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.
91 PP/02 Raise smtp_cmd_buffer_size to 16kB. Patch from Paul Fisher.
94 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
95 Heavily based on revision 40f9a89a from Simon Arlott's tree.
102 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
104 PP/02 Harmonised TLS library version reporting; only show if debugging.
105 Layout now matches that introduced for other libraries in 4.74 PP/03.
107 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
109 PP/04 New "dns_use_edns0" global option.
111 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
114 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
115 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
117 TK/01 Updated PolarSSL code to 0.14.2.
118 Bugzilla 1097. Patch from Andreas Metzler.
120 PP/07 Catch divide-by-zero in ${eval:...}.
123 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
126 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
127 format-string attack -- SECURITY: remote arbitrary code execution.
129 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
130 time unintentionally subject to list matching rules, letting the header
131 cause arbitrary Exim lookups (of items which can occur in lists, *not*
132 arbitrary string expansion). This allowed for information disclosure.
134 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
135 INT_MIN/-1 -- value coerced to INT_MAX.
141 NM/01 Workround for PCRE version dependency in version reporting
144 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
145 This fixes portability to compilers other than gcc, notably
146 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
148 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
149 makefiles for portability to HP-UX and POSIX correctness.
151 PP/01 Permit LOOKUP_foo enabling on the make command-line.
152 Also via indented variable definition in the Makefile.
153 (Debugging by Oliver Heesakkers).
155 PP/02 Restore caching of spamd results with expanded spamd_address.
156 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
158 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
159 Improves build reliability. Fix from: Frank Elsner
161 NM/02 Fix wide character breakage in the rfc2047 coding
162 Fixes bug 1064. Patch from Andrey N. Oktyabrski
164 NM/03 Allow underscore in dnslist lookups
165 Fixes bug 1026. Patch from Graeme Fowler
167 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
168 Code patches from Adam Ciarcinski of NetBSD.
170 NM/04 Fixed exiqgrep to cope with mailq missing size issue
173 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
174 is logged, to avoid truncation. Patch from John Horne.
176 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
177 Patch from Jakob Hirsch.
179 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
180 SQL string expansion failure details.
181 Patch from Andrey Oktyabrski.
183 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
184 Patch from Simon Arlott.
186 PP/09 New lookups functionality failed to compile on old gcc which rejects
187 extern declarations in function scope.
188 Patch from Oliver Fleischmann
190 PP/10 Use sig_atomic_t for flags set from signal handlers.
191 Check getgroups() return and improve debugging.
192 Fixed developed for diagnosis in bug 927 (which turned out to be
195 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
196 Patch from Mark Zealey.
198 PP/12 Bugzilla 1056: Improved spamd server selection.
199 Patch from Mark Zealey.
201 PP/13 Bugzilla 1086: Deal with maildir quota file races.
202 Based on patch from Heiko Schlittermann.
204 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
205 Patch from Uwe Doering, sign-off by Michael Haardt.
207 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
208 variable declaration deep within a block. Bug and patch from
211 PP/15 lookups-Makefile IRIX compatibilty coercion.
213 PP/16 Make DISABLE_DKIM build knob functional.
215 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
216 Patch by Simon Arlott
218 TF/03 Fix valgrind.h portability to C89 compilers that do not support
219 variable argument macros. Our copy now differs from upstream.
225 TF/01 Failure to get a lock on a hints database can have serious
226 consequences so log it to the panic log.
228 TF/02 Log LMTP confirmation messages in the same way as SMTP,
229 controlled using the smtp_confirmation log selector.
231 TF/03 Include the error message when we fail to unlink a spool file.
233 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
234 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
235 for maintaining out-of-tree patches for some time.
237 PP/01 Bugzilla 139: Documentation and portability issues.
238 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
239 Handle per-OS dynamic-module compilation flags.
241 PP/02 Let /dev/null have normal permissions.
242 The 4.73 fixes were a little too stringent and complained about the
243 permissions on /dev/null. Exempt it from some checks.
244 Reported by Andreas M. Kirchwitz.
246 PP/03 Report version information for many libraries, including
247 Exim version information for dynamically loaded libraries. Created
248 version.h, now support a version extension string for distributors
249 who patch heavily. Dynamic module ABI change.
251 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
252 privilege escalation vulnerability whereby the Exim run-time user
253 can cause root to append content of the attacker's choosing to
256 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
259 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
260 If dropping privileges for untrusted macros, we disabled normal logging
261 on the basis that it would fail; for the Exim run-time user, this is not
262 the case, and it resulted in successful deliveries going unlogged.
263 Fixed. Reported by Andreas Metzler.
269 PP/01 Date: & Message-Id: revert to normally being appended to a message,
270 only prepend for the Resent-* case. Fixes regression introduced in
271 Exim 4.70 by NM/22 for Bugzilla 607.
273 PP/02 Include check_rfc2047_length in configure.default because we're seeing
274 increasing numbers of administrators be bitten by this.
276 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
278 PP/03 Bugzilla 994: added openssl_options main configuration option.
280 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
282 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
284 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
286 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
287 without a peer certificate, leading to a segfault because of an
288 assumption that peers always have certificates. Be a little more
289 paranoid. Problem reported by Martin Tscholak.
291 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
292 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
293 NB: ClamAV planning to remove STREAM in "middle of 2010".
294 CL also introduces -bmalware, various -d+acl logging additions and
295 more caution in buffer sizes.
297 PP/09 Implemented reverse_ip expansion operator.
299 PP/10 Bugzilla 937: provide a "debug" ACL control.
301 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
303 PP/12 Bugzilla 973: Implement --version.
305 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
307 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
309 PP/15 Bugzilla 816: support multiple condition rules on Routers.
311 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
312 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
313 ignore trailing whitespace.
315 JJ/02 prevent non-panic DKIM error from being sent to paniclog
317 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
320 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
321 Notification from Dr Andrew Aitchison.
323 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
324 ExtendedDetectionInfo response format.
325 Notification from John Horne.
327 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
330 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
331 XSL and documented dependency on system catalogs, with examples of how
334 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
337 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
338 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
339 configuration file which is writeable by the Exim user or group.
341 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
342 of configuration files to cover files specified with the -C option if
343 they are going to be used with root privileges, not just the default
346 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
347 option (effectively making it always true).
349 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
350 files to be used while preserving root privileges.
352 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
353 that rogue child processes cannot use them.
355 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
356 run-time user, instead of root.
358 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
359 Exim run-time user without dropping privileges.
361 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
362 result string, instead of calling string_vformat() twice with the same
365 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
366 for other users. Others should always drop root privileges if they use
367 -C on the command line, even for a whitelisted configure file.
369 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
371 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
377 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
378 $data_path, and $header_path variables; fixed documentation bugs and
381 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
382 exipick to access non-standard spools, including the "frozen" queue
385 NM/01 Bugzilla 965: Support mysql stored procedures.
386 Patch from Alain Williams
388 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
390 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
391 Patch from Andreas Metzler
393 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
394 Patch from Kirill Miazine
396 NM/05 Bugzilla 671: Added umask to procmail example.
398 JJ/03 installed exipick 20100323.0, fixing doc bug
400 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
401 directory. Notification and patch from Dan Rosenberg.
403 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
405 TK/02 Improve log output when DKIM signing operation fails.
407 MH/01 Treat the transport option dkim_domain as a colon separated
408 list, not as a single string, and sign the message with each element,
409 omitting multiple occurences of the same signer.
411 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
412 Bugzilla 985, 986. Patch by Simon Arlott
414 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
415 Patch by Simon Arlott
417 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
418 MBX locking. Notification from Dan Rosenberg.
424 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
426 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
428 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
430 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
432 NM/04 Bugzilla 915: Flag broken perl installation during build.
438 TK/01 Added patch by Johannes Berg that expands the main option
439 "spamd_address" if it starts with a dollar sign.
441 TK/02 Write list of recipients to X-Envelope-Sender header when building
442 the mbox-format spool file for content scanning (suggested by Jakob
445 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
446 (http://www.dcc-servers.net/) support via dccifd. Activated by
447 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
449 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
450 by Mark Daniel Reidel <mr@df.eu>.
452 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
453 When building exim an external PCRE library is now needed -
454 PCRE is a system library on the majority of modern systems.
455 See entry on PCRE_LIBS in EDITME file.
457 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
458 conversation. Added nologin parameter to request.
459 Patch contributed by Kirill Miazine.
461 TF/01 Do not log submission mode rewrites if they do not change the address.
463 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
465 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
466 log files in place. Contributed by Roberto Lima.
468 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
470 TF/03 Bugzilla 615: When checking the local_parts router precondition
471 after a local_part_suffix or local_part_prefix option, Exim now
472 does not use the address's named list lookup cache, since this
473 contains cached lookups for the whole local part.
475 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
476 Robert Millan. Documentation is in experimental-spec.txt.
478 TF/04 Bugzilla 668: Fix parallel build (make -j).
480 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
482 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
483 Patch provided by Jan Srzednicki.
485 TF/05 Leading white space used to be stripped from $spam_report which
486 wrecked the formatting. Now it is preserved.
488 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
489 that they are available at delivery time.
491 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
493 TF/08 TLS error reporting now respects the incoming_interface and
494 incoming_port log selectors.
496 TF/09 Produce a more useful error message if an SMTP transport's hosts
497 setting expands to an empty string.
499 NM/06 Bugzilla 744: EXPN did not work under TLS.
500 Patch provided by Phil Pennock.
502 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
503 Patch provided by Richard Godbee.
505 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
506 acl_smtp_notquit, added index entry.
508 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
509 Patch provided by Eugene Bujak.
511 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
512 accept(). Patch provided by Maxim Dounin.
514 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
515 Patch provided by Phil Pennock.
517 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
519 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
520 Patch provided by Brad "anomie" Jorsch.
522 NM/14 Bugzilla 622: Added timeout setting to transport filter.
523 Patch provided by Dean Brooks.
525 TK/05 Add native DKIM support (does not depend on external libraries).
527 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
528 Patch provided by Graeme Fowler.
530 NM/16 Bugzilla 851: Documentation example syntax fix.
532 NM/17 Changed NOTICE file to remove references to embedded PCRE.
534 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
537 NM/19 Bugzilla 745: TLS version reporting.
538 Patch provided by Phil Pennock.
540 NM/20 Bugzilla 167: bool: condition support.
541 Patch provided by Phil Pennock.
543 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
544 clients. Patch provided by Phil Pennock.
546 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
547 Patch provided by Brad "anomie" Jorsch.
549 NM/23 Bugzilla 687: Fix misparses in eximstats.
550 Patch provided by Heiko Schlittermann.
552 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
553 Patch provided by Heiko Schlittermann.
555 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
556 plus update to original patch.
558 NM/26 Bugzilla 799: Documentation correction for ratelimit.
560 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
561 Patch provided by David Brownlee.
563 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
565 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
567 NM/30 Bugzilla 888: TLS documentation bugfixes.
569 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
571 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
572 Unlike the original bugzilla I have changed all shell scripts in src tree.
574 NM/33 Bugzilla 898: Transport filter timeout fix.
575 Patch by Todd Rinaldo.
577 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
578 Patch by Serge Demonchaux.
580 NM/35 Bugzilla 39: Base64 decode bug fixes.
581 Patch by Jakob Hirsch.
583 NM/36 Bugzilla 909: Correct connect() call in dcc code.
585 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
587 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
589 NM/39 Bugzilla 911: Fixed MakeLinks build script.
595 TK/01 Add preliminary DKIM support. Currently requires a forked version of
596 ALT-N's libdkim that I have put here:
597 http://duncanthrax.net/exim-experimental/
599 Note to Michael Haardt: I had to rename some vars in sieve.c. They
600 were called 'true' and it seems that C99 defines that as a reserved
601 keyword to be used with 'bool' variable types. That means you could
602 not include C99-style headers which use bools without triggering
603 build errors in sieve.c.
605 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
606 as mailq or other aliases. Changed the --help handling significantly
607 to do whats expected. exim_usage() emits usage/help information.
