255 of $address_data from the recipient verification was clobbered by the
3256 sender verification.
3258 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3259 was its contents. (It was OK if the option was not defined at all.)
3261 11. A "Completed" log line is now written for messages that are removed from
3262 the spool by the -Mrm option.
3264 12. $host_address is now set to the target address during the checking of
3265 ignore_target_hosts.
3267 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3268 being passed; this would have caused $sender_host_name to have been used if
3269 matching the list had actually called for a host name (not very likely,
3270 since this list is usually IP addresses). A host name is now passed as
3273 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3274 code that specifies a non-restarting handler (typically sigaction() in
3275 modern systems) in an attempt to fix a rare and obscure crash bug.
3277 15. Narrowed the window for a race in the daemon that could cause it to ignore
3278 SIGCHLD signals. This is not a major problem, because they are used only to
3279 wake it up if nothing else does.
3281 16. A malformed maildirsize file could cause Exim to calculate negative values
3282 for the mailbox size or file count. Odd effects could occur as a result.
3283 The maildirsize information is now recalculated if the size or filecount
3286 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3287 support for a long time. Removed HAVE_SYS_VFS_H.
3289 18. Updated exipick to current release from John Jetmore.
3291 19. Allow an empty sender to be matched against a lookup in an address list.
3292 Previously the only cases considered were a regular expression, or an
3295 20. Exim went into a mad DNS lookup loop when doing a callout where the
3296 host was specified on the transport, if the DNS lookup yielded more than
3299 21. The RFC2047 encoding function was originally intended for short strings
3300 such as real names; it was not keeping to the 75-character limit for
3301 encoded words that the RFC imposes. It now respects the limit, and
3302 generates multiple encoded words if necessary. To be on the safe side, I
3303 have increased the buffer size for the ${rfc2047: expansion operator from
3306 22. Failure to deliver a bounce message always caused it to be frozen, even if
3307 there was an errors_to setting on the router. The errors_to setting is now
3310 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3311 canonical form (fully expanded) before being placed in
3312 $sender_host_address.
3314 24. Updated eximstats to version 1.33
3316 25. Include certificate and key file names in error message when GnuTLS fails
3317 to set them up, because the GnuTLS error message doesn't include the name
3318 of the failing file when there is a problem reading it.
3320 26. Expand error message when OpenSSL has problems setting up cert/key files.
3323 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3324 (this can affect the format of dates).
3326 28. exim_tidydb, when checking for the continued existence of a message for
3327 which it has found a message-specific retry record, was not finding
3328 messages that were in split spool directories. Consequently, it was
3329 deleting retry records that should have stayed in existence.
3331 29. eximstats updated to version 1.35
3332 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3333 1.35 - bugfix such that pie charts by volume are generated correctly
3335 30. The SPA authentication driver was not abandoning authentication and moving
3336 on to the next authenticator when an expansion was forced to fail,
3337 contradicting the general specification for all authenticators. Instead it
3338 was generating a temporary error. It now behaves as specified.
3340 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3341 (the order specifies the preference for clients). The order is now AES256,
3342 AES128, 3DES, ARCFOUR128.
3344 31. Small patch to Sieve code - explicitly set From: when generating an
3347 32. Exim crashed if a remote delivery caused a very long error message to be
3348 recorded - for instance if somebody sent an entire SpamAssassin report back
3349 as a large number of 550 error lines. This bug was coincidentally fixed by
3350 increasing the size of one of Exim's internal buffers (big_buffer) that
3351 happened as part of the Exiscan merge. However, to be on the safe side, I
3352 have made the code more robust (and fixed the comments that describe what
3355 33. Some experimental protocols are using DNS PTR records for new purposes. The
3356 keys for these records are domain names, not reversed IP addresses. The
3357 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3358 leaves it alone. Component reversal etc. now happens only for IP addresses.
3361 34. The host_aton() function is supposed to be passed a string that is known
3362 to be a valid IP address. However, in the case of IPv6 addresses, it was
3363 not checking this. This is a hostage to fortune. Exim now panics and dies
3364 if the condition is not met. A case was found where this could be provoked
3365 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3366 components; fortuitously, this particular loophole had already been fixed
3367 by change 4.50/55 or 4.44/33 above.
3369 If there are any other similar loopholes, the new check in host_aton()
3370 itself should stop them being exploited. The report I received stated that
3371 data on the command line could provoke the exploit when Exim was running as
3372 exim, but did not say which command line option was involved. All I could
3373 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3374 running as the user.
3377 35. There was a buffer overflow vulnerability in the SPA authentication code
3378 (which came originally from the Samba project). I have added a test to the
3379 spa_base64_to_bits() function which I hope fixes it.
3382 36. The daemon start-up calls getloadavg() while still root for those OS that
3383 need the first call to be done as root, but it missed one case: when
3384 deliver_queue_load_max is set with deliver_drop_privilege. This is
3385 necessary for the benefit of the queue runner, because there is no re-exec
3386 when deliver_drop_privilege is set.
3388 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3389 was in use, was not putting the data itself into the right store pool;
3390 consequently, it could be overwritten for a subsequent message in the same
3391 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3394 38. Sometimes the final signoff response after QUIT could fail to get
3395 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3396 < 0 before doing a fflush(). This bug looks as though it goes back to the
3397 introduction of TLS in release 3.20, but "sometimes" must have been rare
3398 because the tests only now provoked it.
3404 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3405 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3406 Now it returns an integer. A number of calls were still expecting a T/F
3407 return. Fortuitously, in all cases, the tests worked in OK situations,
3408 which is the norm. However, things would have gone wrong on any write
3409 failures on the smtp file descriptor. This function is used when sending
3410 messages over SMTP and also when doing verify callouts.
3412 2. When Exim is called to do synchronous delivery of a locally submitted
3413 message (the -odf or -odi options), it no longer closes stderr before doing
3416 3. Implemented the mua_wrapper option.
3418 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3420 5. Implemented the functions header_remove(), header_testname(),
3421 header_add_at_position(), and receive_remove_recipient(), and exported them
3424 6. If an ACL "warn" statement specified the addition of headers, Exim already
3425 inserted X-ACL-Warn: at the start if there was no header name. However, it
3426 was not making this test for the second and subsequent header lines if
3427 there were newlines in the string. This meant that an invalid header could
3428 be inserted if Exim was badly configured.
3430 7. Allow an ACL "warn" statement to add header lines at the start or after all
3431 the Received: headers, as well as at the end.
3433 8. Added the rcpt_4xx retry error code.
3435 9. Added postmaster_mailfrom=xxx to callout verification option.
3437 10. Added mailfrom=xxxx to the callout verification option, for verify=
3440 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3441 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3443 12. Inserted some casts to stop certain compilers warning when using pointer
3444 differences as field lengths or precisions in printf-type calls (mostly
3445 affecting debugging statements).
3447 13. Added optional readline() support for -be (dynamically loaded).
3449 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3450 same clock tick as a message's arrival, so that its received time was the
3451 same as the "first fail" time on the retry record, and that message
3452 remained on the queue past the ultimate address timeout, every queue runner
3453 would try a delivery (because it was past the ultimate address timeout) but
3454 after another failure, the ultimate address timeout, which should have then
3455 bounced the address, did not kick in. This was a "< instead of <=" error;
3456 in most cases the first failure would have been in the next clock tick
3457 after the received time, and all would be well.
3459 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3460 being recognized when the domain list was tested by the match_domain
3461 condition in an expansion string.
3463 16. Added the ${str2b64: operator.
3465 17. Exim was always calling setrlimit() to set a large limit for the number of
3466 processes, without checking whether the existing limit was already
3467 adequate. (It did check for the limit on file descriptors.) Furthermore,
3468 errors from getrlimit() and setrlimit() were being ignored. Now they are
3469 logged to the main and panic logs, but Exim does carry on, to try to do its
3470 job under whatever limits there are.
3472 18. Imported PCRE 5.0.
3474 19. Trivial typo in log message " temporarily refused connection" (the leading
3477 20. If the log selector return_path_on_delivery was set and an address was
3478 redirected to /dev/null, the delivery process crashed because it assumed
3479 that a return path would always be set for a "successful" delivery. In this
3480 case, the whole delivery is bypassed as an optimization, and therefore no
3483 21. Internal re-arrangement: the function for sending a challenge and reading
3484 a response while authentication was assuming a zero-terminated challenge
3485 string. It's now changed to take a pointer and a length, to allow for
3486 binary data in such strings.
3488 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3490 23. Exim was not respecting finduser_retries when seeking the login of the
3491 uid under which it was called; it was always trying 10 times. (The default
3492 setting of finduser_retries is zero.) Also, it was sleeping after the final
3493 failure, which is pointless.
3495 24. Implemented tls_on_connect_ports.
3497 25. Implemented acl_smtp_predata.
3499 26. If the domain in control=submission is set empty, Exim assumes that the
3500 authenticated id is a complete email address when it generates From: or
3501 Sender: header lines.
3503 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3504 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3505 chown and chgrp in /bin and hostname in /usr/bin.
3507 28. Exim was keeping the "process log" file open after each use, just as it
3508 does for the main log. This opens the possibility of it remaining open for
3509 long periods when the USR1 signal hits a daemon. Occasional processlog
3510 errors were reported, that could have been caused by this. Anyway, it seems
3511 much more sensible not to leave this file open at all, so that is what now
3514 29. The long-running daemon process does not normally write to the log once it
3515 has entered its main loop, and it closes the log before doing so. This is
3516 so that log files can straightforwardly be renamed and moved. However,
3517 there are a couple of unusual error situations where the daemon does write
3518 log entries, and I had neglected to close the log afterwards.
3520 30. The text of an SMTP error response that was received during a remote
3521 delivery was being truncated at 512 bytes. This is too short for some of
3522 the long messages that one sometimes sees. I've increased the limit to
3525 31. It is now possible to make retry rules that apply only when a message has a
3526 specific sender, in particular, an empty sender.
3528 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3529 it possible to be selective about when SMTP synchronization is enforced.
3531 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3533 32. Implemented hosts_connection_nolog.
3535 33. Added an ACL for QUIT.
3537 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3540 35. Added mailbox_size and mailbox_filecount to appendfile.
3542 36. Added control = no_multiline_responses to ACLs.
3544 37. There was a bug in the logic of the code that waits for the clock to tick
3545 in the case where the clock went backwards by a substantial amount such
3546 that the microsecond fraction of "now" was more than the microsecond
3547 fraction of "then" (but the whole seconds number was less).
3549 38. Added support for the libradius Radius client library this is found on
3550 FreeBSD (previously only the radiusclient library was supported).
3556 1. When certain lookups returned multiple values in the form name=value, the
3557 quoting of the values was not always being done properly. Specifically:
3558 (a) If the value started with a double quote, but contained no whitespace,
3560 (b) If the value contained whitespace other than a space character (i.e.
3561 tabs or newlines or carriage returns) it was not quoted.
3562 This fix has been applied to the mysql and pgsql lookups by writing a
3563 separate quoting function and calling it from the lookup code. The fix
3564 should probably also be applied to nisplus, ibase and oracle lookups, but
3565 since I cannot test any of those, I have not disturbed their existing code.
3567 2. A hit in the callout cache for a specific address caused a log line with no
3568 reason for rejecting RCPT. Now it says "Previous (cached) callout
3569 verification failure".
3571 3. There was an off-by-one bug in the queryprogram router. An over-long
3572 return line was truncated at 256 instead of 255 characters, thereby
3573 overflowing its buffer with the terminating zero. As well as fixing this, I
3574 have increased the buffer size to 1024 (and made a note to document this).
3576 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3577 when Exim is waiting for an SMTP response from a remote server, Exim
3578 restarts its select() call on the socket, thereby resetting its timeout.
3579 This is not a problem when such interrupts are rare. Somebody set up a cron
3580 job to run exiwhat every 2 minutes, which is less than the normal select()
3581 timeout (5 or 10 minutes). This meant that the select() timeout never
3582 kicked in because it was always reset. I have fixed this by comparing the
3583 time when an interrupt arrives with the time at the start of the first call
3584 to select(). If more time than the timeout has elapsed, the interrupt is
3585 treated as a timeout.
3587 5. Some internal re-factoring in preparation for the addition of Sieve
3588 extensions (by MH). In particular, the "personal" test is moved to a
3589 separate function, and given an option for scanning Cc: and Bcc: (which is
3590 not set for Exim filters).
3592 6. When Exim created an email address using the login of the caller as the
3593 local part (e.g. when creating a From: or Sender: header line), it was not
3594 quoting the local part when it contained special characters such as @.
3596 7. Installed new OpenBSD configuration files.
3598 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3599 try to make them clearer.
3601 9. Callout options, other than the timeout value, were being ignored when
3602 verifying sender addresses in header lines. For example, when using
3604 verify = header_sender/callout=no_cache
3606 the cache was (incorrectly) being used.
3608 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3609 only the Cygwin environment.
3611 11. When return_path_on_delivery was set as a log selector, if different remote
3612 addresses in the same message used different return paths and parallel
3613 remote delivery occurred, the wrong values would sometimes be logged.
3614 (Whenever a remote delivery process finished, the return path value from
3615 the most recently started remote delivery process was logged.)
3617 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3618 lines when AUTH and/or TLS are in use. This is the "received protocol"
3619 field. Exim used to use "asmtp" for authenticated SMTP, without any
3620 indication (in the protocol name) for TLS use. Now it follows the RFC and
3621 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3622 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3623 names appear in log lines as well as in Received: header lines.
3625 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3626 extensions, and comparison tests, and to fix some bugs.
3628 14. Changes to the "personal" filter test:
3630 (1) The test was buggy in that it was just doing the equivalent of
3631 "contains" tests on header lines. For example, if a user's address was
3632 anne@some.where, the "personal" test would incorrectly be true for
3634 To: susanne@some.where
3636 This test is now done by extracting each address from the header in turn,
3637 and checking the entire address. Other tests that are part of "personal"
3638 are now done using regular expressions (for example, to check local parts
3639 of addresses in From: header lines).
3641 (2) The list of non-personal local parts in From: addresses has been
3642 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3643 taken from the Sieve specification recommendations.
3645 (3) If the message contains any header line starting with "List-" it is
3646 treated as non-personal.
3648 (4) The test for "circular" in the Subject: header line has been removed
3649 because it now seems ill-conceived.
3651 15. Minor typos in src/EDITME comments corrected.
3653 16. Installed latest exipick from John Jetmore.
3655 17. If headers_add on a router specified a text string that was too long for
3656 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3657 of string_sprintf() is now avoided.
3659 18. $message_body_size was not set (it was always zero) when running the DATA
3660 ACL and the local_scan() function.
3662 19. For the "mail" command in an Exim filter, no default was being set for
3663 the once_repeat time, causing a random time value to be used if "once" was
3664 specified. (If the value happened to be <= 0, no repeat happened.) The
3665 default is now 0s, meaning "never repeat". The "vacation" command was OK
3666 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3667 (I found it when inspecting the code).
3669 20. There is now an overall timeout for performing a callout verification. It
3670 defaults to 4 times the callout timeout, which applies to individual SMTP
3671 commands during the callout. The overall timeout applies when there is more
3672 than one host that can be tried. The timeout is checked before trying the
3673 next host. This prevents very long delays if there are a large number of
3674 hosts and all are timing out (e.g. when the network connections are timing
3675 out). The value of the overall timeout can be changed by specifying an
3676 additional sub-option for "callout", called "maxwait". For example:
3678 verify = sender/callout=5s,maxwait=20s
3680 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3681 to the end before writing, but this should make it even safer).
3683 22. Exim was forgetting that it had advertised PIPELINING for the second and
3684 subsequent messages on an SMTP connection. It was also not resetting its
3685 memory on STARTTLS and an internal HELO.
3687 23. When Exim logs an SMTP synchronization error within a session, it now
3688 records whether PIPELINING has been advertised or not.
3690 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3691 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3692 rather than long int.
3694 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3696 26. Added the never_mail option to autoreply.
3702 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3703 crash if the getsockname() call failed; this can happen if a connection is
3704 closed very soon after it is established. The problem was simply in the
3705 order in which certain operations were done, causing Exim to try to write
3706 to the SMTP stream before it had set up the file descriptor. The bug has
3707 been fixed by making things happen in the correct order.
3713 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3714 before the connection was closed, thus losing the rejection response.
3716 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3717 some early Solaris releases, but causes trouble in current releases where
3718 socklen_t is defined.
3720 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3723 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3726 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3727 that releasing the top of it at the end releases what was used for sub-
3728 expansions (unless the block got too big). However, discard this block if
3729 the first thing is a variable or header, so that we can use its block when
3730 it is dynamic (useful for very large $message_headers, for example).
3732 6. Lookups now cache *every* query, not just the most recent. A new, separate
3733 store pool is used for this. It can be recovered when all lookup caches are
3734 flushed. Lookups now release memory at the end of their result strings.
3735 This has involved some general refactoring of the lookup sources.
3737 7. Some code has been added to the store_xxx() functions to reduce the amount
3738 of flapping under certain conditions.
3740 8. log_incoming_interface used to affect only the <= reception log lines. Now
3741 it causes the local interface and port to be added to several more SMTP log
3742 lines, for example "SMTP connection from", and rejection lines.
3744 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3746 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3748 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3749 could be overwritten at the end of the current message (or the start of a
3750 new message if it was set in a HELO ACL). The value is now preserved for
3751 the duration of the SMTP connection.
3753 12. If a transport had a headers_rewrite setting, and a matching header line
3754 contained an unqualified address, that address was qualified, even if it
3755 did not match any rewriting rules. The underlying bug was that the values
3756 of the flags that permit the existence of unqualified sender and recipient
3757 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3758 non-local messages, and by -bnq for local messages) were not being
3759 preserved with the message after it was received.
3761 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3762 "next input=" as part of the text comprising the host identity instead of
3763 the correct text. The code was using the same buffer for two different
3764 strings. However, depending on which order the printing function evaluated
3765 its arguments, the bug did not always show up. Under Linux, for example, my
3766 test suite worked just fine.
3768 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3769 doesn't work with some older versions of Perl. It has been changed to "my",
3770 which in any case is probably the better facility to use.
3772 15. A really picky compiler found some instances of statements for creating
3773 error messages that either had too many or two few arguments for the format
3776 16. The size of the buffer for calls to the DNS resolver has been increased
3777 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3778 for addresses that have a lot of PTR records. This alleviates a problem; it
3779 does not fully solve it.
3781 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3782 buffer now truncates the list and logs the incident, which is the same
3783 action as happens when Exim is looking up a host name and its aliases.
3784 Previously in this situation something unpredictable would happen;
3785 sometimes it was "internal error: store_reset failed".
3787 18. If a server dropped the connection unexpectedly when an Exim client was
3788 using GnuTLS and trying to read a response, the client delivery process
3789 crashed while trying to generate an error log message.