609 SC/01 Added the -bylocaldomain option to eximstats.
611 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
613 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
615 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
621 PH/01 Another patch from the Sieve maintainer.
623 PH/02 When an IPv6 address is converted to a string for single-key lookup
624 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
625 dots are used instead of colons so that keys in lsearch files need not
626 contain colons. This was done some time before quoting was made available
627 in lsearch files. However, iplsearch files do require colons in IPv6 keys
628 (notated using the quote facility) so as to distinguish them from IPv4
629 keys. This meant that lookups for IP addresses in host lists did not work
630 for iplsearch lookups.
632 This has been fixed by arranging for IPv6 addresses to be expressed with
633 colons if the lookup type is iplsearch. This is not incompatible, because
634 previously such lookups could never work.
636 The situation is now rather anomolous, since one *can* have colons in
637 ordinary lsearch keys. However, making the change in all cases is
638 incompatible and would probably break a number of configurations.
640 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
643 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
644 conversion specification without a maximum field width, thereby enabling
645 a rogue spamd server to cause a buffer overflow. While nobody in their
646 right mind would setup Exim to query an untrusted spamd server, an
647 attacker that gains access to a server running spamd could potentially
648 exploit this vulnerability to run arbitrary code as the Exim user.
650 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
651 $primary_hostname instead of what libspf2 thinks the hosts name is.
653 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
654 a directory entry by the name of the lookup key. Previously, if a
655 symlink pointed to a non-existing file or a file in a directory that
656 Exim lacked permissions to read, a lookup for a key matching that
657 symlink would fail. Now it is enough that a matching directory entry
658 exists, symlink or not. (Bugzilla 503.)
660 PH/03 The body_linecount and body_zerocount variables are now exported in the
663 PH/04 Added the $dnslist_matched variable.
665 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
666 This means they are set thereafter only if the connection becomes
669 PH/06 Added the client_condition to authenticators so that some can be skipped
670 by clients under certain conditions.
672 PH/07 The error message for a badly-placed control=no_multiline_responses left
673 "_responses" off the end of the name.
675 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
677 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
678 (without spaces) instead of just copying the configuration text.
680 PH/10 Added the /noupdate option to the ratelimit ACL condition.
682 PH/11 Added $max_received_linelength.
684 PH/12 Added +ignore_defer and +include_defer to host lists.
686 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
687 way in which PCRE > 7.0 is built.
689 PH/14 Implemented queue_only_load_latch.
691 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
692 MAIL command. The effect was to mangle the value on 64-bit systems.
694 PH/16 Another patch from the Sieve maintainer.
696 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
698 PH/18 If a system quota error occurred while trying to create the file for
699 a maildir delivery, the message "Mailbox is full" was not appended to the
700 bounce if the delivery eventually timed out. Change 4.67/27 below applied
701 only to a quota excession during the actual writing of the file.
703 PH/19 It seems that peer DN values may contain newlines (and other non-printing
704 characters?) which causes problems in log lines. The DN values are now
705 passed through string_printing() before being added to log lines.
707 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
708 and InterBase are left for another time.)
710 PH/21 Added message_body_newlines option.
712 PH/22 Guard against possible overflow in moan_check_errorcopy().
714 PH/23 POSIX allows open() to be a macro; guard against that.
716 PH/24 If the recipient of an error message contained an @ in the local part
717 (suitably quoted, of course), incorrect values were put in $domain and
718 $local_part during the evaluation of errors_copy.
724 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
725 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
728 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
729 issue a MAIL command.
731 PH/02 In an ACL statement such as
733 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
735 if a client was not listed at all, or was listed with a value other than
736 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
737 the condition was not true (as it should be), so access was not denied.
738 The bug was that the ! inversion was incorrectly passed on to the second
739 item. This has been fixed.
741 PH/03 Added additional dnslists conditions == and =& which are different from
742 = and & when the dns lookup returns more than one IP address.
744 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
745 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
747 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
748 FSYNC, which compiles an option called disable_fsync that allows for
749 bypassing fsync(). The documentation is heavily laced with warnings.
751 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
753 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
754 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
755 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
756 including adding "make clean"; (3) Added -fPIC when compiling the test
757 dynamically loaded module, to get rid of a warning.
759 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
760 message fails, move_frozen_messages = true and ignore_bounce_errors_after
761 = 0s. The bug is otherwise harmless.
763 PH/07 There was a bug in the dovecot authenticator such that the value of
764 $auth1 could be overwritten, and so not correctly preserved, after a
765 successful authentication. This usually meant that the value preserved by
766 the server_setid option was incorrect.
768 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
770 PH/09 Installed PCRE release 7.0.
772 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
773 run for batched SMTP input. It is now run at the start of every message
774 in the batch. While fixing this I discovered that the process information
775 (output by running exiwhat) was not always getting set for -bs and -bS
776 input. This is fixed, and it now also says "batched" for BSMTP.
778 PH/11 Added control=no_pipelining.
780 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
781 patch, slightly modified), and move the expansion of helo_data till after
782 the connection is made in the smtp transport (so it can use these
785 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
787 PH/14 Added log_selector = +pid.
789 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
791 PH/16 Add ${if forany and ${if forall.
793 PH/17 Added dsn_from option to vary the From: line in DSNs.
795 PH/18 Flush SMTP output before performing a callout, unless control =
796 no_callout_flush is set.
798 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
799 was true (the default) a successful delivery failed to delete the retry
800 item, thus causing premature timeout of the address. The bug is now
803 PH/20 Added hosts_avoid_pipelining to the smtp transport.
805 PH/21 Long custom messages for fakedefer and fakereject are now split up
806 into multiline reponses in the same way that messages for "deny" and
807 other ACL rejections are.
809 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
810 with slight modification.
812 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
813 draft, changing the syntax and factoring some duplicate code".
815 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
816 for deliveries of the second and subsequent messages over the same SMTP
819 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
820 ${reduce, with only minor "tidies".
822 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
824 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
825 expansion side effects.
827 PH/27 When a message times out after an over-quota error from an Exim-imposed
828 quota, the bounce message says "mailbox is full". This message was not
829 being given when it was a system quota that was exceeded. It now should
832 MH/03 Made $recipients available in local_scan(). local_scan() already has
833 better access to the recipient list through recipients_list[], but
834 $recipients can be useful in postmaster-provided expansion strings.
836 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
837 in the case of a MAIL command with additional options following the
838 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
839 were accidentally chopped off.
841 PH/29 SMTP synchronization checks are implemented when a command is read -
842 there is a check that no more input is waiting when there shouldn't be
843 any. However, for some commands, a delay in an ACL can mean that it is
844 some time before the response is written. In this time, more input might
845 arrive, invalidly. So now there are extra checks after an ACL has run for
846 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
847 pipelining has not been advertised.
849 PH/30 MH's patch to allow iscntrl() characters to be list separators.
851 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
852 returned in the SMTP response when smtp_return_error_details was false.
855 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
856 instead of the C I/O that was originally supplied, because problems were
859 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
860 Exim which did not show up earlier: it was assuming that a call to
861 SSL_CTX_set_info_callback() might give an error value. In fact, there is
862 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
863 was a macro that became an assignment, so it seemed to work. This has
864 changed to a proper function call with a void return, hence the compile
865 error. Exim's code has been fixed.
867 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
870 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
872 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
874 PH/37 If a message is not accepted after it has had an id assigned (e.g.
875 because it turns out to be too big or there is a timeout) there is no
876 "Completed" line in the log. When some messages of this type were
877 selected by exigrep, they were listed as "not completed". Others were
878 picked up by some special patterns. I have improved the selection
879 criteria to be more general.
881 PH/38 The host_find_failed option in the manualroute router can now be set
882 to "ignore", to completely ignore a host whose IP address cannot be
883 found. If all hosts are ignored, the behaviour is controlled by the new
884 host_all_ignored option.
886 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
887 homing or because of multiple MX records with /mx) generated more than
888 one IP address, and the following item turned out to be the local host,
889 all the secondary addresses of the first item were incorrectly removed
890 from the list, along with the local host and any following hosts (which
891 is what is supposed to happen).
893 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
894 whoever called Exim into the -H file. In the case of the daemon it was
895 behaving confusingly. When first started, it used values for whoever
896 started the daemon, but after a SIGHUP it used the Exim user (because it
897 calls itself on a restart). I have changed the code so that it now always
900 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
901 message are rejected with the same error (e.g. no authentication or bad
902 sender address), and a DATA command is nevertheless sent (as can happen
903 with PIPELINING or a stupid MUA), the error message that was given to the
904 RCPT commands is included in the rejection of the DATA command. This is
905 intended to be helpful for MUAs that show only the final error to their
908 PH/42 Another patch from the Sieve maintainer.
910 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
911 Eximstats - Fixed some broken HTML links and added missing column headers
913 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
914 columns for Rejects, Temp Rejects, Ham, and Spam rows.
916 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
918 PH/43 Yet another patch from the Sieve maintainer.
920 PH/44 I found a way to check for a TCP/IP connection going away before sending
921 the response to the final '.' that terminates a message, but only in the
922 case where the client has not sent further data following the '.'
923 (unfortunately, this is allowed). However, in many cases there won't be
924 any further data because there won't be any more messages to send. A call
925 to select() can be used: if it shows that the input is "ready", there is
926 either input waiting, or the socket has been closed. An attempt to read
927 the next input character can distinguish the two cases. Previously, Exim
928 would have sent an OK response which the client would never have see.
929 This could lead to message repetition. This fix should cure that, at
930 least in a lot of common cases.
932 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
933 advertised in response to EHLO.
939 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
940 fixed by 4.65/MH/01 (is this a record?) are fixed:
942 (i) An empty string was always treated as zero by the numeric comparison
943 operators. This behaviour has been restored.
945 (ii) It is documented that the numeric comparison operators always treat
946 their arguments as decimal numbers. This was broken in that numbers
947 starting with 0 were being interpreted as octal.
949 While fixing these problems I realized that there was another issue that
950 hadn't been noticed. Values of message_size_limit (both the global option
951 and the transport option) were treated as octal if they started with 0.
952 The documentation was vague. These values are now always treated as
953 decimal, and I will make that clear in the documentation.
959 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
960 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
963 MH/01 Don't check that the operands of numeric comparison operators are
964 integers when their expansion is in "skipping" mode (fixes bug
965 introduced by 4.64-PH/07).
967 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
968 child addresses, Exim now panics and dies. Previously, because the count
969 is held in a short int, deliveries were likely to be lost. As such a
970 large number of recipients for a single message is ridiculous
971 (performance will be very, very poor), I have chosen to impose a limit
972 rather than extend the field.
978 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
979 leftover -K file (the existence of which was triggered by #402).
980 While we were at it, introduced process PID as part of the -K
981 filename. This should rule out race conditions when creating
984 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
985 processing considerably. Previous code took too long for large mails,
986 triggering a timeout which in turn triggers #401.
988 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
989 in the DK code in transports.c. sendfile() is not really portable,
990 hence the _LINUX specificness.
992 TF/01 In the add_headers option to the mail command in an Exim filter,
993 there was a bug that Exim would claim a syntax error in any
994 header after the first one which had an odd number of characters
997 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
998 callout verification, Exim cached a "reject" for the entire domain. This
999 is correct for most verifications, but it is not correct for a recipient
1000 verification with use_sender or use_postmaster set, because in that case
1001 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1002 case of MAIL FROM:<> rejection from other early rejections (e.g.
1003 rejection of HELO). When verifying a recipient using a non-null MAIL
1004 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1005 Whatever the result of the callout, the value of the domain cache is
1006 left unchanged (for any other kind of callout, getting as far as trying
1007 RCPT means that the domain itself is ok).
1009 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1012 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1013 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1014 session, the daemon ignores ECONNECT errors and logs others; it now
1015 ignores EPIPE as well.
1017 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1018 (quoted-printable decoding).
1020 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1021 later the small subsequent patch to fix an introduced bug.
1023 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1025 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1027 PH/08 An error is now given if message_size_limit is specified negative.
1029 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1030 to be given (somewhat) arbitrary names.
1032 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1035 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1036 miscellaneous code fixes
1038 PH/10 Added the log_reject_target ACL modifier to specify where to log
1041 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1042 hostname. This is wrong, because it relates to the incoming message (and
1043 probably the interface on which it is arriving) and not to the outgoing
1044 callout (which could be using a different interface). This has been
1045 changed to use the value of the helo_data option from the smtp transport
1046 instead - this is what is used when a message is actually being sent. If
1047 there is no remote transport (possible with a router that sets up host
1048 addresses), $smtp_active_hostname is used.