3791 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3792 string, for example:
3794 warn message = H1: something\nH2: something
3796 the text was added as a single header line from Exim's point of view
3797 though it ended up OK in the delivered message. However, searching for the
3798 second and subsequent header lines using $h_h2: did not work. This has been
3799 fixed. Similarly, if a system filter added multiple headers in this way,
3800 the routers could not see them.
3802 20. Expanded the error message when iplsearch is called with an invalid key to
3803 suggest using net-iplsearch in a host list.
3805 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3806 ACLs are no longer actually imposed (and a message to that effect is
3809 22. If a "gecos" field in a passwd entry contained escaped characters, in
3810 particular, if it contained a \" sequence, Exim got it wrong when building
3811 a From: or a Sender: header from that name. A second bug also caused
3812 incorrect handling when an unquoted " was present following a character
3813 that needed quoting.
3815 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3816 was not being matched caselessly.
3818 24. Arranged for all hyphens in the exim.8 source to be escaped with
3821 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3822 properly. Recipient callout cache records were still being keyed to include
3823 the sender, even when use_sender was set false. This led to far more
3824 callouts that were necessary. The sender is no longer included in the key
3825 when use_sender is false.
3827 26. Added "control = submission" modifier to ACLs.
3829 27. Added the ${base62d: operator to decode base 62 numbers.
3831 28. dnsdb lookups can now access SRV records.
3833 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3834 the configuration file.
3836 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3837 (-v) mode. This makes the output for a verbose queue run more intelligible.
3839 31. Added a use_postmaster feature to recipient callouts.
3841 32. Added the $body_zerocount variable, containing the number of binary zero
3842 bytes in the message body.
3844 33. The time of last modification of the "new" subdirectory is now used as the
3845 "mailbox time last read" when there is a quota error for a maildir
3848 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3850 35. Added +ignore_unknown as a special item in host lists.
3852 36. Code for decoding IPv6 addresses in host lists is now included, even if
3853 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3854 address was recognized as an IP address, but was then not correctly decoded
3855 into binary, causing unexpected and incorrect effects when compared with
3862 1. Very minor rewording of debugging text in manualroute to say "list of
3863 hosts" instead of "hostlist".
3865 2. If verify=header_syntax was set, and a header line with an unqualified
3866 address (no domain) and a large number of spaces between the end of the
3867 name and the colon was received, the reception process suffered a buffer
3868 overflow, and (when I tested it) crashed. This was caused by some obsolete
3869 code that should have been removed. The fix is to remove it!
3871 3. When running in the test harness, delay a bit after writing a bounce
3872 message to get a bit more predictability in the log output.
3874 4. Added a call to search_tidyup() just before forking a reception process. In
3875 theory, someone could use a lookup in the expansion of smtp_accept_max_
3876 per_host which, without the tidyup, could leave open a database connection.
3878 5. Added the variables $recipient_data and $sender_data which get set from a
3879 lookup success in an ACL "recipients" or "senders" condition, or a router
3880 "senders" option, similar to $domain_data and $local_part_data.
3882 6. Moved the writing of debug_print from before to after the "senders" test
3885 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3886 problems for message scanning, either using a data ACL, or using
3887 local_scan() because the Received: header was not generated till after they
3888 were called (in order to set the time as the time of reception completion).
3889 I have revised the way this works. The header is now generated after the
3890 body is received, but before the ACL or local_scan() are called. After they
3891 are run, the timestamp in the header is updated.
3897 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3898 before starting a queue runner without re-exec. This happened only when
3899 deliver_drop_privilege was set or when the Exim user was set to root. The
3900 effect of the bug was that timeouts during subsequent deliveries caused
3901 crashes instead of being properly handled. The handler is now left at its
3902 default (and expected) setting.
3904 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3905 message, again when deliver_drop_privilege is set or Exim is run as root.
3906 The bug described in (1) was not present in this case, but the tidying up
3907 of the other signals was missing. I have made the two cases consistent.
3909 3. The ignore_target_hosts setting on a manualroute router was being ignored
3910 for hosts that were looked up using the /MX notation.
3912 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3915 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3916 operated on the sender address. After changing the $sender_address to <>
3917 for the sender address verify, Exim was re-instated it as the original
3918 (before rewriting) address, but remembering that it had rewritten it, so it
3919 wasn't rewriting it again. This bug also had the effect of breaking the
3920 sender address verification caching when the sender address was rewritten.
3922 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3923 This has been changed so that if the ip literal address matches
3924 ignore_target_hosts, the router declines.
3926 7. Added expansion conditions match_domain, match_address, and match_local_
3927 part (NOT match_host).
3929 8. The placeholder for the Received: header didn't have a length field set.
3931 9. Added code to Exim itself and to exim_lock to test for a specific race
3932 condition that could lead to file corruption when using MBX delivery. The
3933 issue is with the lockfile that is created in /tmp. If this file is removed
3934 after a process has opened it but before that process has acquired a lock,
3935 there is the potential for a second process to recreate the file and also
3936 acquire a lock. This could lead to two Exim processes writing to the file
3937 at the same time. The added code performs the same test as UW imapd; it
3938 checks after acquiring the lock that its file descriptor still refers to
3939 the same named file.
3941 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3942 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3945 11. Added the smtp_active_hostname option. If used, this will typically be made
3946 to depend on the incoming interface address. Because $interface_address is
3947 not set up until the daemon has forked a reception process, error responses
3948 that can happen earlier (such as "too many connections") no longer contain
3951 12. If an expansion in a condition on a "warn" statement fails because a lookup
3952 defers, the "warn" statement is abandoned, and the next ACL statement is
3953 processed. Previously this caused the whole ACL to be aborted.
3955 13. Added the iplsearch lookup type.
3957 14. Added ident_timeout as a log selector.
3959 15. Added tls_certificate_verified as a log selector.
3961 16. Added a global option tls_require_ciphers (compare the smtp transport
3962 option of the same name). This controls incoming TLS connections.
3964 17. I finally figured out how to make tls_require_ciphers do a similar thing
3965 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3966 before starting the TLS session.
3968 18. Tabs are now shown as \t in -bP output.
3970 19. If the log selector return_path_on_delivery was set, Exim crashed when
3971 bouncing a message because it had too many Received: header lines.
3973 20. If two routers both had headers_remove settings, and the first one included
3974 a superfluous trailing colon, the final name in the first list and the
3975 first name in the second list were incorrectly joined into one item (with a
3976 colon in the middle).
3982 1. Added -C and -D options to the exinext utility, mainly to make it easier
3983 to include in the automated testing, but these could be helpful when
3984 multiple configurations are in use.
3986 2. The exinext utility was not formatting the output nicely when there was
3987 an alternate port involved in the retry record key, nor when there was a
3988 message id as well (for retries that were specific to a specific message
3989 and a specific host). It was also confused by IPv6 addresses, because of
3990 the additional colons they contain. I have fixed the IPv4 problem, and
3991 patched it up to do a reasonable job for IPv6.
3993 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3994 delivery, the log line now contains "pipelined" if PIPELINING was used.
3996 4. An SMTP transport process used to panic and die if the bind() call to set
3997 an explicit outgoing interface failed. This has been changed; it is now
3998 treated in the same way as a connect() failure.
4000 5. A reference to $sender_host_name in the part of a conditional expansion
4001 that was being skipped was still causing a DNS lookup. This no longer
4004 6. The def: expansion condition was not recognizing references to header lines
4005 that used bh_ and bheader_.
4007 7. Added the _cache feature to named lists.
4009 8. The code for checking quota_filecount in the appendfile transport was
4010 allowing one more file than it should have been.
4012 9. For compatibility with Sendmail, the command line option
4020 and sets the incoming protocol and host name (for trusted callers). The
4021 host name and its colon can be omitted when only the protocol is to be set.
4022 Note the Exim already has two private options, -pd and -ps, that refer to
4023 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4024 "s", but I don't think that's a major issue.
4026 10. A number of refactoring changes to the code, none of which should affect
4029 (a) The number of logging options was getting close to filling up the
4030 32-bit word that was used as a bit map. I have split them into two classes:
4031 those that are passed in the argument to log_write(), and those that are
4032 only ever tested independently outside of that function. These are now in
4033 separate 32-bit words, so there is plenty of room for expansion again.
4034 There is no change in the user interface or the logging behaviour.
4036 (b) When building, for example, log lines, the code previously used a
4037 macro that called string_cat() twice, in order to add two strings. This is
4038 not really sufficiently general. Furthermore, there was one instance where
4039 it was actually wrong because one of the argument was used twice, and in
4040 one call a function was used. (As it happened, calling the function twice
4041 did not affect the overall behaviour.) The macro has been replaced by a
4042 function that can join an arbitrary number of extra strings onto a growing
4045 (c) The code for expansion conditions now uses a table and a binary chop
4046 instead of a serial search (which was left over from when there were very
4047 few conditions). Also, it now recognizes conditions like "pam" even when
4048 the relevant support is not compiled in: a suitably worded error message is
4049 given if an attempt is made to use such a condition.
4051 11. Added ${time_interval:xxxxx}.
4053 12. A bug was causing one of the ddress fields not to be passed back correctly
4054 from remote delivery subprocesses. The field in question was not being
4055 subsequently used, so this caused to problems in practice.
4057 13. Added new log selectors queue_time and deliver_time.
4059 14. Might have fixed a bug in maildirsizefile handling that threw up
4060 "unexpected character" debug warnings, and recalculated the data
4061 unnecessarily. In any case, I expanded the warning message to give more
4064 15. Added the message "Restricted characters in address" to the statements in
4065 the default ACL that block characters like @ and % in local parts.