1050 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1051 tweaks were necessary in order to get it to work (see also 21 below):
1052 (a) The code assumed that strncpy() returns a negative number on buffer
1053 overflow, which isn't the case. Replaced with Exim's string_format()
1055 (b) There were several signed/unsigned issues. I just did the minimum
1056 hacking in of casts. There is scope for a larger refactoring.
1057 (c) The code used strcasecmp() which is not a standard C function.
1058 Replaced with Exim's strcmpic() function.
1059 (d) The code set only $1; it now sets $auth1 as well.
1060 (e) A simple test gave the error "authentication client didn't specify
1061 service in request". It would seem that Dovecot has changed its
1062 interface. Fortunately there's a specification; I followed it and
1063 changed what the client sends and it appears to be working now.
1065 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1068 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1069 address A is aliased to B and C, where B exists and C does not. Without
1070 -v the output is "A verified" because verification stops after a
1071 successful redirection if more than one address is generated. However,
1072 with -v the child addresses are also verified. Exim was outputting "A
1073 failed to verify" and then showing the successful verification for C,
1074 with its parentage. It now outputs "B failed to verify", showing B's
1075 parentage before showing the successful verification of C.
1077 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1078 look up a TXT record in a specific list after matching in a combined
1081 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1082 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1083 they consult the DNS. I had assumed they would set it the way they
1084 wanted; and indeed my experiments on Linux seem to show that in some
1085 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1086 To be on the safe side, however, I have now made the interface to
1087 host_find_byname() similar to host_find_bydns(), with an argument
1088 containing the DNS resolver options. The host_find_byname() function now
1089 sets these options at its start, just as host_find_bydns() does. The smtp
1090 transport options dns_qualify_single and dns_search_parents are passed to
1091 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1092 of host_find_byname() use the default settings of RES_DEFNAMES
1093 (qualify_single) but not RES_DNSRCH (search_parents).
1095 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1096 spool_read_header() do less string testing, by means of a preliminary
1097 switch on the second character of optional "-foo" lines. (This is
1098 overdue, caused by the large number of possibilities that now exist.
1099 Originally there were few.) While I was there, I also converted the
1100 str(n)cmp tests so they don't re-test the leading "-" and the first
1101 character, in the hope this might squeeze out yet more improvement.
1103 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1104 flag allowing group syntax was set by the header_syntax check but not
1105 turned off, possible causing trouble later; (2) The flag was not being
1106 set at all for the header_verify test, causing "group"-style headers to
1107 be rejected. I have now set it in this case, and also caused header_
1108 verify to ignore an empty address taken from a group. While doing this, I
1109 came across some other cases where the code for allowing group syntax
1110 while scanning a header line wasn't quite right (mostly, not resetting
1111 the flag correctly in the right place). These bugs could have caused
1112 trouble for malformed header lines. I hope it is now all correct.
1114 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1115 with the "reply" argument non-NULL. The code, however (which originally
1116 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1117 but it didn't always do it. This confused somebody who was copying the
1118 code for some other use. I have removed all the tests.
1120 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1121 feature that was used to support insecure browsers during the U.S. crypto
1122 embargo. It requires special client support, and Exim is probably the
1123 only MTA that supported it -- and would never use it because real RSA is
1124 always available. This code has been removed, because it had the bad
1125 effect of slowing Exim down by computing (never used) parameters for the
1126 RSA_EXPORT functionality.
1128 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1129 authenticator to fail if there's a tab character in the incoming data
1130 (there should never be unless someone is messing about, as it's supposed
1131 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1132 if the connection is using TLS or if the remote IP is the same as the
1133 local IP, and the "valid-client-cert option" if a client certificate has
1136 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1137 authenticators. This can be used for authorization after authentication
1138 succeeds. (In the case of plaintext, it servers for both authentication
1141 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1142 if any retry times were supplied.
1144 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1145 connection, where there is no client IP address to check. In this
1146 situation, the verify now always succeeds.
1148 PH/25 Applied John Jetmore's -Mset patch.
1150 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1152 PH/27 In a string expansion for a processed (not raw) header when multiple
1153 headers of the same name were present, leading whitespace was being
1154 removed from all of them, but trailing whitespace was being removed only
1155 from the last one. Now trailing whitespace is removed from each header
1156 before concatenation. Completely empty headers in a concatenation (as
1157 before) are ignored.
1159 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1160 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1162 PH/29 [Removed. This was a change that I later backed out, and forgot to
1163 correct the ChangeLog entry (that I had efficiently created) before
1164 committing the later change.]
1166 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1167 address errors (4xx response to RCPT) over the same connection as other
1168 messages routed to the same hosts. Such deliveries are always "forced",
1169 so retry times are not inspected. This resulted in far too many retries
1170 for the affected addresses. The effect occurred only when there were more
1171 hosts than the hosts_max_try setting in the smtp transport when it had
1172 the 4xx errors. Those hosts that it had tried were not added to the list
1173 of hosts for which the message was waiting, so if all were tried, there
1174 was no problem. Two fixes have been applied:
1176 (i) If there are any address or message errors in an SMTP delivery, none
1177 of the hosts (tried or untried) are now added to the list of hosts
1178 for which the message is waiting, so the message should not be a
1179 candidate for sending over the same connection that was used for a
1180 successful delivery of some other message. This seems entirely
1181 reasonable: after all the message is NOT "waiting for some host".
1182 This is so "obvious" that I'm not sure why it wasn't done
1183 previously. Hope I haven't missed anything, but it can't do any
1184 harm, as the worst effect is to miss an optimization.
1186 (ii) If, despite (i), such a delivery is accidentally attempted, the
1187 routing retry time is respected, so at least it doesn't keep
1188 hammering the server.
1190 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1191 in ${readsocket because some servers need this prod.
1193 PH/32 Added some extra debug output when updating a wait-xxx database.
1195 PH/33 The hint "could be header name not terminated by colon", which has been
1196 given for certain expansion errors for a long time, was not being given
1197 for the ${if def:h_colon_omitted{... case.
1199 PH/34 The spec says: "With one important exception, whenever a domain list is
1200 being scanned, $domain contains the subject domain." There was at least
1201 one case where this was not true.
1203 PH/35 The error "getsockname() failed: connection reset by peer" was being
1204 written to the panic log as well as the main log, but it isn't really
1205 panic-worthy as it just means the connection died rather early on. I have
1206 removed the panic log writing for the ECONNRESET error when getsockname()
1209 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1210 runs only) independently of the message's sender address. This meant
1211 that, if the 4xx error was in fact related to the sender, a different
1212 message to the same recipient with a different sender could confuse
1213 things. In particualar, this can happen when sending to a greylisting
1214 server, but other circumstances could also provoke similar problems.
1215 I have changed the default so that the retry time for these errors is now
1216 based a combination of the sender and recipient addresses. This change
1217 can be overridden by setting address_retry_include_sender=false in the
1220 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1221 remote server are returned as part of bounce messages. This was not
1222 happening for LMTP over a pipe (the lmtp transport), but now it is the
1223 same for both kinds of LMTP.
1225 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1226 in header lines that were in fact CNAMEs. This is no longer the case.
1228 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1229 and queue runs started by the daemon processed all messages. This has
1230 been fixed so that -R and -S can now usefully be given with -q<time>.
1232 PH/40 Import PCRE release 6.7 (fixes some bugs).
1234 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1236 PH/42 Give an error if -q is specified more than once.
1238 PH/43 Renamed the variables $interface_address and $interface_port as
1239 $received_ip_address and $received_port, to make it clear that these
1240 values apply to message reception, and not to the outgoing interface when
1241 a message is delivered. (The old names remain recognized, of course.)
1243 PH/44 There was no timeout on the connect() call when using a Unix domain
1244 socket in the ${readsocket expansion. There now is.
1246 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1247 be meaningful with "accept".
1249 SC/01 Eximstats V1.43
1250 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1252 SC/02 Eximstats V1.44
1253 Use a glob alias rather than an array ref in the generated
1254 parser. This improves both readability and performance.
1256 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1257 Collect SpamAssassin and rejection statistics.
1258 Don't display local sender or destination tables unless
1259 there is data to show.
1260 Added average volumes into the top table text output.
1262 SC/04 Eximstats V1.46
1263 Collect data on the number of addresses (recipients)
1264 as well as the number of messages.
1266 SC/05 Eximstats V1.47
1267 Added 'Message too big' to the list of mail rejection
1268 reasons (thanks to Marco Gaiarin).
1270 SC/06 Eximstats V1.48
1271 Mainlog lines which have GMT offsets and are too short to
1272 have a flag are now skipped.
1274 SC/07 Eximstats V1.49 (Alain Williams)
1275 Added the -emptyok flag.
1277 SC/08 Eximstats V1.50
1278 Fixes for obtaining the IP address from reject messages.
1280 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1281 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1282 whitesspace changes from 4.64-PH/27
1284 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1287 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1288 are found, allow negative numbers in numeric criteria)
1290 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1292 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1293 to match changes made in 4.64-PH/43
1295 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1297 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1298 "domain" is actually an IP address, and force a failure. This locks out
1299 those revolvers/nameservers that support "A-for-A" lookups, in
1300 contravention of the specifications.
1302 PH/48 When a host name was looked up from an IP address, and the subsequent
1303 forward lookup of the name timed out, the host name was left in
1304 $sender_host_name, contrary to the specification.
1306 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1307 restricted to single-key lookups, Exim was not diagnosing an error if
1308 * or *@ was used with a query-style lookup.
1310 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1312 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1313 long ago, but noone interested enough thought of it. Let's just say that
1314 the "1.1" means that there are some new functions that weren't there at
1315 some point in the past.
1317 PH/51 Error processing for expansion failure of helo_data from an smtp
1318 transport during callout processing was broken.
1320 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1321 tested/used via the -bh/-bhc/-bs options.
1323 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1324 bug, fixed in subsequent PCRE releases).
1326 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1327 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1329 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1335 SC/01 Use a glob alias rather than an array ref in eximstats generated
1336 parser. This improves both readability and performance.
1338 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1339 Don't display local sender or destination tables in eximstats unless
1340 there is data to show.
1341 Added average volumes into the eximstats top table text output.
1343 SC/03 Collect data on the number of addresses (recipients) as well
1344 as the number of messages in eximstats.
1346 TF/01 Correct an error in the documentation for the redirect router. Exim
1347 does (usually) call initgroups() when daemonizing.
1349 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1350 with consistent privilege compared to when running as a daemon.
1352 TF/03 Note in the spec that $authenticated_id is not set for local
1353 submissions from trusted users.
1355 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1356 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1358 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1359 by adding some example configuration directives to the default
1360 configuration file. A little bit of work is required to uncomment the
1361 directives and define how usernames and passwords are checked, but
1362 there is now a framework to start from.
1364 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1365 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1366 without this. I don't know how relevant this is to other LDAP libraries.
1368 PH/02 Add the verb name to the "unknown ACL verb" error.
1370 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1372 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1374 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1375 directory not expanded when it should be if an expanded home directory
1376 was set for the address (which is overridden by the transport).
1378 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1381 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1382 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1383 because it is too late at that time, and has no effect.
1385 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1386 security issue with \' (bugzilla #107). I could not use the
1387 PQescapeStringConn() function, because it needs a PGconn value as one of
1390 PH/08 When testing addresses using -bt, indicate those final addresses that
1391 are duplicates that would not cause an additional delivery. At least one
1392 person was confused, thinking that -bt output corresponded to deliveries.
1393 (Suppressing duplicates isn't a good idea as you lose the information
1394 about possibly different redirections that led to the duplicates.)
1396 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1397 systems where poll() doesn't work, in particular OS X.
1399 PH/10 Added more information to debugging output for retry time not reached.
1401 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1402 operations in malware.c.
1404 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1407 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1408 syslog_duplication set false, log lines that would normally be written
1409 both the the main log and to the reject log were not written to syslog at
1412 PH/14 In the default configuration, change the use of "message" in ACL warn
1413 statements to "add_header".
1415 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1416 not followed by a command (e.g. "seen endif").
1418 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1419 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1422 PH/17 Added extra conditions to the default value of delay_warning_condition
1426 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1427 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1428 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1431 The Auto-Submitted: and various List- headers are standardised, whereas I
1432 don't think Precedence: ever was.
1434 PH/18 Refactored debugging code in route_finduser() to show more information,
1435 in particular, the error code if getpwnam() issues one.
1437 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1438 This is apparently needed in addition to the PH/07 change above to avoid
1439 any possible encoding problems.
1441 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1442 but not after initializing Perl.
1444 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1445 output them only if debugging. By default they are written stderr,
1446 apparently, which is not desirable.
1448 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1451 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1454 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1456 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1457 authenticated or an ident call has been made. Suppress the default
1458 values for $authenticated_id and $authenticated_sender (but permit -oMai
1459 and -oMas) when testing with -bh.
1461 PH/24 Re-jigged the order of the tests in the default configuration so that the
1462 tests for valid domains and recipients precede the DNS black list and CSA
1463 tests, on the grounds that those ones are more expensive.
1465 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1466 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1467 This bug exists in every version of Exim that I still have, right back to
1470 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1471 However, an attempt to turn on case-sensitivity in a regex key by
1472 including (?-i) didn't work because the subject string was already
1473 lowercased, and the effects were non-intuitive. It turns out that a
1474 one-line patch can be used to allow (?-i) to work as expected.
1480 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1481 other effects) broke the use of negated acl sub-conditions.
1483 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1486 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1487 "Deny" causes Exim to reject the incoming connection with a 554 error.
1488 Unfortunately, if there is a major crisis, such as a disk failure,
1489 tcp-wrappers gives "deny", whereas what one would like would be some
1490 kind of temporary error. A kludge has been added to help with this.
1491 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1492 554 error is used if errno is still zero or contains ENOENT (which occurs
1493 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1496 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1498 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1499 errors. Otherwise a message that provokes a temporary error (when other
1500 messages do not) can cause a whole host to time out.
1502 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1503 addresses were routed directly to files or pipes from a redirect router.
1504 File deliveries just didn't batch; pipe deliveries might have suffered
1507 PH/06 A failure to get a lock for a hints database would erroneously always say
1508 "Failed to get write lock", even when it was really a read lock.
1510 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1511 of 0600. This has been changed to use the value of the lockfile_mode
1512 option (which defaults to 0600).
1514 PH/08 Applied small patch from the Sieve maintainer.
1516 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1517 folder from quota calculations, a direct delivery into this folder messed
1518 up the contents of the maildirsize file. This was because the regex was
1519 used only to exclude .Trash (or whatever) when the size of the mailbox
1520 was calculated. There was no check that a delivery was happening into an
1521 excluded directory. This bug has been fixed by ignoring all quota
1522 processing for deliveries into excluded directories.
1524 PH/10 Added the maildirfolder_create_regex option to appendfile.
1530 PH/01 The code for finding all the local interface addresses on a FreeBSD
1531 system running IPv6 was broken. This may well have applied to all BSD
1532 systems, as well as to others that have similar system calls. The broken
1533 code found IPv4 interfaces correctly, but gave incorrect values for the
1534 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1535 that it would not match correctly against @[] and not recognize the IPv6
1538 PH/02 The ipliteral router was not recognizing addresses of the form user@
1539 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1541 PH/03 Added disable_ipv6.
1543 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1544 decoded form, because it is most often used to construct To: headers
1545 lines in autoreplies, and the decoded form may well be syntactically
1546 invalid. However, $reply_address has leading white space removed, and all
1547 newlines turned into spaces so that the autoreply transport does not
1550 PH/05 If group was specified without a user on a router, and no group or user
1551 was specified on a transport, the group from the router was ignored.
1553 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1554 for visible compile-time settings that can be used to change these
1555 numbers, for those that want even more. Backwards compatibility with old
1556 spool files has been maintained. However, going back to a previous Exim
1557 release will lost any variables that are in spool files.
1559 PH/07 Two small changes when running in the test harness: increase delay when
1560 passing a TCP/IP connection to a new process, in case the original
1561 process has to generate a bounce, and remove special handling of
1562 127.0.0.2 (sic), which is no longer necessary.
1564 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1565 be the same on different OS.
1567 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1570 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1571 whether --show-vars was specified or not
1573 JJ/02 exipick: Added support for new ACL variable spool format introduced
1576 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1577 syntactically invalid From: or Reply-to: line, and a filter used this to
1578 generate an autoreply, and therefore failed to obtain an address for the
1579 autoreply, Exim could try to deliver to a non-existent relative file
1580 name, causing unrelated and misleading errors. What now happens is that
1581 it logs this as a hard delivery error, but does not attempt to create a
1584 PH/11 The exinext utility has a -C option for testing purposes, but although
1585 the given file was scanned by exinext itself; it wasn't being passed on
1586 when Exim was called.
1588 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1589 an end-of-file indication when reading a command response.
1591 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1592 compiled. In many other places in Exim, IPv6 addresses are always
1593 recognized, so I have changed this. It also means that IPv4 domain
1594 literals of the form [IPV4:n.n.n.n] are now always recognized.
1596 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1597 used if the router is not running as root, for example, when verifying at
1598 ACL time, or when using -bh. The debugging output from this situation was
1599 non-existent - all you got was a failure to exec. I have made two
1602 (a) Failures to set uid/gid, the current directory, or a process leader
1603 in a subprocess such as that created by queryprogram now generate
1604 suitable debugging ouput when -d is set.
1606 (b) The queryprogram router detects when it is not running as root,
1607 outputs suitable debugging information if -d is set, and then runs
1608 the subprocess without attempting to change uid/gid.
1610 PH/15 Minor change to Makefile for building test_host (undocumented testing
1613 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1614 additional section of a DNS packet that returns MX or SRV records.
1615 Instead, it always explicitly searches for A/AAAA records. This avoids
1616 major problems that occur when a DNS server includes only records of one
1617 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1618 fixed another bug: if SRV records were looked up and the corresponding
1619 address records were *not* found in the additional section, the port
1620 values from the SRV records were lost.
1622 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1623 using the correct key (the original address) when searching the retry
1624 rules in order to find which one to use for generating the retry hint.
1626 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1627 adding the default one. Similarly, if it contains a Reply-To: header, the
1628 errors_reply_to option, if set, is not used.
1630 PH/19 When calculating a retry time, Exim used to measure the "time since
1631 failure" by looking at the "first failed" field in the retry record. Now
1632 it does not use this if it is later than than the arrival time of the
1633 message. Instead it uses the arrival time. This makes for better
1634 behaviour in cases where some deliveries succeed, thus re-setting the
1635 "first failed" field. An example is a quota failure for a huge message
1636 when small messages continue to be delivered. Without this change, the
1637 "time since failure" will always be short, possible causing more frequent
1638 delivery attempts for the huge message than are intended.
1639 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1641 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1642 $1, $2, $3) because the numerical variables can be reset during some
1643 expansion items (e.g. "match"), thereby losing the authentication data.
1645 PH/21 Make -bV show the size of off_t variables so that the test suite can
1646 decide whether to run tests for quotas > 2G.
1648 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1649 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1650 filecount value is greater than 2G or if a quota value is greater than 2G
1651 on a system where the size of off_t is not greater than 4, a panic error
1654 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1655 never match. The debug and -bh output now contains an explicit error
1656 message indicating a malformed IPv4 address or mask.
1658 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1659 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1660 PH/23 above applies.
1662 PH/25 Do not write to syslog when running in the test harness. The only
1663 occasion when this arises is a failure to open the main or panic logs
1664 (for which there is an explicit test).
1666 PH/26 Added the /no_tell option to "control=freeze".
1668 PH/27 If a host name lookup failed very early in a connection, for example, if
1669 the IP address matched host_lookup and the reverse lookup yielded a name
1670 that did not have a forward lookup, an error message of the form "no IP
1671 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1672 could be logged. Now it outputs the IP address instead of "NULL".
1674 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1675 allows the sender and the authenticated sender to be set when
1676 submitting a message from within Exim. Since child_open_exim() is
1677 documented for local_scan(), the new function should be too.
1679 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1680 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1681 results in an empty string is now treated as unset.
1683 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1685 PH/31 Added sender_verify_fail logging option.
1687 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1688 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1689 tidied the source and removed it altogether.
1691 PH/33 When a queue run was abandoned because the load average was too high, a
1692 log line was always written; now it is written only if the queue_run log
1693 selector is set. In addition, the log line for abandonment now contains
1694 information about the queue run such as the pid. This is always present
1695 in "start" and "stop" lines but was omitted from the "abandon" line.
1697 PH/34 Omit spaces between a header name and the colon in the error message that
1698 is given when verify = headers_syntax fails (if there are lots of them,
1699 the message gets confusing).
1701 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1702 names, as there are now some PTR records that contain slashes. This check
1703 is only to protect against broken name servers that fall over on strange
1704 characters, so the fact that it applies to all lookups doesn't matter.
1706 PH/36 Now that the new test suite is complete, we can remove some of the
1707 special code in Exim that was needed for the old test suite. For example,
1708 sorting DNS records because real resolvers return them in an arbitrary
1709 order. The new test suite's fake resolver always returns records in the
1712 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1713 bounces) except when queue_only is set, to avoid logging races between
1714 the different processes.
1716 PH/38 Panic-die if .include specifies a non-absolute path.
1718 PH/39 A tweak to the "H" retry rule from its user.
1720 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1721 a label. They prevented compilation on older perls.
1723 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1724 a warning to be raised on newish perls.
1726 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1727 on queue. Changes to match documented behaviour of showing count of
1728 messages matching specified criteria.
1730 PH/40 Changed the default ident timeout from 30s to 5s.
1732 PH/41 Added support for the use of login_cap features, on those BSD systems
1733 that have them, for controlling the resources used by pipe deliveries.
1735 PH/42 The content-scanning code uses fopen() to create files in which to put
1736 message data. Previously it was not paying any attention to the mode of
1737 the files. Exim runs with umask(0) because the rest of the code creates
1738 files with open(), and sets the required mode explicitly. Thus, these
1739 files were ending up world-writeable. This was not a big issue, because,
1740 being within the spool directory, they were not world-accessible. I have
1741 created a function called modefopen, which takes an additional mode
1742 argument. It sets umask(777), creates the file, chmods it to the required
1743 mode, then resets the umask. All the relevant calls to fopen() in the
1744 content scanning code have been changed to use this function.
1746 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1747 to 24 hours. This avoids potential overflow problems when processing G
1748 and H retry rules. I suspect nobody ever tinkers with this value.
1750 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1752 PH/45 When the plaintext authenticator is running as a client, the server's
1753 challenges are checked to ensure they are valid base64 strings. By
1754 default, the authentication attempt is cancelled if an invalid string is
1755 received. Setting client_ignore_invalid_base64 true ignores these errors.
1756 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1757 they are received. Thus, the responses can be made to depend on the
1758 challenges. If an invalid string is ignored, an empty string is placed in
1761 PH/46 Messages that are created by the autoreply transport now contains a
1762 References: header, in accordance with RFCs 2822 and 3834.
1764 PH/47 Added authenticated_sender_force to the smtp transport.
1766 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1768 PH/49 Installed latest patch from the Sieve maintainer.
1770 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1771 was also set, the appendfile transport was unnecessarily scanning a
1772 directory of message files (e.g. for maildir delivery) to find the count
1773 of files (along with the size), even though it did not need this
1774 information. It now does the scan only if it needs to find either the
1775 size of the count of files.
1777 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1779 PH/52 Two bugs concerned with error handling when the smtp transport is
1782 (i) Exim was not creating retry information for temporary errors given
1783 for individual recipients after the DATA command when the smtp transport
1784 was used in LMTP mode. This meant that they could be retried too
1785 frequently, and not timed out correctly.
1787 (ii) Exim was setting the flag that allows error details to be returned
1788 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1789 recipients that were returned after the DATA command.
1791 PH/53 This is related to PH/52, but is more general: for any failing address,
1792 when detailed error information was permitted to be returned to the
1793 sender, but the error was temporary, then after the final timeout, only
1794 "retry timeout exceeded" was returned. Now it returns the full error as
1795 well as "retry timeout exceeded".
1797 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1798 do this, and (what is worse) MTAs that accept it.