4067 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4068 Three changes have been made:
4070 (a) There was a serious bug; a negative response to MAIL caused the whole
4071 recipient domain to be cached as invalid, thereby blocking all messages
4072 to all local parts at the same domain, from all senders. This bug has
4073 been fixed. The domain is no longer cached after a negative response to
4074 MAIL if the sender used is not empty.
4076 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4079 (c) A new callout option, "use_sender" has been added for people who want
4080 the modified behaviour.
4086 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4089 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4090 indeed breaks things for older releases.
4092 3. Added additional logging to the case where there is a problem reading data
4093 from a filter that is running in a subprocess using a pipe, in order to
4094 try to track down a specific problem.
4096 4. Testing facility fudge: when running in the test harness and attempting
4097 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4098 getting "No route to host". Convert this to a timeout.
4100 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4103 6. Some OS don't have socklen_t but use size_t instead. This affects the
4104 fifth argument of getsockopt() amongst other things. This is now
4105 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4106 can be set for individual OS. I have set it for SunOS5, OSF1, and
4107 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4108 some earlier ones do not.
4110 7. Change 4.30/15 was not doing the test caselessly.
4112 8. The standard form for an IPv6 address literal was being rejected by address
4113 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4114 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4115 this, as well as the form without the "IPv6" on the front (but only when
4116 address literals are enabled, of course).
4118 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4120 10. Exim crashed if a message with an empty sender address specified by -f
4121 encountered a router with an errors_to setting. This could be provoked only
4122 by a command such as
4126 where an empty string was supplied; "<>" did not hit this bug.
4128 11. Installed PCRE release 4.5.
4130 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4131 remained set. It is now erased.
4133 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4134 times from message ids (which are base 36 rather than the normal 62).
4136 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4137 were being counted as actual protocol errors, and logged if the log
4138 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4139 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4140 and DATA following a set of rejected RCPTs do not count as protocol errors.
4141 In other words, Exim assumes they were pipelined, though this may not
4142 actually be the case. Of course, in all cases the client gets an
4143 appropriate error code.
4145 15. If a lookup fails in an ACL condition, a message about the failure may
4146 be available; it is used if testing the ACL cannot continue, because most
4147 such messages specify what the cause of the deferral is. However, some
4148 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4149 that caused an old message to be retained and used if a later statement
4150 caused a defer, replacing the real cause of the deferral.
4152 16. If an IP address had so many PTR records that the DNS lookup buffer
4153 was not large enough to hold them, Exim could crash while trying to process
4154 the truncated data. It now detects and logs this case.
4156 17. Further to 4.21/58, another change has been made: if (and only if) the
4157 first line of a message (the first header line) ends with CRLF, a bare LF
4158 in a subsequent header line has a space inserted after it, so as not to
4159 terminate the header.
4161 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4162 unnecessarily, used atoi() instead of strtol(), and didn't check the
4163 termination when getting file sizes from file names by regex.
4165 19. Completely re-implemented the support for maildirsize files, in the light
4166 of a number of problems with the previous contributed implementation
4167 (4.30/29). In particular:
4169 . If the quota is zero, the maildirsize file is maintained, but no quota is
4172 . If the maildir directory does not exist, it is created before any attempt
4173 to write a maildirsize file.
4175 . The quota value in the file is just a cache; if the quota is changed in
4176 the transport, the new value overrides.
4178 . A regular expression is available for excluding directories from the
4181 20. The autoreply transport checks the characters in options that define the
4182 message's headers; it allows continued headers, but it was checking with
4183 isspace() after an embedded newline instead of explicitly looking for a
4186 21. If all the "regular" hosts to which an address was routed had passed their
4187 expiry times, and had not reached their retry times, the address was
4188 bounced, even if fallback hosts were defined. Now Exim should go on to try
4191 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4192 equivalent code in the SMTP transport. Some hosts send humungous responses
4193 to HELO/EHLO, more than 1024 it seems.
4195 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4196 gives compiler warnings in some environments. I've now done it "properly",
4199 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4200 (because of problems with the built-in one) was declared to return uschar *
4201 instead of char *, causing compiler failure.
4203 25. Fixed a file descriptor leak when processing alias/forward files.
4205 26. Fixed a minor format string issue in dbfn.c.
4207 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4209 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4210 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4213 29. When a sender address is verified, it is cached, to save repeating the test
4214 when there is more than one recipient in a message. However, when the
4215 verification involves a callout, it is possible for different callout
4216 options to be set for different recipients. It is too complicated to keep
4217 track of this in the cache, so now Exim always runs a verification when a
4218 callout is required, relying on the callout cache for the optimization.
4219 The overhead is duplication of the address routing, but this should not be
4222 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4223 to be verified with callout=postmaster, and the main callout worked but the
4224 postmaster check failed, the verification correctly failed. However, if a
4225 subsequent RCPT command asked for sender verification *without* the
4226 postmaster check, incorrect caching caused this verification also to fail,
4229 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4230 it was not caching the DNS options (qualify_single, search_parents) that
4231 were used when the lookup failed. A subsequent lookup with different
4232 options therefore always gave the same answer, though there were cases
4233 where it should not have. (Example: a "domains = !$mx_any" option on a
4234 dnslookup router: the "domains" option is always processed without any
4235 widening, but the router might have qualify_single set.) Now Exim uses the
4236 cached value only when the same options are set.
4238 32. Added John Jetmore's "exipick" utility to the distribution.
4240 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4241 than a timeout (e.g. a certificate is required, and is not provided), an
4242 Exim server now closes the connection immediately. Previously it waited for
4243 the client to close - but if the client is SSL, it seems that they each
4244 wait for each other, leading to a delay before one of them times out.
4246 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4247 maintained 0.8.x compatibility because I don't think many are using it, and
4248 it is clearly obsolete.
4250 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4253 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4254 client certificate was expired. A simple patch fixes this, though I don't
4255 understand the full logic of why the verify callback is called multiple
4258 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4259 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4260 which causes problems with some clients (such as the Certicom SSL Plus
4261 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4262 disables the coutermeasure allowing Eudora to connect."
4264 38. Exim was not checking that a write() to a log file succeeded. This could
4265 lead to Bad Things if a log got too big, in particular if it hit a file
4266 size limit. Exim now panics and dies if it cannot write to a log file, just
4267 as it does if it cannot open a log file.
4269 39. Modified OS/Makefile-Linux so that it now contains
4271 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4273 The two -D definitions ensure that Exim is compiled with large file
4274 support, which makes it possible to handle log files that are bigger than
4277 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4278 instance) a domain was checked against a named list that involved a lookup,
4279 causing $domain_data to be set, then another domain was checked against the
4280 same list, then the first domain was re-checked, the value of $domain_data
4281 after the final check could be wrong. In particular, if the second check
4282 failed, it could be set empty. This bug probably also applied to
4285 41. The strip_trailing_dot option was not being applied to the address given
4286 with the -f command-line option.
4288 42. The code for reading a message's header from the spool was incrementing
4289 $received_count, but never initializing it. This meant that the value was
4290 incorrect (doubled) while delivering a message in the same process in which
4291 it was received. In the most common configuration of Exim, this never
4292 happens - a fresh exec is done - but it can happen when
4293 deliver_drop_privilege is set.
4295 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4296 it now includes up to 150 characters of the unexpected data in the log
4299 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4300 and building data strings. The size of both of these buffers was 10 000
4301 bytes - far larger than anybody would *ever* want, thought I. Needless to
4302 say, somebody hit the limit. I have increased the maximum line length to
4303 20 000 and the maximum data length of concatenated lines to 100 000. I have
4304 also fixed two bugs, because there was no checking on these buffers. Tsk,
4305 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4306 buffer is too small.
4308 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4309 lsearch lookups. Now it does.
4311 46. When parsing a route_list item in a manualroute router, a fixed-length
4312 buffer was used for the list of hosts. I made this 1024 bytes long,
4313 thinking that nobody would ever have a list of hosts that long. Wrong.
4314 Somebody had a whole pile of complicated expansion conditions, and the
4315 string was silently truncated, leading to an expansion error. It turns out
4316 that it is easier to change to an unlimited length (owing to other changes
4317 that have happened since this code was originally written) than to build
4318 structure for giving a limitation error. The length of the item that
4319 expands into the list of hosts is now unlimited.
4321 47. The lsearch lookup could not handle data where the length of text line was
4322 more than 4095 characters. Such lines were truncated, leading to shortened
4323 data being returned. It should now handle lines of any length.
4325 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4326 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4329 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4330 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4331 no longer happens in comment lines. A list of replacements is now placed
4332 at the head of all of the source files, except those whose only change is
4333 to replace PERL_COMMAND in the very first #! line.
4335 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4336 messages on the queue, with a bottom-up merge sort, using code contributed
4337 by Michael Haardt. This should make operations like -bp somewhat faster on
4338 large queues. It won't affect queue runners, except when queue_run_in_order
4341 51. Installed eximstats 1.31 in the distribution.
4343 52. Added support for SRV lookups to the dnslookup router.
4345 53. If an ACL referred to $message_body or $message_body_end, the value was not
4346 reset for any messages that followed in the same SMTP session.
4348 54. The store-handling optimization for building very long strings was not
4349 differentiating between the different store pools. I don't think this
4350 actually made any difference in practice, but I've tidied it.
4352 55. While running the routers to verify a sender address, $sender_address
4353 was still set to the sender address. This is wrong, because when routing to
4354 send a bounce to the sender, it would be empty. Therefore, I have changed
4355 it so that, while verifying a sender address, $sender_address is set to <>.
4356 (There is no change to what happens when verifying a recipient address.)