1800 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1801 will now be deprecated.
1803 PH/56 New os.c-cygwin from the Cygwin maintainer.
1805 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1806 formats (previously only available in exim formats via -bpr, -bpru,
1807 and -bpra. Now also available in native and exiqgrep formats)
1809 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1810 with very large, slow to parse queues
1812 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1814 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1816 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1817 responses to authentication challenges, though it was showing the
1818 challenges; (ii) I've removed the CR characters from the debug output for
1821 PH/58 Allow for the insertion of a newline as well as a space when a string
1822 is turned into more than one encoded-word during RFC 2047 encoding. The
1823 Sieve code now uses this.
1825 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1826 data_4xx, lost_connection, tls_required.
1828 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1829 message was being sent as an SMTP response.
1831 PH/61 Add -l and -k options to exicyclog.
1833 PH/62 When verifying, if an address was redirected to one new address, so that
1834 verification continued, and the new address failed or deferred after
1835 having set something in $address_data, the value of $address_data was not
1836 passed back to the ACL. This was different to the case when no
1837 redirection occurred. The value is now passed back in both cases.
1839 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1840 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1841 use login_cap.h, so on its own it isn't the distinguishing feature. The
1842 new name refers directly to the setclassresources() function.
1844 PH/65 Added configuration files for NetBSD3.
1846 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1848 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1849 is preferred over IPv4.
1851 PH/68 The bounce_return_message and bounce_return_body options were not being
1852 honoured for bounces generated during the reception of non-SMTP messages.
1853 In particular, this applied to messages rejected by the ACL. This bug has
1854 been fixed. However, if bounce_return_message is true and bounce_return_
1855 body is false, the headers that are returned for a non-SMTP message
1856 include only those that have been read before the error was detected.
1857 (In the case of an ACL rejection, they have all been read.)
1859 PH/69 The HTML version of the specification is now built in a directory called
1860 spec_html instead of spec.html, because the latter looks like a path with
1861 a MIME-type, and this confuses some software.
1863 PH/70 Catch two compiler warnings in sieve.c.
1865 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1866 function verify_get_ident() calls ip_connect() to connect a socket, but
1867 if the "connect()" function timed out, ip_connect() used to close the
1868 socket. However, verify_get_ident() also closes the socket later, and in
1869 between Exim writes to the log, which may get opened at this point. When
1870 the socket was closed in ip_connect(), the log could get the same file
1871 descriptor number as the socket. This naturally causes chaos. The fix is
1872 not to close the socket in ip_connect(); the socket should be closed by
1873 the function that creates it. There was only one place in the code where
1874 this was missing, in the iplookup router, which I don't think anybody now
1875 uses, but I've fixed it anyway.
1877 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1878 well as to direct DNS lookups. Otherwise the handling of names in host
1879 lists is inconsistent and therefore confusing.
1885 PH/01 Two changes to the default runtime configuration:
1887 (1) Move the checks for relay_from_hosts and authenticated clients from
1888 after to before the (commented out) DNS black list checks.
1890 (2) Add control=submission to the relay_from_hosts and authenticated
1891 clients checks, on the grounds that messages accepted by these
1892 statements are most likely to be submissions.
1894 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1896 (1) Generate an error if the third argument for the ${prvs expansion is
1899 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1902 (3) Reset the variables that are obtained from the first argument of
1903 ${prvscheck and used in the second argument before leaving the code,
1904 because their memory is reclaimed, so using them afterwards may do
1907 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1908 one (it's much easier than Tom thought :-).
1910 (5) Because of (4), we can now allow for the use of $prvscheck_result
1911 inside the third argument.
1913 PH/03 For some reason, the default setting of PATH when running a command from
1914 a pipe transport was just "/usr/bin". I have changed it to
1917 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1918 anything to be listed in the output from -bV.
1920 PH/05 When a filter generated an autoreply, the entire To: header line was
1921 quoted in the delivery log line, like this:
1923 => >A.N.Other <ano@some.domain> <original@ddress> ...
1925 This has been changed so that it extracts the operative address. There
1926 may be more than one such address. If so, they are comma-separated, like
1929 => >ano@some.domain,ona@other.domain <original@ddress> ...
1931 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1932 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1933 not being looked up in the rDNS to get a host name, Exim was showing the
1934 IP address twice in Received: lines, even though the IP addresses were
1935 identical. For example:
1937 Received: from [1.2.3.4] (helo=[1.2.3.4])
1939 However, if the real host name was known, it was omitting the HELO data
1940 if it matched the actual IP address. This has been tidied up so that it
1941 doesn't show the same IP address twice.
1943 PH/07 When both +timestamp and +memory debugging was on, the value given by
1944 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1945 called by the debug printing, thereby overwriting the timestamp buffer.
1946 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1948 PH/08 When the original message was included in an autoreply transport, it
1949 always said "this is a copy of the message, including all the headers",
1950 even if body_only or headers_only was set. It now gives an appropriate
1953 PH/09 Applied a patch from the Sieve maintainer which:
1955 o fixes some comments
1956 o adds the (disabled) notify extension core
1957 o adds some debug output for the result of if/elsif tests
1958 o points to the current vacation draft in the documentation
1959 and documents the missing references header update
1963 o fixes a bug in processing the envelope test (when testing
1964 multiple envelope elements, the last element determinted the
1967 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1968 Electronic Mail") by including:
1970 Auto-submitted: auto-generated
1972 in the messages that it generates (bounce messages and others, such as
1973 warnings). In the case of bounce messages for non-SMTP mesages, there was
1974 also a typo: it was using "Auto_submitted" (underscore instead of
1975 hyphen). Since every message generated by Exim is necessarily in response
1976 to another message, thes have all been changed to:
1978 Auto-Submitted: auto-replied
1980 in accordance with these statements in the RFC:
1982 The auto-replied keyword:
1984 - SHOULD be used on messages sent in direct response to another
1985 message by an automatic process,
1987 - MUST NOT be used on manually-generated messages,
1989 - MAY be used on Delivery Status Notifications (DSNs) and Message
1990 Disposition Notifications (MDNs),
1992 - MUST NOT be used on messages generated by automatic or periodic
1993 processes, except for messages which are automatic responses to
1996 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1997 to the default Received: header definition.
1999 PH/12 Added log selector acl_warn_skipped (default on).
2001 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2002 variables because (a) they are in the wrong storage pool and (b) even if
2003 they were copied, it wouldn't work properly because of the caching.
2005 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2006 checking when decoding. Apparently there are clients that generate
2007 overlong encoded strings. Why am I not surprised?
2009 PH/15 If the first argument of "${if match_address" was not empty, but did not
2010 contain an "@" character, Exim crashed. Now it writes a panic log message
2011 and treats the condition as false.
2013 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2015 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2016 extension "envelope-auth". The code is finished and in agreement with
2017 other implementations, but there is no documentation so far and in fact,
2018 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2019 not changing the active code.
2021 Print executed "if" and "elsif" statements when debugging is used. This
2022 helps a great deal to understand what a filter does.
2024 Document more things not specified clearly in RFC3028. I had all this
2025 sorted out, when out of a sudden new issues came to my mind. Oops."
2027 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2030 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2031 canonical form (as documented). However, after a host name lookup from
2032 the IP address, check_host() was doing a simple string comparison with
2033 addresses acquired from the DNS when checking that the found name did
2034 have the original IP as one of its addresses. Since any found IPv6
2035 addresses are likely to be in abbreviated form, the comparison could
2036 fail. Luckily, there already exists a function for doing the comparison
2037 by converting both addresses to binary, so now that is used instead of
2038 the text comparison.
2040 PH/20 There was another similar case to PH/19, when a complete host name was
2041 given in a host list; looking up its IP address could give an abbreviated
2042 form, whereas the current host's name might or might not be abbreviated.
2043 The same fix has been applied.
2049 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2050 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2053 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2054 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2056 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2058 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2059 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2060 Also, for "Auto-Submitted", treat anything other than "no" as
2061 non-personal, in accordance with RFC 3834. (Previously it treated
2062 anything starting "auto-" as non-personal.)
2064 TF/01 The control=submission/name=... option had a problem with syntax
2065 errors if the name included a slash character. The /name= option
2066 now slurps the rest of the string, so it can include any characters
2067 but it must come last in the list of options (after /sender_retain
2070 PH/05 Some modifications to the interface to the fake nameserver for the new
2078 TK/01 Added the "success_on_redirect" address verification option. See
2079 NewStuff for rationale and an example.
2081 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2083 PH/02 Patch to exigrep to allow it to work on syslog lines.
2085 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2086 fread() to skip over the body file's header line, because in Cygwin the
2087 header line is locked and is inaccessible.
2089 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2090 co-exist for some time) to make it clear that it is the Exim ID that is
2091 referenced, not the Message-ID: header line.
2093 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2094 string_format() function, because snprintf() does not exist on all
2097 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2098 ${stat: expansion item.
2100 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2101 protocol synchronization error", to keep the pedants happy.
2103 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2104 well as for IRIX systems, when gcc is being used. See the host.c source
2107 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2109 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2112 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2113 transport and to the smtp transport in LMTP mode.
2115 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2117 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2118 run a filter in a subprocess. This could lead to confusion in subsequent
2119 lookups in the parent process. There should also be a search_tidyup() at
2120 the end of the subprocess.
2122 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2123 only if the host matched helo_try_verify_hosts, which caused the
2124 verification to occur when the EHLO/HELO command was issued. The ACL just
2125 tested the remembered result. Now, if a previous verification attempt has
2126 not happened, "verify = helo" does it there and then.
2128 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2130 TK/03 Fix log output including CR from clamd.
2132 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2133 exist provoked a memory error which could cause a segfault.
2135 PH/15 Installed PCRE 6.2
2137 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2139 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2140 of the problem. Specifically, suggested +O2 rather than +O1 for the
2143 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2145 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2146 record, the port was not correctly logged when the outgoing_port log
2147 selector was set (it logged the transort's default port).
2149 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2150 fallback_hosts, and "hosts" in the smtp transport.
2152 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2153 host errors such as "Connection refused".
2155 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2156 authentication with radiusclient 0.4.9:
2158 - Error returned from rc_read_config was caught wrongly
2159 - Username/password not passed on to radius server due to wrong length.
2161 The presumption is that some radiusclient API changes for 4.51/PH/17
2162 were not taken care of correctly. The code is still untested by me (my
2163 Linux distribution still has 0.3.2 of radiusclient), but it was
2164 contributed by a Radius user.
2166 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2167 expanding the "port" option of the smtp transport.
2169 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2170 while reading a MIME header. Thanks to Tom Hughes for a patch.
2172 PH/24 Include config.h inside local_scan.h so that configuration settings are
2175 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2176 that in an ACL for RCPT (for example), you can examine exactly what was
2179 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2180 commands, but it was not correctly comparing the address with the actual
2181 client host address. Thus, it would show the EHLO address in Received:
2182 header lines when this was not necessary.
2184 PH/27 Added the % operator to ${eval:}.
2186 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2187 it should be ignoring failures (because with -C, for example, it has lost
2188 privilege). It wasn't ignoring creation failures other than "already
2191 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2194 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2195 because an input line was too long, either on its own, or by virtue of
2196 too many continuations, the temporary file was not being removed, and the
2197 return code was incorrect.
2199 PH/31 Missing "BOOL" in function definition in filtertest.c.
2201 PH/32 Applied Sieve patches from the maintainer.
2203 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2205 PH/33 Added "verify = not_blind".
2207 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2208 Local/Makefile (with some defaults set). These are used in built scripts
2209 such as exicyclog, but they have never been used in the exim_install
2210 script (though there are many overriding facilities there). I have
2211 arranged that the exim_install script now takes note of these two
2214 PH/35 Installed configuration files for Dragonfly.
2216 PH/36 When a locally submitted message by a trusted user did not contain a
2217 From: header, and the sender address was obtained from -f or from an SMTP
2218 MAIL command, and the trusted user did not use -F to supply a sender
2219 name, $originator_name was incorrectly used when constructing a From:
2220 header. Furthermore, $originator_name was used for submission mode
2221 messages from external hosts without From: headers in a similar way,
2222 which is clearly wrong.
2224 PH/37 Added control=suppress_local_fixups.
2226 PH/38 When log_selector = +received_sender was set, and the addition of the
2227 sender made the log line's construction buffer exactly full, or one byte
2228 less than full, an overflow happened when the terminating "\n" was
2231 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2232 when the result of a list match is failure because a DNS lookup failed.
2234 PH/40 RM_COMMAND is now used in the building process.
2236 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2237 the "build-* directories that it finds.
2239 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2240 address in a domain literal was a prefix of an interface address.
2242 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2243 when verifying a sender address, unless rewrite_headers is false.
2245 PH/44 Wrote a long comment about why errors_to addresses are verified as
2246 recipients, not senders.
2248 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2249 the ratelimit ACL was added.
2251 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2253 PH/46 Added extra information about PostgreSQL errors to the error string.
2255 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2256 avoiding the need to install special zones in a real server. This is
2257 backwards compatible; if it can't find the fake resolver, it drops back.
2258 Thus, both old and new test suites can be run.
2260 TF/02 Added util/ratelimit.pl
2262 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2263 clock is set back in time.
2265 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2266 Candler <B.Candler@pobox.com>.
2268 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2269 OK for addresses that are the result of redirections.
2271 PH/48 A number of further additions for the benefit of the new test suite,
2272 including a fake gethostbyname() that interfaces to the fake DNS resolver
2275 TF/06 The fix for widen_domains has also been applied to qualify_single and
2276 search_parents which are the other dnslookup options that can cause
2279 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2282 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2284 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2285 service" if it encountered an MX record with an empty target hostname.
2286 The message is now "an MX or SRV record indicated no SMTP service".
2288 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2289 if the DNS of the sending site is misconfigured. This is quite a
2290 common situation. This change restores the behaviour of treating a
2291 helo verification defer as a failure.
2293 PH/51 If self=fail was set on a router, the bounce message did not include the
2294 actual error message.
2300 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2302 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2303 command itself ended in error, the underlying message about the transport
2304 filter timeout was being overwritten with the pipe command error. Now the
2305 underlying error message should be appended to the second error message.
2307 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2309 PH/02 Reduce the amount of output that "make" produces by default. Full output
2310 can still be requested.
2312 PH/03 The warning log line about a condition test deferring for a "warn" verb
2313 was being output only once per connection, rather than after each
2314 occurrence (because it was using the same function as for successful
2315 "warn" verbs). This seems wrong, so I have changed it.
2317 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2318 it should not have, which might have caused a crash in the right
2319 circumstances, but probably never did.
2321 PH/04 Installed a modified version of Tony Finch's patch to make submission
2322 mode fix the return path as well as the Sender: header line, and to
2323 add a /name= option so that you can make the user's friendly name appear
2326 TF/03 Added the control = fakedefer ACL modifier.
2328 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2329 Mark Lowes for thorough testing.
2331 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2333 TK/03 Merged latest SRS patch from Miles Wilton.
2335 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2336 with the definition in sysexits.h (which is #included earlier).
2337 Fortunately, Exim does not actually use EX_OK. The code used to try to
2338 preserve the sysexits.h value, by assumimg that macro definitions were
2339 scanned for macro replacements. I have been disabused of this notion,
2340 so now the code just undefines EX_OK before #including unistd.h.
2342 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2343 in the smtp transport. When a block could not be written in a single
2344 write() function, the timeout was being re-applied to each part-write.
2345 This seems wrong - if the receiver was accepting one byte at a time it
2346 would take for ever. The timeout is now adjusted when this happens. It
2347 doesn't have to be particularly precise.
2349 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2350 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2352 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2353 without a domain if the check to <postmaster@domain> fails.
2355 SC/01 Eximstats: added -xls and the ability to specify output files
2356 (patch written by Frank Heydlauf).
2358 SC/02 Eximstats: use FileHandles for outputing results.
2360 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2362 SC/04 Eximstats: fixed display of large numbers with -nvr option
2364 SC/05 Eximstats: fixed merging of reports with empty tables.
2366 SC/06 Eximstats: added the -include_original_destination flag
2368 SC/07 Eximstats: removed tabs and trailing whitespace.
2370 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2372 TK/06 MBOX spool code: Add real "From " MBOX separator line
2373 so the .eml file is really in mbox format (even though
2374 most programs do not really care). Patch from Alex Miller.
2376 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2377 The latter is generated from $received_to and is only set if the
2378 message has one envelope recipient. SA can use these headers,
2379 obviously out-of-the-box. Patch from Alex Miller.
2381 PH/08 The ${def test on a variable was returning false if the variable's
2382 value was "0", contrary to what the specification has always said!
2383 The result should be true unless the variable is empty.
2385 PH/09 The syntax error of a character other than { following "${if
2386 def:variable_name" (after optional whitespace) was not being diagnosed.
2387 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2388 accidental colon was present, for example, could give incorrect results.
2390 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2391 result is used (not including appendfile, where other changes are about
2394 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2395 This involved changing a lot of size variables from int to off_t. It
2396 should work with maildirs and everything.
2398 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2399 spamd dying while we are connected to it.
2401 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2404 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2405 function for BDB 4.3.
2407 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2409 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2410 the sources for PCRE have been re-arranged and more files are now
2413 PH/15 The code I had for printing potentially long long variables in PH/11
2414 above was not the best (it lost precision). The length of off_t variables
2415 is now inspected at build time, and an appropriate printing format (%ld
2416 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2417 to be "long long int" or "long int". This is needed for the internal
2418 formatting function string_vformat().
2420 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2421 the configuration file to be ":syslog", then the script "guesses" where
2422 the logs files are, rather than using the compiled in default. In our
2423 case the guess is not the same as the compiled default, so the script
2424 suddenly stopped working when I started to use syslog. The patch checks
2425 to see if log_file_path is "". If so, it attempts to read it from exim
2426 with no configuration file to get the compiled in version, before it
2427 falls back to the previous guessing code."
2429 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2430 implementing BATV in an Exim configuration. See NewStuff for the gory
2433 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2434 Makefile that are specific to HP-UX.
2436 PH/18 If the "use_postmaster" option was set for a recipient callout together
2437 with the "random" option, the postmaster address was used as the MAIL
2438 FROM address for the random test, but not for the subsequent recipient
2439 test. It is now used for both.
2441 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2442 patch removes a few documentation additions to RFC 3028, because the
2443 latest draft now contains them. It adds the new en;ascii-case comparator
2444 and a new error check for 8bit text in MIME parts. Comparator and
2445 require names are now matched exactly. I enabled the subaddress
2446 extension, but it is not well tested yet (read: it works for me)."
2448 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2449 rework some of the code of TK/09 above to avoid the hardwired use of
2450 "%lld" and "long long". Replaced the call to snprintf() with a call to
2453 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2454 records point to non-existent hosts", "retry timeout exceeded", and
2455 "retry time not reached for any host after a long failure period".
2457 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2458 experimental DomainKeys support:
2460 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2461 (2) On an error such as an illegally used "control", the wrong name for
2462 the control was given.
2464 These problems did NOT occur unless DomainKeys support was compiled.
2466 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2468 PH/24 Added ${if match_ip condition.
2470 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2471 databases so that it will be absolutely obvious if a crash occurs in the
2472 DB library. This is a regular occurrence (often caused by mis-matched
2475 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2476 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2477 picked up on a user's system that detects such things. There doesn't seem
2478 to be a gcc warning option for this - only an attribute that has to be
2479 put on the function's prototype. It seems that in Fedora Core 4 they have
2480 set this on a number of new functions. No doubt there will be more in due
2483 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2484 specify a transport. However, if an address that was verified by such a
2485 router was the subject of a callout, Exim crashed because it tried to
2486 read the rcpt_include_affixes from the non-existent transport. Now it
2487 just assumes that the setting of that option is false. This bug was
2488 introduced by 4.51/PH/31.
2490 PH/28 Changed -d+all to exclude +memory, because that information is very
2491 rarely of interest, but it makes the output a lot bigger. People tend to
2492 do -d+all out of habit.
2494 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2495 code in os-type was giving problems when libc.so lives in lib64, like on
2498 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2499 aren't the modern standard, and it seems that some systems' include files
2500 don't always have them. Exim was already checking for some of the newer
2501 ones like T_AAAA, and defining it itself. I've added checks for all the
2502 record types that Exim uses.
2504 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2505 not automatically generating a new one, as it is supposed to. This
2506 prevented TLS from working. If the file did exist, but contained invalid
2507 data, a new version was generated, as expected. It was only the case of a
2508 non-existent file that was broken.
2510 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2511 with a change in libdomainkeys > 0.64.
2513 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2514 from DNS. If the selector record carries the flag, it now has
2515 precedence over the domain-wide flag.
2517 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2519 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2520 the use of a port name that isn't defined in /etc/services) occurred, the
2521 message was deferred as in a normal delivery, and thus remained on the
2522 spool, instead of being failed because of the mua_wrapper setting. This
2523 is now fixed, and I tidied up some of the mua_wrapper messages at the
2526 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2527 the messages in a hash of arrays rather than using individual hashes.
2528 This is a bit cleaner and results in dramatic memory savings, albeit
2529 at a slight CPU cost.
2531 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2532 as requested by Marc Sherman.
2534 SC/10 Eximstats: added histograms for user specified patterns as requested
2537 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2539 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2540 fopen() in the content-scanning modules that did not already have it.
2546 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2547 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2549 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2551 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2553 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2554 to test Sieve filters that use "vacation".
2556 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2557 that changes the way the GnuTLS parameters are stored in the cache file.
2558 The new format can be generated externally. For backward compatibility,
2559 if the data in the cache doesn't make sense, Exim assumes it has read an
2560 old-format file, and it generates new data and writes a new file. This
2561 means that you can't go back to an older release without removing the
2564 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2565 work if there are any delivery delays because "one_time" forces the
2566 parent to be marked "delivered", so its unseen clone is never tried
2567 again. For this reason, Exim now forbids the simultaneous setting of
2570 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2571 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2572 entry if you want to know the details. The fix, however, neglected to
2573 consider the case when local delivery batching is involved. The test for
2574 "previously delivered" was not happening when checking to see if an
2575 address could be batched with a previous (undelivered) one; under
2576 certain circumstances this could lead to multiple deliveries to the same
2579 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2580 in its include files, and this causes problems building Exim.
2582 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2583 header_syntax) but Exim was just ignoring anything given after a slash.
2584 In particular, this caused confusion with an attempt to use "verify =
2585 reverse_host_lookup/defer_ok". An error is now given when options are
2586 supplied for verify items that do not have them. (Maybe reverse_host_
2587 lookup should have a defer_ok option, but that's a different point.)
2589 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2590 defined by RFC 821) to 2048, because there were problems with some AUTH
2591 commands, and RFC 1869 says the size should be increased for extended
2592 SMTP commands that take arguments.
2594 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2597 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2598 "unknown" error; now it says that the functionality isn't in the binary.
2600 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2601 an address' error message when a string expansion fails (syntax or
2602 whatever). Otherwise the password may appear in the log. Following change
2603 PH/42 below, there is no longer a chance of it appearing in a bounce
2606 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2608 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2609 of its addresses was ever tried. (Bugzilla bug #2.)
2611 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2612 the result incorrectly in the debug output. (It correctly added a newline
2613 to what was transported.)
2615 TF/01 Added $received_time.
2617 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2618 commented out examples of how to interface to a virus scanner and to
2619 SpamAssassin. Also added commented examples of av_scanner and
2620 spamd_address settings.
2622 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2623 and controls are allowed in which ACLs. There were a couple of minor
2624 errors. Some of the entries in the conditions table (which is a table of
2625 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2626 negation of where the condition IS allowed.
2628 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2630 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2631 header file does not have a version number, so I've had to invent a new
2632 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2633 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2634 radiusclient), but it was contributed by a Radius user.
2636 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2637 files or directories, for OpenSSL.
2639 PH/19 When an Exim process that is running as root has to create an Exim log
2640 file, it does so in a subprocess that runs as exim:exim so as to get the
2641 ownership right at creation (otherwise, other Exim processes might see
2642 the file with the wrong ownership). There was no test for failure of this
2643 fork() call, which would lead to the process getting stuck as it waited
2644 for a non-existent subprocess. Forks do occasionally fail when resources
2645 run out. I reviewed all the other calls to fork(); they all seem to check
2648 PH/20 When checking for unexpected SMTP input at connect time (before writing
2649 the banner), Exim was not dealing correctly with a non-positive return
2650 from the read() function. If the client had disconnected by this time,
2651 the result was a log entry for a synchronization error with an empty
2652 string after "input=" when read() returned zero. If read() returned -1
2653 (an event I could not check), uninitialized data bytes were printed.