4358 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4359 target A or AAAA records (if not already returned) without resetting the
4360 qualify_single or search_parents options of the DNS resolver. These are
4361 inappropriate in this case because the targets of MX and SRV records must
4362 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4363 target that, when qualified, matched something in the local domain. These
4364 two options are now turned off when doing these lookups.
4366 57. It seems that at least some releases of Reiserfs (which does not have the
4367 concept of a fixed number of inodes) returns zero and not -1 for the
4368 number of available inodes. This interacted badly with check_spool_inodes,
4369 which assumed that -1 was the "no such thing" setting. What I have done is
4370 to check that the total number of inodes is greater than zero before doing
4371 the test of how many are available.
4373 58. When a "warn" ACL statement has a log_message modifier, the message is
4374 remembered, and not repeated. This is to avoid a lot of repetition when a
4375 message has many recipients that cause the same warning to be written.
4376 Howewer, Exim was preserving the list of already written lines for an
4377 entire SMTP session, which doesn't seem right. The memory is now reset if a
4378 new message is started.
4380 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4381 debugging output unless log_rewrite was also set.
4383 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4384 of (int)(handle) when we know that handle contains (void *)(-1).
4386 61. The Exim daemon panic-logs an error return when it closes the incoming
4387 connection. However "connection reset by peer" seems to be common, and
4388 isn't really an error worthy of noting specially, so that particular error
4391 62. When Exim is trying to find all the local interfaces, it used to panic and
4392 die if the ioctl to get the interface flags failed. However, it seems that
4393 on at least one OS (Solaris 9) it is possible to have an interface that is
4394 included in the list of interfaces, but for which you get a failure error
4395 for this call. This happens when the interface is not "plumbed" into a
4396 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4397 failure of the "get flags" call assumes that the interface is down.
4399 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4400 makes life easier for people who are doing arithmetic on fields extracted
4401 from dates, where you often get leading zeros that should not be
4402 interpreted as octal.
4404 64. Added qualify_domain to the redirect router, to override the global
4407 65. If a pathologically long header line contained very many addresses (the
4408 report of this problem mentioned 10 000) and each of them was rewritten,
4409 Exim could use up a very large amount of memory. (It kept on making new
4410 copies of the header line as it rewrote, and never released the old ones.)
4411 At the expense of a bit more processing, the header rewriting function has
4412 been changed so that it no longer eats memory in this way.
4414 66. The generation of the Received: header has been moved from the time that a
4415 message starts to be received, to the time that it finishes. The timestamp
4416 in the Received: header should now be very close to that of the <= log
4417 line. There are two side-effects of this change:
4419 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4420 logged header lines no longer include the local Received: line, because
4421 it has not yet been created. The same applies to a copy of the message
4422 that is returned to a non-SMTP sender when a message is rejected.
4424 (b) When a filter file is tested using -bf, no additional Received: header
4425 is added to the test message. After some thought, I decided that this
4428 This change does not affect the value of $received_for. It is still set
4429 after address rewriting, but before local_scan() is called.
4431 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4433 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4434 gave an unhelpful panic error message, and a defer error. I have managed to
4435 change this behaviour so that it now rejects any supplied certificate,
4436 which seems right, as the list of acceptable certificates is empty.
4438 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4439 gave an unhelpful defer error. I have not managed to make this reject any
4440 supplied certificates, but the error message it gives is "no certificate
4441 supplied", which is not helpful.
4443 70. exigrep's output now also includes lines that are not associated with any
4444 message, but which match the given pattern. Implemented by a patch from
4445 Martin Sluka, which also tidied up the Perl a bit.
4447 71. Recipient callout verification, like sender verification, was using <> in
4448 the MAIL FROM command. This isn't really the right thing, since the actual
4449 sender may affect whether the remote host accepts the recipient or not. I
4450 have changed it to use the actual sender in the callout; this means that
4451 the cache record is now keyed on a recipient/sender pair, not just the
4452 recipient address. There doesn't seem to be a real danger of callout loops,
4453 since a callout by the remote host to check the sender would use <>.
4454 [SEE ABOVE: changed after hitting problems.]
4456 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4457 temporary errors. However, in the case of such a code being given after
4458 the end of a data transmission (i.e. after ".") Exim was failing to write
4459 a retry record for the message. (Yes, there was some broken host that was
4460 actually sending 8xx at this point.)
4462 73. An unknown lookup type in a host list could cause Exim to panic-die when
4463 the list was checked. (An example that provoked this was putting <; in the
4464 middle of a list instead of at the start.) If this happened during a DATA
4465 ACL check, a -D file could be left lying around. This kind of configuration
4466 error no longer causes Exim to die; instead it causes a defer errror. The
4467 incident is still logged to the main and panic logs.
4469 74. Buglet left over from Exim 3 conversion. The message "too many messages
4470 in one connection" was written to the rejectlog but not the mainlog, except
4471 when address rewriting (yes!) was being logged.
4473 75. Added write_rejectlog option.
4475 76. When a system filter was run not as root (that is, when system_filter_user
4476 was set), the values of the $n variables were not being returned to the
4477 main process; thus, they were not subsequently available in the $sn
4480 77. Added +return_path_on_delivery log selector.
4482 78. A connection timeout was being treated differently from recipients deferred
4483 when testing hosts_max_try with a message that was older than the host's
4484 retry timeout. (The host should not be counted, thus allowing all hosts to
4485 be tried at least once before bouncing.) This may have been the cause of an
4486 occasionally reported bug whereby a message would remain on the queue
4487 longer than the retry timeout, but would be bounced if a delivery was
4488 forced. I say "may" because I never totally pinned down the problem;
4489 setting up timeout/retry tests is difficult. See also the next item.
4491 79. The ultimate address timeout was not being applied to errors that involved
4492 a combination of host plus message (for example, a timeout on a MAIL
4493 command). When an address resolved to a number of possible hosts, and they
4494 were not all tried for each delivery (e.g. because of hosts_max_try), a
4495 message could remain on the queue longer than the retry timeout.
4497 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4500 81. Fixed an obscure SMTP outgoing bug which required at least the following
4501 conditions: (a) there was another message waiting for the same server;
4502 (b) the server returned 5xx to all RCPT commands in the first message so
4503 that the message was not completed; (c) the server dropped the connection
4504 or gave a negative response to the RSET that Exim sends to abort the
4505 transaction. The observed case was a dropped connection after DATA that had
4506 been sent in pipelining mode. That is, the server had advertised PIPELINING
4507 but was not implementing it correctly. The effect of the bug was incorrect
4508 behaviour, such as trying another host, and this could lead to a crash.
4514 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4515 and daemon.c were passed as pointers to ints; they should have been
4516 pointers to socklen_t variables (which are typically unsigned ints).
4518 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4521 3. Fixed a really odd bug that affected only the testing scheme; patching a
4522 certain fixed string in the binary changed the value of another string that
4523 happened to be identical to the end of the original first string.
4525 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4526 name", it returns that address as the IP address. On some operating
4527 systems (e.g. Solaris), it also passes back the IP address string as the
4528 "host name". However, on others (e.g. Linux), it passes back an empty
4529 string. Exim wasn't checking for this, and was changing the host name to an
4530 empty string, assuming it had been canonicized.
4532 5. Although rare, it is permitted to have more than one PTR record for a given
4533 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4534 all the names associated with an address, because they do in Solaris.
4535 However, it seems that they do not in Linux for data that comes from the
4536 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4537 I found this out when I moved to a new Linux workstation and tried to run
4538 the Exim test suite.
4540 To get round this problem I have changed the code so that it now does its
4541 own call to the DNS to look up PTR records when searching for a host name.
4542 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4543 addresses that are only in /etc/hosts are still found.
4545 This behaviour is, however, controlled by an option called host_lookup_
4546 order, which defaults to "bydns:byaddr". If people want to use the other
4547 order, or indeed, just use one or the other means of lookup, they can
4548 specify it in this variable.
4550 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4551 some operating systems, this comes back from gethostbyaddr() as an empty
4552 string, and this is what Exim used to test for. However, it seems that in
4553 other systems, "." is yielded. Exim now tests for this case too.
4555 7. The values of check_spool_space and check_log_space are now held internally
4556 as a number of kilobytes instead of an absolute number of bytes. If a
4557 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4558 kilobyte. This means that much larger values can be stored.
4560 8. Exim monitor: an attempt to get the action menu when not actually pointing
4561 at a message produces an empty menu entitled "No message selected". This
4562 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4563 no entries in it ("Shell widget menu has zero width and/or height"). So I
4564 have added a single, blank menu entry in this case.
4566 9. Added ${quote_local_part.
4568 10. MIME decoding is now applied to the contents of Subject: header lines when
4571 11. Now that a reference to $sender_host_address automatically causes a reverse
4572 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4573 host lookup before query-style lookups in lists that might use this
4574 variable. This has therefore been abolished, and the "net-" prefix is no
4575 longer necessary for query-style lookups.
4577 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4578 have been a typo for LFLAGS, so it has been changed.
4580 13. The install script calls Exim with "-C /dev/null" in order to find the
4581 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4582 to be output. Howeve, since Exim outputs its version number before the
4583 error, it didn't break the script. It just looked ugly. I fixed this by
4584 always allowing "-C /dev/null" if the caller is root.
4586 14. Ignore overlarge ACL variable number when reading spool file - insurance
4587 against a later release with more variables having written the file.
4589 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4590 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4591 this, as well as the form without the "IPv6" on the front.
4593 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4594 OS/Makefile-Darwin file.
4596 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4597 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4599 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4600 variable. However, some OS use ints for inodes. I've added cast to long int
4601 to get rid of the compiler warning.