2654 There were reports of junk text (parts of files, etc) appearing after
2657 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2659 PH/22 Added support for macro redefinition, and (re)definition in between
2660 driver and ACL definitions.
2662 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2663 forgetting to use the resulting value; it was using the unexpanded value.
2665 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2666 hadn't been configured. The fix is from Juergen Kreileder, who
2667 understands it better than I do:
2669 "Here's what I see happening with three configured cyrus_sasl
2670 authenticators configured (plain, login, cram-md5):
2672 On startup auth_cyrus_sasl_init() gets called for each of these.
2673 This means three calls to sasl_listmech() without a specified mech_list.
2674 => SASL tests which mechs of all available mechs actually work
2675 => three warnings about OTP not working
2676 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2678 With the patch, sasl_listmech() also gets called three times. But now
2679 SASL's mech_list option is set to the server_mech specified in the the
2680 authenticator. Or in other words, the answer from sasl_listmech()
2681 gets limited to just the mech you're testing for (which is different
2683 => the return list contains just 'plain' or 'login', 'cram-md5' or
2684 nothing depending on the value of ob->server_mech.
2686 I've just tested the patch: Authentication still works fine,
2687 unavailable mechs specified in the exim configuration are still
2688 caught, and the auth.log warnings about OTP are gone."
2690 PH/25 When debugging is enabled, the contents of the command line are added
2691 to the debugging output, even when log_selector=+arguments is not
2694 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2695 answer is "GNU", and only if the return is "GNU/something" is the answer
2698 PH/27 $acl_verify_message is now set immediately after the failure of a
2699 verification in an ACL, and so is available in subsequent modifiers. In
2700 particular, the message can be preserved by coding like this:
2702 warn !verify = sender
2703 set acl_m0 = $acl_verify_message
2705 Previously, $acl_verify_message was set only while expanding "message"
2706 and "log_message" when a very denied access.
2708 PH/28 Modified OS/os.c-Linux with
2710 -#ifndef OS_LOAD_AVERAGE
2711 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2713 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2714 nomenclature these days.)
2716 PH/29 Installed patch from the Sieve maintainer that adds the options
2717 sieve_useraddress and sieve_subaddress to the redirect router.
2719 PH/30 In these circumstances:
2720 . Two addresses routed to the same list of hosts;
2721 . First host does not offer TLS;
2722 . First host accepts first address;
2723 . First host gives temporary error to second address;
2724 . Second host offers TLS and a TLS session is established;
2725 . Second host accepts second address.
2726 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2727 and peerdn, if requested) that were in fact used only for the second
2730 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2731 attention to any local part prefix or suffix that was matched by the
2732 router that accepted the address. It now behaves in the same way as it
2733 does for delivery: the affixes are removed from the local part unless
2734 rcpt_include_affixes is set on the transport.
2736 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2737 timeout during the DATA phase of an incoming message.
2739 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2740 applied a patch sanctioned by the Sieve maintainer.
2742 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2743 the uid or gid is negative. A case of a negative gid caused this to be
2744 noticed. The fix allows for either to be negative.
2746 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2747 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2750 PH/36 The change PH/12 above was broken. Fixed it.
2752 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2753 the grounds that routing the same address twice would always produce the
2754 same answer. This might have been true once, but it is certainly no
2755 longer true now. Routing a child address may depend on the previous
2756 routing that produced that child. Some complicated redirection strategies
2757 went wrong when messages had multiple recipients, and made Exim's
2758 behaviour dependent on the order in which the addresses were given.
2760 I have moved the duplicate checking until after the routing is complete.
2761 Exim scans the addresses that are assigned to local and remote
2762 transports, and removes any duplicates. This means that more work will be
2763 done, as duplicates will always all be routed, but duplicates are
2764 presumably rare, so I don't expect this is of any significance.
2766 For deliveries to pipes, files, and autoreplies, the duplicate checking
2767 still happens during the routing process, since they are not going to be
2770 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2771 It corrects a timeout issue with spamd. This is Ian's comment: "The
2772 background is that sometimes spamd either never reads data from a
2773 connection it has accepted, or it never writes response data. The exiscan
2774 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2775 blindly assumes that writes won't block so it may never time out."
2777 PH/39 Allow G after quota size as well as K and M.
2779 PH/40 The value set for $authenticated_id in an authenticator may not contain
2780 binary zeroes or newlines because the value is written to log lines and
2781 to spool files. There was no check on this. Now the value is run through
2782 the string_printing() function so that such characters are converted to
2783 printable escape sequences.
2785 PH/41 $message_linecount is a new variable that contains the total number of
2786 lines in the message. Compare $body_linecount, which is the count for the
2789 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2790 bounce and delay warning messages, except in certain special cases, which
2793 (a) An SMTP error message from a remote host;
2794 (b) A message specified in a :fail: redirection;
2795 (c) A message specified in a "fail" command in a system filter;
2796 (d) A message specified in a FAIL return from the queryprogram router;
2797 (e) A message specified by the cannot_route_message router option.
2799 In these cases only, Exim does include the error details in bounce and
2800 warning messages. There are also a few cases where bland messages such
2801 as "unrouteable address" or "local delivery error" are given.
2803 PH/43 $value is now also set for the "else" part of a ${run expansion.
2805 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2806 being worked on, but at least Exim now implements the latest version to
2809 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2810 process to complete was treated as a delivery failure, a timeout while
2811 writing the message to the pipe was logged, but erroneously treated as a
2812 successful delivery. Such timeouts include transport filter timeouts. For
2813 consistency with the overall process timeout, these timeouts are now
2814 treated as errors, giving rise to delivery failures by default. However,
2815 there is now a new Boolean option for the pipe transport called
2816 timeout_defer, which, if set TRUE, converts the failures into defers for
2817 both kinds of timeout. A transport filter timeout is now identified in
2820 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2821 systems where "make" and "gmake" are different, calling "gmake" at top
2822 level broke things. I've arranged for the value of $(MAKE) to be passed
2823 from the Makefile to this script so that it can call the same version of
2827 A note about Exim versions 4.44 and 4.50
2828 ----------------------------------------
2830 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2831 changes of various kinds. As a consequence, a big documentation update was
2832 needed. This delayed the release for rather longer than seemed good, especially
2833 in the light of a couple of (minor) security issues. Therefore, the changes
2834 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2835 release. So 4.44 and 4.50 are in effect two different branches that both start
2838 I have left the 4.50 change log unchanged; it contains all the changes since
2839 4.43. The change log for 4.44 is below; many of its items are identical to
2840 those for 4.50. This seems to be the most sensible way to preserve the
2841 historical information.
2847 1. Minor wording change to the doc/README.SIEVE file.
2849 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2850 computation of the current number of files was incorrect.
2852 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2853 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2856 4. Give more explanation in the error message when the command for a transport
2857 filter fails to execute.
2859 5. There are several places where Exim runs a non-Exim command in a
2860 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2861 was being done only for the command run by the queryprogram router. It is
2862 now done for all such subprocesses. The other cases are: ${run, transport
2863 filters, and the commands run by the lmtp and pipe transports.
2865 6. Added CONFIGURE_GROUP build-time option.
2867 7. Some older OS have a limit of 256 on the maximum number of file
2868 descriptors. Exim was using setrlimit() to set 1000 as a large value
2869 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2870 systems. I've change it so that if it can't get 1000, it tries for 256.
2872 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2873 was an oversight, and furthermore, ever since the addition of extra
2874 controls (e.g. 4.43/32), the checks on when to allow different forms of
2875 "control" were broken. There should now be diagnostics for all cases when a
2876 control that does not make sense is encountered.
2878 9. Added the /retain_sender option to "control=submission".
2880 10. $recipients is now available in the predata ACL (oversight).
2882 11. Tidy the search cache before the fork to do a delivery from a message
2883 received from the command line. Otherwise the child will trigger a lookup
2884 failure and thereby defer the delivery if it tries to use (for example) a
2885 cached ldap connection that the parent has called unbind on.
2887 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2888 of $address_data from the recipient verification was clobbered by the
2889 sender verification.
2891 13. The value of address_data from a sender verification is now available in
2892 $sender_address_data in subsequent conditions in the ACL statement.
2894 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2896 15. Added a new option "connect=<time>" to callout options, to set a different
2899 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2900 was its contents. (It was OK if the option was not defined at all.)
2902 17. A "Completed" log line is now written for messages that are removed from
2903 the spool by the -Mrm option.
2905 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2906 information about exactly what failed.
2908 19. Added -dd to debug only the daemon process.
2910 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2911 handles timeouts, both on the server side and network timeouts. Renamed the
2912 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2914 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2915 It is now set to "smtps".
2917 22. $host_address is now set to the target address during the checking of
2918 ignore_target_hosts.
2920 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2921 being passed; this would have caused $sender_host_name to have been used if
2922 matching the list had actually called for a host name (not very likely,
2923 since this list is usually IP addresses). A host name is now passed as
2926 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2927 code that specifies a non-restarting handler (typically sigaction() in
2928 modern systems) in an attempt to fix a rare and obscure crash bug.
2930 25. Narrowed the window for a race in the daemon that could cause it to ignore
2931 SIGCHLD signals. This is not a major problem, because they are used only to
2932 wake it up if nothing else does.
2934 26. A malformed maildirsize file could cause Exim to calculate negative values
2935 for the mailbox size or file count. Odd effects could occur as a result.
2936 The maildirsize information is now recalculated if the size or filecount
2939 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2940 support for a long time. Removed HAVE_SYS_VFS_H.
2942 28. Installed the latest version of exipick from John Jetmore.
2944 29. In an address list, if the pattern was not a regular expression, an empty
2945 subject address (from a bounce message) matched only if the pattern was an
2946 empty string. Non-empty patterns were not even tested. This was the wrong
2947 because it is perfectly reasonable to use an empty address as part of a
2948 database query. An empty address is now tested by patterns that are
2949 lookups. However, all the other forms of pattern expect the subject to
2950 contain a local part and a domain, and therefore, for them, an empty
2951 address still always fails if the pattern is not itself empty.
2953 30. Exim went into a mad DNS loop when attempting to do a callout where the
2954 host was specified on an smtp transport, and looking it up yielded more
2955 than one IP address.
2957 31. Re-factored the code for checking spool and log partition space into a
2958 function that finds that data and another that does the check. The former
2959 is then used to implement four new variables: $spool_space, $log_space,
2960 $spool_inodes, and $log_inodes.
2962 32. The RFC2047 encoding function was originally intended for short strings
2963 such as real names; it was not keeping to the 75-character limit for
2964 encoded words that the RFC imposes. It now respects the limit, and
2965 generates multiple encoded words if necessary. To be on the safe side, I
2966 have increased the buffer size for the ${rfc2047: expansion operator from
2969 33. It is now permitted to omit both strings after an "if" condition; if the
2970 condition is true, the result is "true". As before, when the second string
2971 is omitted, a false condition yields an empty string. This makes it less
2972 cumbersome to write custom ACL and router conditions.
2974 34. Failure to deliver a bounce message always caused it to be frozen, even if
2975 there was an errors_to setting on the router. The errors_to setting is now
2978 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2979 canonical form (fully expanded) before being placed in
2980 $sender_host_address.
2982 36. The table in the code that translates DNS record types into text (T_A to
2983 "A" for instance) was missing entries for NS and CNAME. It is just possible
2984 that this could have caused confusion if both these types were looked up
2985 for the same domain, because the text type is used as part of Exim's
2986 per-process caching. But the chance of anyone hitting this buglet seems
2989 37. The dnsdb lookup has been extended in a number of ways.
2991 (1) There is a new type, "zns", which walks up the domain tree until it
2992 finds some nameserver records. It should be used with care.
2994 (2) There is a new type, "mxh", which is like "mx" except that it returns
2995 just the host names, not the priorities.