4603 19. I had forgotten to lock out "/../" in configuration file names when
4604 ALT_CONFIG_PREFIX was set.
4606 20. Routers used for verification do not need to specify transports. However,
4607 if such a router generated a host list, and callout was configured, Exim
4608 crashed, because it could not find a port number from the (non-existent)
4609 transport. It now assumes port 25 in this circumstance.
4611 21. Added the -t option to exigrep.
4613 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4614 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4615 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4616 them is set, LOOKUP_LSEARCH is forced.
4618 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4620 24. Added sender and host information to the "rejected by local_scan()" log
4621 line; previously there was no indication of these.
4623 25. Added .include_if_exists.
4625 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4626 Linux. It turns out that not all file systems support this. Apparently some
4627 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4628 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4629 not supported on the file descriptor, is now ignored when Exim is trying to
4630 sync a directory. This applies only to Linux.
4632 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4634 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4635 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4638 29. Added support for maildirsize files from supplied patch (modified a bit).
4640 30. The use of :fail: followed by an empty string could lead Exim to respond to
4641 sender verification failures with (e.g.):
4643 550 Verification failed for <xxx>
4644 550 Sender verify failed
4646 where the first response line was missing the '-' that indicates it is not
4647 the final line of the response.
4649 31. The loop for finding the name of the user that called Exim had a hardwired
4650 limit of 10; it now uses the value of finduser_retries, which is used for
4651 all other user lookups.
4653 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4656 33. Exim was neglecting to zero errno before one call of strtol() when
4657 expanding a string and expecting an integer value. On some systems this
4658 resulted in spurious "integer overflow" errors. Also, it was casting the
4659 result into an int without checking.
4661 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4662 did not work. The code looks as if it has *never* worked, though it appears
4663 to have been documented since at least releast 1.62. I have made it work.
4665 35. The "timeout_DNS" error in retry rules, also documented since at least
4666 1.62, also never worked. As it isn't clear exactly what this means, and
4667 clearly it isn't a major issue, I have abolished the feature by treating it
4668 as "timeout", and writing a warning to the main and panic logs.
4670 36. The display of retry rules for -brt wasn't always showing the error code
4673 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4674 timeout_connect_A, timeout_connect_MX.
4676 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4677 to the empty sender.
4679 39. The daemon was not analysing the content of -oX till after it had closed
4680 stderr and disconnected from the controlling terminal. This meant that any
4681 syntax errors were only noted on the panic log, and the return code from
4682 the command was 0. By re-arranging the code a little, I've made the
4683 decoding happen first, so such errors now appear on stderr, and the return
4684 code is 1. However, the actual setting up of the sockets still happens in
4685 the disconnected process, so errors there are still only recorded on the
4688 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4689 connections (as happens on some IP stacks) was logged at start up time as
4690 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4691 it from "IPv6 and IPv4", which means that two separate sockets are being
4694 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4695 says whether AF_INET or AF_INET6 was passed as an argument.
4697 42. Exiwhat output was messed up when time zones were included in log
4700 43. Exiwhat now gives more information about the daemon's listening ports,
4701 and whether -tls-on-connect was used.
4703 44. The "port" option of the smtp transport is now expanded.
4705 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4706 silently ignored. Now an error message is written to the main and panic
4709 46. There's a new ACL modifier called "logwrite" which writes to a log file
4710 as soon as it is encountered.
4712 47. Added $local_user_uid and $local_user_gid at routing time.
4714 48. Exim crashed when trying to verify a sender address that was being
4717 49. Exim was recognizing only a space character after ".include". It now also
4718 recognizes a tab character.
4720 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4721 extracting the relevant information from the specification. The man page no
4722 longer contains scrambled data for the -d option, and I've added a section
4723 at the front about calling Exim under different names.
4725 51. Added "extra_headers" argument to the "mail" command in filter files.
4727 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4730 53. Installed eximstats 1.29.
4732 54. Added transport_filter_timeout as a generic transport option.
4734 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4735 Cc: header lines. This was required by RFC 822, but it not required by RFC
4738 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4739 incoming messages that did not have them. Now it does so only if the
4740 message originates locally, that is, if there is no associated remote host
4741 address. When Resent- header lines are present, this applies to the Resent-
4742 lines rather than the non-Resent- lines.
4744 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4745 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4747 58. Messages for configuration errors now include the name of the main
4748 configuration file - useful now that there may be more than one file in a
4749 list (.included file names were always shown).
4751 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4752 for those rare installations that do not start the daemon as root or run it
4753 setuid root. I've cut out the call to initgroups() if the daemon is not
4756 60. The Exim user and group can now be bound into the binary as text strings
4757 that are looked up at the start of Exim's processing.
4759 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4761 62. Added $mailstore_basename variable.
4763 63. Installed patch to sieve.c from Michael Haardt.
4765 64. When Exim failed to open the panic log after failing to open the main log,
4766 the original message it was trying to log was written to stderr and debug
4767 output, but if they were not available (the usual case in production), it
4768 was lost. Now it is written to syslog before the two lines that record the
4769 failures to open the logs.
4771 65. Users' Exim filters run in subprocesses under the user's uid. It is
4772 possible for a "deliver" command or an alias in a "personal" command to
4773 provoke an address rewrite. If logging of address rewriting is configured,
4774 this fails because the process is not running as root or exim. There may be
4775 a better way of dealing with this, but for the moment (because 4.30 needs
4776 to be released), I have disabled address rewrite logging when running a
4777 filter in a non-root, non-exim process.
4783 1. The buildconfig auxiliary program wasn't quoting the value set for
4784 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4785 not defined. This bug was masked in 4.22 by the effect that was fixed in
4788 2. Some messages that were rejected after a message id was allocated were
4789 shown as "incomplete" by exigrep. It no longer does this for messages that
4790 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4792 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4793 have allow_domain_literals set, the ID did not get logged in the <= line.
4794 Domain literals are now always recognized in Message-ID: header lines.
4796 4. The first argument for a ${extract expansion item is the key name or field
4797 number. Leading and trailing spaces in this item were not being ignored,
4798 causing some misleading effects.
4800 5. When deliver_drop_privilege was set, single queue runner processes started
4801 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4802 same command in the process it spins off) were not dropping privilege.
4804 6. When the daemon running as "exim" started a queue runner, it always
4805 re-executed Exim in the spun-off process. This is a waste of effort when
4806 deliver_drop_privilege is set. The new process now just calls the
4807 queue-runner function directly.
4813 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4816 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4817 ignored. Though the use of -oP was forcing the writing of a pid file, it
4818 was always written to the default place.
4820 3. If the message "no IP address found for host xxxx" is generated during
4821 incoming verification, it is now followed by identification of the incoming
4822 connection (so you can more easily find what provoked it).
4824 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4826 5. Added some features to "harden" Exim a bit more against certain attacks:
4828 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4829 be put in Local/Makefile. This is like the never_users runtime option,
4830 but it cannot be overridden. The default setting is "root".
4832 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4833 prefix string with which any file named in a -C command line option
4836 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4837 is retained for -C and -D only if the caller of Exim is root. Without
4838 it, the exim user may also use -C and -D and retain privilege.
4840 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4841 command line option is disabled.
4843 6. Macro names set by the -D option must start with an upper case letter, just
4844 like macro names defined in the configuration file.
4846 7. Added "dereference=" facility to LDAP.
4848 8. Two instances of the typo "uknown" in the source files are fixed.
4850 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4851 the Configure-Makefile script screwed up while processing it.
4853 10. Incorporated PCRE 4.4.
4855 11. The SMTP synchronization check was not operating right at the start of an
4856 SMTP session. For example, it could not catch a HELO sent before the client
4857 waited for the greeting. There is now a check for outstanding input at the
4858 point when the greeting is written. Because of the duplex, asynchronous
4859 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4860 way, but not yet received, when the check is performed.
4862 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4863 on TCP/IP sockets, because this apparently causes some broken clients to
4866 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4867 unchanged) from the Cygwin maintainer.
4869 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4870 was supported instead of testing for mbx. Effectively a typo.
4872 15. The spa authenticator server code was not checking that the input it
4873 received was valid base64.
4875 16. The debug output line for the "set" modifier in ACLs was not showing the
4876 name of the variable that was being set.
4878 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4880 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4881 lookup on its own. Something else was needed to trigger the lookup. For
4882 example, a match in host_lookup or the need for a host name in a host list.
4883 Now, if $sender_host_name is referenced and the host name has not yet been
4884 looked up, a lookup is performed. If the lookup fails, the variable remains
4885 empty, and $host_lookup_failed is set to "1".
4887 19. Added "eqi" as a case-independent comparison operator.
4889 20. The saslauthd authentication condition could segfault if neither service
4890 nor realm was specified.
4892 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4893 error message that was logged was misleading, and incoming SMTP
4894 connections were dropped. The message is now more accurate, and temporary
4895 errors are given to SMTP connections.
4897 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4898 (except RSET) with a 421 error, until QUIT is received. However, it was
4899 failing to send a response to QUIT.
4901 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4902 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4903 ACL is now run after the helo_try_verify_hosts code.
4905 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4906 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4907 case-independent, so other case variants are also recognized). Apparently
4908 some systems use these upper case variants.
4910 25. If more than two messages were waiting for the same host, and a transport
4911 filter was specified for the transport, Exim sent two messages over the
4912 same TCP/IP connection, and then failed with "socket operation on non-
4913 socket" when it tried to send the third.
4915 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4917 27. The extern definition of crypt16() in expand.c was not being excluded when
4918 the OS had its own crypt16() function.
4920 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4921 as a preferred synonym for return_size_limit, both as an option and as an
4924 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4926 30. Changed the default configuration ACL to relax the local part checking rule
4927 for addresses that are not in any local domains. For these addresses,
4928 slashes and pipe symbols are allowed within local parts, but the sequence
4929 /../ is explicitly forbidden.
4931 31. SPA server authentication was not clearing the challenge buffer before
4934 32. log_message in a "warn" ACL statement was writing to the reject log as
4935 well as to the main log, which contradicts the documentation and doesn't
4936 seem right (because no rejection is happening). So I have stopped it.
4938 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4939 However, I am unable to do any testing of this.
4941 34. Fixed an infelicity in the appendfile transport. When checking directories
4942 for a mailbox, to see if any needed to be created, it was accidentally
4943 using path names with one or more superfluous leading slashes; tracing
4944 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4946 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4947 contents are added to the log line that is written for every discarded
4948 recipient. (Previously a log_message setting was ignored.)
4950 36. The ${quote: operator now quotes the string if it is empty.
4952 37. The install script runs exim in order to find its version number. If for
4953 some reason other than non-existence or emptiness, which it checks, it
4954 could not run './exim', it was installing it with an empty version number,
4955 i.e. as "exim-". This error state is now caught, and the installation is
4958 38. An argument was missing from the function that creates an error message
4959 when Exim fails to connect to the socket for saslauthd authentication.
4960 This could cause Exim to crash, or give a corrupted message.
4962 39. Added isip, isip4, and isip6 to ${if conditions.
4964 40. The ACL variables $acl_xx are now saved with the message, and can be
4965 accessed later in routers, transports, and filters.
4967 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4968 strings in the file are not string-expanded.
4970 42. If a MAIL command specified a SIZE value that was too large to fit into an
4971 int variable, the check against message_size_limit failed. Such values are
4972 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4973 day this will have to be increased, but I don't think I want to be around
4974 when emails are that large.
4981 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4982 iconv() is not standard in FreeBSD.
4984 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4985 IPv6 enabled. The observed symptom was a segmentation fault on return from
4986 the function os_common_find_running_interfaces() in src/os.c.
4988 3. In the check_special_case() function in daemon.c I had used "errno" as an
4989 argument name, which causes warnings on some systems. This was basically a
4990 typo, since it was named "eno" in the comments!
4992 4. The code that waits for the clock to tick (at a resolution of some fraction
4993 of a second) so as to ensure message-id uniqueness was always waiting for
4994 at least one whole tick, when it could have waited for less. [This is
4995 almost certainly not relevant at current processor speeds, where it is
4996 unlikely to ever wait at all. But we try to future-proof.]
4998 5. The function that sleeps for a time interval that includes fractions of a
4999 second contained a race. It did not block SIGALRM between setting the
5000 timer, and suspending (a couple of lines later). If the interval was short
5001 and the sigsuspend() was delayed until after it had expired, the suspension
5002 never ended. On busy systems this could lead to processes getting stuck for
5005 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5006 process, before it forks any delivery processes. The open lookup caching
5007 mechanism meant that the open file or database connection was passed into
5008 the delivery process. The problem was that delivery processes always tidy
5009 up cached lookup data. This could cause a problem for the next delivery
5010 process started by the queue runner, because the external queue runner
5011 process does not know about the closure. So the next delivery process
5012 still has data in the lookup cache. In the case of a file lookup, there was
5013 no problem because closing a file descriptor in a subprocess doesn't affect
5014 the parent. However, if the lookup was caching a connection to a database,
5015 the connection was closed, and the second delivery process was likely to
5016 see errors such as "PGSQL: query failed: server closed the connection
5017 unexpectedly". The problem has been fixed by closing all cached lookups
5018 in a queue runner before running a delivery process.
5020 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5021 seem to have the "const" qualifier which it has on other OS. I've
5024 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5025 *of the same type* (client or server) with the same public name that an
5026 error should be diagnosed.
5028 9. When Exim looked up a host name for an IP address, but failed to find the
5029 original IP address when looking up the host name (a safety check), it
5030 output the message "<ip address> does not match any IP for NULL", which was
5031 confusing, to say the least. The bug was that the host name should have
5032 appeared instead of "NULL".
5034 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5035 user that is built into the binary, and the -C or -D options is used, root
5036 privilege is dropped before the configuration file is read. In addition,
5037 logging is switched to stderr instead of the normal log files. If the
5038 configuration then re-defines the Exim user, the unprivileged environment
5039 is probably not what is expected, so Exim logs a panic warning message (but
5042 However, if deliver_drop_privilege is set, the unprivileged state may well
5043 be exactly what is intended, so the warning has been cut out in that case,
5044 and Exim is allowed to try to write to its normal log files.
5050 1. smtp_return_error_details was not giving details for temporary sender
5051 or receiver verification errors.
5053 2. Diagnose a configuration error if two authenticators have the same public
5056 3. Exim used not to create the message log file for a message until the first
5057 delivery attempt. This could be confusing when incoming messages were held
5058 for policy or load reasons. The message log file is now created at the time
5059 the message is received, and an initial "Received" line is written to it.
5061 4. The automatically generated man page for command line options had a minor
5062 bug that caused no ill effects; however, a more serious problem was that
5063 the procedure for building the man page automatically didn't always
5064 operate. Consequently, release 4.20 contains an out-of-date version. This
5065 shouldn't happen again.
5067 5. When building Exim with embedded Perl support, the script that builds the
5068 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5069 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5071 6. The freeze_tell option was not being used for messages that were frozen on
5072 arrival, either by an ACL or by local_scan().
5074 7. Added the smtp_incomplete_transaction log selector.
5076 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5077 was accepting AUTH without a new EHLO.
5079 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5080 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5083 10. Logging of TCP/IP connections (when configured) now happens in the main
5084 daemon process instead of the child process, so that the TCP/IP connection
5085 count is more accurate (but it can never be perfect).
5087 11. The use of "drop" in a nested ACL was not being handled correctly in the
5088 outer ACL. Now, if condition failure induced by the nested "drop" causes
5089 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5090 or "require"), the connection is dropped.
5092 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5093 that yield "discard" can now be used with an "accept" or a "discard" verb,
5094 but an error is generated for any others (because I can't see a useful way
5095 to define what should happen).
5097 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5098 are now processed in the same way as lines in the Exim configuration file.
5099 In particular, continuation lines are supported.
5101 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5103 15. Added -ti meaning -t -i.
5105 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5106 domains, and warn by logging if others are found.
5108 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5109 returned from GIFCONF when Exim is trying to find the list of interfaces on
5110 a host. The code in os.c has been modified to copy each ifreq to an aligned
5111 structure in all cases.
5113 Also, in some cases, the returned ifreq's were being copied to a 'struct
5114 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5115 means the last couple of bytes of an IPv6 address could be chopped if the
5116 ifreq contained only a normal sockaddr (14 bytes storage).
5118 18. Named domain lists were not supported in the hosts_treat_as_local option.
5119 An entry such as +xxxx was not recognized, and was treated as a literal
5122 19. Ensure that header lines added by a DATA ACL are included in the reject log
5123 if the ACL subsequently rejects the message.
5125 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5126 MD5 (which is deprecated).
5128 21. When testing a filter file using -bf, Exim was writing a message when it
5129 took the sender from a "From " line in the message, but it was not doing so
5130 when it took $return_path from a Return-Path: header line. It now does.
5132 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5133 with a valid header line field name (a series of printing characters
5134 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5136 23. Changed "disc" in the source to "disk" to conform to the documentation and
5137 the book and for uniformity.
5139 24. Ignore Sendmail's -Ooption=value command line item.
5141 25. When execve() failed while trying to run a command in a pipe transport,
5142 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5143 could be confused with a return value of 69 from the command itself. This
5144 has been changed to 127, the value the shell returns if it is asked to run
5145 a non-existent command. The wording for the related log line suggests a
5146 non-existent command as the problem.
5148 26. If received_header_text expands to an empty string, do not add a Received:
5149 header line to the message. (Well, it adds a token one on the spool, but
5150 marks it "old" so that it doesn't get used or transmitted.)
5152 27. Installed eximstats 1.28 (addition of -nt option).
5154 28. There was no check for failure on the call to getsockname() in the daemon
5155 code. This can fail if there is a shortage of resources on the system, with
5156 ENOMEM, for example. A temporary error is now given on failure.
5158 29. Contrary to the C standard, it seems that in some environments, the
5159 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5160 program. Exim now does this explicitly; it affects the formatting of
5161 timestamps using strftime().
5163 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5164 complaints. I've now initialized all the variables, to avoid this.
5166 32. Header lines added by a system filter were not being "seen" during
5167 transport-time rewrites.
5169 33. The info_callback() function passed to OpenSSL is set up with type void
5170 (*)(SSL *, int, int), as described somewhere. However, when calling the
5171 function (actually a macro) that sets it up, the type void(*)() is
5172 expected. I've put in a cast to prevent warnings from picky compilers.
5174 34. If a DNS black list lookup found a CNAME record, but there were no A
5175 records associated with the domain it pointed at, Exim crashed.
5177 35. If a DNS black list lookup returned more than one A record, Exim ignored
5178 all but the first. It now scans all returned addresses if a particular IP
5179 value is being sought. In this situation, the contents of the
5180 $dnslist_value variable are a list of all the addresses, separated by a
5183 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5184 to accept a host name and all its aliases if the forward lookup for any of
5185 them yielded the IP address of the incoming connection. Now it accepts only
5186 those names whose forward lookup yields the correct IP address. Any other
5187 names are discarded. This closes a loophole whereby a rogue DNS
5188 administrator could create reverse DNS records to break through a
5189 wildcarded host restriction in an ACL.