2997 (3) It is now possible to give a list of domains (or IP addresses) to be
2998 looked up. The behaviour when one of the lookups defers can be
2999 controlled by a keyword.
3001 (4) It is now possible to specify the separator character for use when
3002 multiple records are returned.
3004 38. The dnslists ACL condition has been extended: it is now possible to supply
3005 a list of IP addresses and/or domains to be looked up in a particular DNS
3008 39. Added log_selector=+queue_time_overall.
3010 40. When running the queue in the test harness, wait just a tad after forking a
3011 delivery process, to get repeatability of debugging output.
3013 41. Include certificate and key file names in error message when GnuTLS fails
3014 to set them up, because the GnuTLS error message doesn't include the name
3015 of the failing file when there is a problem reading it.
3017 42. Allow both -bf and -bF in the same test run.
3019 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3021 44. The "Exiscan patch" is now merged into the mainline Exim source.
3023 45. Sometimes the final signoff response after QUIT could fail to get
3024 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3025 < 0 before doing a fflush(). This bug looks as though it goes back to the
3026 introduction of TLS in release 3.20, but "sometimes" must have been rare
3027 because the tests only now provoked it.
3029 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3030 (this can affect the format of dates).
3032 47. exim_tidydb, when checking for the continued existence of a message for
3033 which it has found a message-specific retry record, was not finding
3034 messages that were in split spool directories. Consequently, it was
3035 deleting retry records that should have stayed in existence.
3037 48. Steve fixed some bugs in eximstats.
3039 49. The SPA authentication driver was not abandoning authentication and moving
3040 on to the next authenticator when an expansion was forced to fail,
3041 contradicting the general specification for all authenticators. Instead it
3042 was generating a temporary error. It now behaves as specified.
3044 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3045 (the order specifies the preference for clients). The order is now AES256,
3046 AES128, 3DES, ARCFOUR128.
3048 51. Small patch to Sieve code - explicitly set From: when generating an
3051 52. Exim crashed if a remote delivery caused a very long error message to be
3052 recorded - for instance if somebody sent an entire SpamAssassin report back
3053 as a large number of 550 error lines. This bug was coincidentally fixed by
3054 increasing the size of one of Exim's internal buffers (big_buffer) that
3055 happened as part of the Exiscan merge. However, to be on the safe side, I
3056 have made the code more robust (and fixed the comments that describe what
3059 53. Now that there can be additional text after "Completed" in log lines (if
3060 the queue_time_overall log selector is set), a one-byte patch to exigrep
3061 was needed to allow it to recognize "Completed" as not the last thing in
3064 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3065 patch that reportedly fixes this has been added. I am not expert enough to
3066 create a test for it. This is what the patch creator wrote:
3068 "I found a little strange behaviour of ldap code when working with
3069 Windows 2003 AD Domain, where users was placed in more than one
3070 Organization Units. When I tried to give exim partial DN, the exit code
3071 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3072 But simultaneously result of request was absolutely normal ldap result,
3073 so I produce this patch..."
3075 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3076 REFERENCE, so I have modified the code to exclude the patch when that macro
3079 55. Some experimental protocols are using DNS PTR records for new purposes. The
3080 keys for these records are domain names, not reversed IP addresses. The
3081 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3082 leaves it alone. Component reversal etc. now happens only for IP addresses.
3085 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3087 57. Double the size of the debug message buffer (to 2048) so that more of very
3088 long debug lines gets shown.
3090 58. The exicyclog utility now does better if the number of log files to keep
3091 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3093 59. Two changes related to the smtp_active_hostname option:
3095 (1) $smtp_active_hostname is now available as a variable.
3096 (2) The default for smtp_banner uses $smtp_active_hostname instead
3097 of $primary_hostname.
3099 60. The host_aton() function is supposed to be passed a string that is known
3100 to be a valid IP address. However, in the case of IPv6 addresses, it was
3101 not checking this. This is a hostage to fortune. Exim now panics and dies
3102 if the condition is not met. A case was found where this could be provoked
3103 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3104 components; fortuitously, this particular loophole had already been fixed
3105 by change 4.50/55 above.
3107 If there are any other similar loopholes, the new check in host_aton()
3108 itself should stop them being exploited. The report I received stated that
3109 data on the command line could provoke the exploit when Exim was running as
3110 exim, but did not say which command line option was involved. All I could
3111 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3112 running as the user.
3115 61. There was a buffer overflow vulnerability in the SPA authentication code
3116 (which came originally from the Samba project). I have added a test to the
3117 spa_base64_to_bits() function which I hope fixes it.
3120 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3121 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3123 63. The daemon start-up calls getloadavg() while still root for those OS that
3124 need the first call to be done as root, but it missed one case: when
3125 deliver_queue_load_max is set with deliver_drop_privilege. This is
3126 necessary for the benefit of the queue runner, because there is no re-exec
3127 when deliver_drop_privilege is set.
3129 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3130 This has been fixed.
3132 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3133 was in use, was not putting the data itself into the right store pool;
3134 consequently, it could be overwritten for a subsequent message in the same
3135 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3138 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3140 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3141 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3142 the function were treating the return as a boolean value, which happened to
3143 work because 0=false and not-0=true, but is not correct code.
3145 68. The host_aton() function was not handling scoped IPv6 addresses (those
3146 with, for example, "%eth0" on the end) correctly.
3148 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3149 negated items (that is, ~something) in unsigned ints. Some compilers
3150 apparently mutter when there is no cast.
3152 70. If an address verification called from an ACL failed, and did not produce a
3153 user-specific message (i.e. there was only a "system" message), nothing was
3154 put in $acl_verify_message. In this situation, it now puts the system
3157 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3158 session; change 4.31/43 added the unwanted input to the log line - except
3159 that it did not do this in the start of session case. It now does.
3161 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3162 This is wrong and can cause the other end to generate a synchronization
3163 error if it is another Exim or anything else that does the synchronization
3164 check. A QUIT command is no longer sent after a timeout.
3166 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3167 during host lookups.
3169 74. The defer_ok option of callout verification was not working if it was used
3170 when verifying addresses in header lines, that is, for this case:
3172 verify = header_sender/callout=defer_ok
3174 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3175 those file descriptors could be used for SMTP connections. If anything
3176 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3177 could be sent to the SMTP client, causing chaos. The daemon now opens
3178 stdin, stdout, and stderr to /dev/null when it puts itself into the
3181 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3182 log by default. The user can override this with suitable Perl magic.
3184 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3185 the log message when discard triggers, was not working for the DATA ACL or
3186 for the non-SMTP ACL.
3188 78. Error message wording change in sieve.c.
3190 79. If smtp_accept_max_per_host was set, the number of connections could be
3191 restricted to fewer than expected, because the daemon was trying to set up
3192 a new connection before checking whether the processes handling previous
3193 connections had finished. The check for completed processes is now done
3194 earlier. On busy systems, this bug wouldn't be noticed because something
3195 else would have woken the daemon, and it would have reaped the completed
3198 80. If a message was submitted locally by a user whose login name contained one
3199 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3200 It caused a spool format error. I have fixed the spool reading code. A
3201 related problem was that the "from" clause in the Received: line became
3202 illegal because of the space(s). It is now covered by ${quote_local_part.
3204 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3207 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3209 83. Patch from Sieve maintainer to fix unterminated string problem in
3210 "vacation" handling.
3212 84. Some minor changes to the Linux configuration files to help with other
3213 OS variants using glibc.
3215 85. One more patch for Sieve to update vacation handling to latest spec.
3218 ----------------------------------------------------
3219 See the note above about the 4.44 and 4.50 releases.
3220 ----------------------------------------------------
3226 1. Change 4.43/35 introduced a bug that caused file counts to be
3227 incorrectly computed when quota_filecount was set in an appendfile
3230 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3231 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3234 3. Give more explanation in the error message when the command for a transport
3235 filter fails to execute.
3237 4. There are several places where Exim runs a non-Exim command in a
3238 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3239 was being done only for the command run by the queryprogram router. It is
3240 now done for all such subprocesses. The other cases are: ${run, transport
3241 filters, and the commands run by the lmtp and pipe transports.
3243 5. Some older OS have a limit of 256 on the maximum number of file
3244 descriptors. Exim was using setrlimit() to set 1000 as a large value
3245 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3246 systems. I've change it so that if it can't get 1000, it tries for 256.
3248 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3249 was an oversight, and furthermore, ever since the addition of extra
3250 controls (e.g. 4.43/32), the checks on when to allow different forms of
3251 "control" were broken. There should now be diagnostics for all cases when a
3252 control that does not make sense is encountered.
3254 7. $recipients is now available in the predata ACL (oversight).
3256 8. Tidy the search cache before the fork to do a delivery from a message
3257 received from the command line. Otherwise the child will trigger a lookup
3258 failure and thereby defer the delivery if it tries to use (for example) a
3259 cached ldap connection that the parent has called unbind on.
3261 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3262 of $address_data from the recipient verification was clobbered by the
3263 sender verification.
3265 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3266 was its contents. (It was OK if the option was not defined at all.)
3268 11. A "Completed" log line is now written for messages that are removed from
3269 the spool by the -Mrm option.
3271 12. $host_address is now set to the target address during the checking of
3272 ignore_target_hosts.
3274 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3275 being passed; this would have caused $sender_host_name to have been used if
3276 matching the list had actually called for a host name (not very likely,
3277 since this list is usually IP addresses). A host name is now passed as
3280 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3281 code that specifies a non-restarting handler (typically sigaction() in
3282 modern systems) in an attempt to fix a rare and obscure crash bug.
3284 15. Narrowed the window for a race in the daemon that could cause it to ignore
3285 SIGCHLD signals. This is not a major problem, because they are used only to
3286 wake it up if nothing else does.
3288 16. A malformed maildirsize file could cause Exim to calculate negative values
3289 for the mailbox size or file count. Odd effects could occur as a result.
3290 The maildirsize information is now recalculated if the size or filecount
3293 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3294 support for a long time. Removed HAVE_SYS_VFS_H.
3296 18. Updated exipick to current release from John Jetmore.
3298 19. Allow an empty sender to be matched against a lookup in an address list.
3299 Previously the only cases considered were a regular expression, or an
3302 20. Exim went into a mad DNS lookup loop when doing a callout where the
3303 host was specified on the transport, if the DNS lookup yielded more than
3306 21. The RFC2047 encoding function was originally intended for short strings
3307 such as real names; it was not keeping to the 75-character limit for
3308 encoded words that the RFC imposes. It now respects the limit, and
3309 generates multiple encoded words if necessary. To be on the safe side, I
3310 have increased the buffer size for the ${rfc2047: expansion operator from
3313 22. Failure to deliver a bounce message always caused it to be frozen, even if
3314 there was an errors_to setting on the router. The errors_to setting is now
3317 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3318 canonical form (fully expanded) before being placed in
3319 $sender_host_address.
3321 24. Updated eximstats to version 1.33
3323 25. Include certificate and key file names in error message when GnuTLS fails
3324 to set them up, because the GnuTLS error message doesn't include the name
3325 of the failing file when there is a problem reading it.
3327 26. Expand error message when OpenSSL has problems setting up cert/key files.
3330 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3331 (this can affect the format of dates).
3333 28. exim_tidydb, when checking for the continued existence of a message for
3334 which it has found a message-specific retry record, was not finding
3335 messages that were in split spool directories. Consequently, it was
3336 deleting retry records that should have stayed in existence.
3338 29. eximstats updated to version 1.35
3339 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3340 1.35 - bugfix such that pie charts by volume are generated correctly
3342 30. The SPA authentication driver was not abandoning authentication and moving
3343 on to the next authenticator when an expansion was forced to fail,
3344 contradicting the general specification for all authenticators. Instead it
3345 was generating a temporary error. It now behaves as specified.
3347 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3348 (the order specifies the preference for clients). The order is now AES256,
3349 AES128, 3DES, ARCFOUR128.
3351 31. Small patch to Sieve code - explicitly set From: when generating an
3354 32. Exim crashed if a remote delivery caused a very long error message to be
3355 recorded - for instance if somebody sent an entire SpamAssassin report back
3356 as a large number of 550 error lines. This bug was coincidentally fixed by
3357 increasing the size of one of Exim's internal buffers (big_buffer) that
3358 happened as part of the Exiscan merge. However, to be