5191 37. If a user filter or a system filter that ran in a subprocess used any of
5192 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5193 the wrong values were passed to the pipe command ($thisaddress had the
5194 value of $0, $0 had the value of $1, etc). This bug was introduced by
5195 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5197 38. Improved the line breaking for long SMTP error messages from ACLs.
5198 Previously, if there was no break point between 40 and 75 characters, Exim
5199 left the rest of the message alone. Two changes have been made: (a) I've
5200 reduced the minimum length to 35 characters; (b) if it can't find a break
5201 point between 35 and 75 characters, it looks ahead and uses the first one
5202 that it finds. This may give the occasional overlong line, but at least the
5203 remaining text gets split now.
5205 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5206 file descriptors might be low, and that setting 1000 would always raise it.
5207 It turns out that in some environments, the limit is already over 1000 and
5208 that lowering it causes trouble. So now Exim takes care not to decrease it.
5210 40. When delivering a message, the value of $return_path is set to $sender_
5211 address at the start of routing (routers may change the value). By an
5212 oversight, this default was not being set up when an address was tested by
5213 -bt or -bv, which affected the outcome if any router or filter referred to
5216 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5217 log only when "message" or "log_message" are set. However, if one of the
5218 conditions was an address verification, or a call to a nested ACL, the
5219 messages generated by the underlying test were being passed through. This
5220 no longer happens. The underlying message is available in $acl_verify_
5221 message for both "message" and "log_message" expansions, so it can be
5222 passed through if needed.
5224 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5225 new expansion $bh_ to give the encoded byte string without charset
5226 translation. Translation happens only if iconv() is available; HAVE_ICONV
5227 indicates this at build time. HEADERS_CHARSET gives the charset to
5228 translate to; headers_charset can change it in the configuration, and
5229 "headers charset" can change it in an individual filter file.
5231 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5232 that creates RFC 2047 encoded "words" labels them as that charset instead
5233 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5234 expansion operator; (ii) when Exim creates a From: line for a local
5235 message; (iii) when a header line is rewritten to include a "phrase" part.
5237 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5238 buggy, causing it to skip the first lines of messages whose message ID
5239 ended in 'D'. This would not have bitten before Exim release 4.14, because
5240 message IDs were unlikely to end in 'D' before then. The effect was to have
5241 incorrect size information for certain domains.
5243 45. #include "config.h" was missing at the start of the crypt16.c module. This
5244 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5247 46. If there was a timeout during a "random" callout check, Exim treated it as
5248 a failure of the random address, and carried on sending RSET and the real
5249 address. If the delay was just some slowness somewhere, the response to the
5250 original RCPT would be taken as a response to RSET and so on, causing
5251 mayhem of various kinds.
5253 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5254 when I implemented it. It didn't allow for the fact that some option values
5255 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5256 the right test for positive values.
5258 48. Domain names in DNS records are case-independent. Exim always looks them up
5259 in lower case. Some resolvers return domain names in exactly the case they
5260 appear in the zone file, that is, they may contain uppercase letters. Not
5261 all resolvers do this - some return always lower case. Exim was treating a
5262 change of case by a resolver as a change of domain, similar to a widening
5263 of a domain abbreviation. This triggered its re-routing code and so it was
5264 trying to route what was effectively the same domain again. This normally
5265 caused routing to fail (because the router wouldn't handle the domain
5266 twice). Now Exim checks for this case specially, and just changes the
5267 casing of the domain that it ultimately uses when it transmits the message
5270 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5273 50. If a filter generated a file delivery with a non-absolute name (possible if
5274 no home directory exists for the router), the forbid_file option was not
5277 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5278 the existing equality matching.
5280 52. Exim was using ints instead of ino_t variables in some places where it was
5281 dealing with inode numbers.
5283 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5284 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5285 called TMPDIR, and if it finds it is different, it changes its value.
5287 54. The smtp_printf() function is now made available to local_scan() so
5288 additional output lines can be written before returning. There is also an
5289 smtp_fflush() function to enable the detection of a dropped connection.
5290 The variables smtp_input and smtp_batched_input are exported to
5293 55. Changed the default runtime configuration: the message "Unknown user"
5294 has been removed from the ACL, and instead placed on the localuser router,
5295 using the cannot_route_message feature. This means that any verification
5296 failures that generate their own messages won't get overridden. Similarly,
5297 the "Unrouteable address" message that was in the ACL for unverifiable
5298 relay addresses has also been removed.
5300 56. Added hosts_avoid_esmtp to the smtp transport.
5302 57. The exicyclog script was not checking for the esoteric option
5303 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5304 will work only if exicyclog is run under the appropriate euid.
5306 58. Following a discussion on the list, the rules by which Exim recognises line
5307 endings on incoming messages have been changed. The -dropcr and drop_cr
5308 options are now no-ops, retained only for backwards compatibility. The
5309 following line terminators are recognized: LF CRLF CR. However, special
5310 processing applies to CR:
5312 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5313 nor a local message in the state where . is a terminator.
5315 (ii) If a bare CR is encountered in a header line, an extra space is added
5316 after the line terminator so as not to end the header. The reasoning
5317 behind this is that bare CRs in header lines are most likely either
5318 to be mistakes, or people trying to play silly games.
5320 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5321 was being incorrectly given as 18 bytes larger than it should have been.
5322 This is a VOB (very old bug).
5324 60. This may never have affected anything current, but just in case it has:
5325 When the local host is found other than at the start of a list of hosts,
5326 the local host, those with the same MX, and any that follow, are discarded.
5327 When the list in question was part of a longer list of hosts, the following
5328 hosts (not currently being processed) were also being discarded. This no
5329 longer happens. I'm not sure if this situation could ever has previously
5332 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5335 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5336 header. This is something that is recommended in a new Internet Draft, and
5337 is something that is documented as being done by Sendmail. There are two
5338 possible values. For messages generated by the autoreply transport, Exim
5341 Auto-Submitted: auto-replied
5343 whereas for all other generated messages (e.g. bounces) it adds
5345 Auto-Submitted: auto-generated
5347 63. The "personal" condition in filters now includes a test for the
5348 Auto-Submitted: header. If it contains the string "auto-" the message it
5349 not considered personal.
5351 64. Added rcpt_include_affixes as a generic transport option.
5353 65. Added queue_only_override (default true).
5355 66. Added the syslog_duplication option.
5357 67. If what should have been the first header line of a message consisted of
5358 a space followed by a colon, Exim was mis-interpreting it as a header line.
5359 It isn't of course - it is syntactically invalid and should therefore be
5360 treated as the start of the message body. The misbehaviour could have
5361 caused a number of strange effects, including loss of data in subsequent
5362 header lines, and spool format errors.
5364 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5365 client host had authenticated. This control can now be exercised by an ACL
5366 for more flexibility.
5368 69. By default, callouts do not happen when testing with -bh. There is now a
5369 variant, -bhc, which does actually run the callout code, including
5370 consulting and updating the callout cache.
5372 70. Added support for saslauthd authentication, courtesy of Alexander
5375 71. If statvfs() failed on the spool or log directories while checking their
5376 size for availability, Exim confusingly gave the error "space shortage".
5377 Furthermore, in debugging mode it crashed with a floating point exception.
5378 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5379 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5380 serious problem, Exim now writes to the main and panic logs when this
5381 happens, with details of the failure. It then refuses to accept the
5382 incoming message, giving the message "spool directory problem" or "log
5383 directory problem" with a 421 code for SMTP messages.
5385 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5386 0, 1, and 2 exist, because some OS complain for execs without them (see
5387 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5388 descriptors. However, the code omitted to check that the open succeeded,
5389 causing mysterious errors if for some reason the permissions on /dev/null
5390 got screwed. Now Exim writes a message to the main and panic logs, and
5391 bombs out if it can't open /dev/null.
5393 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5394 interact so that it is all more flexible. It is supposed to remain
5395 backwards compatible. Also added extra_local_interfaces.
5397 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5398 to bomb out with an assertion failure - to the client this appears as a
5399 connection drop. This problem occurs in the part of the code that was taken
5400 from the Samba project. Fortunately, the assertion is in a very simple
5401 function, so I have fixed this by reproducing the function inline in the
5402 one place where it is called, and arranging for authentication to fail
5403 instead of killing the process with assert().
5405 75. The SPA client code was not working when the server requested OEM rather
5406 than Unicode encoding.
5408 76. Added code to make require_files with a specific uid setting more usable in
5409 the case where statting the file as root fails - usually a non-root-mounted
5410 NFS file system. When this happens and the failure is EACCES, Exim now
5411 forks a subprocess and does the per-uid checking as the relevant uid.
5413 77. Added process_log_path.
5415 78. If log_file_path was not explicitly set, a setting of check_log_space or
5416 check_log_inodes was ignored.
5418 79. If a space check for the spool or log partitions fails, the incident is now
5419 logged. Of course, in the latter case the data may get lost...
5421 80. Added the %p formatting code to string_format() so that it can be used to
5422 print addresses in debug_print(). Adjusted all the address printing in the
5423 debugging in store.c to use %p rather than %d.
5425 81. There was a concern that a line of code in smtp_in.c could overflow a
5426 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5427 initially expressed, the concern was not well-founded, because trailing
5428 spaces are removed early. However, if the trailing spaces were followed by
5429 a NULL, they did not get removed, so the overflow was possible. Two fixes
5432 (a) I re-wrote the offending code in a cleaner fashion.
5433 (b) If an incoming SMTP command contains a NULL character, it is rejected
5436 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5437 runs initgroups(), so that if the Exim user is in any additional groups,
5438 they will be used during message reception.
5444 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob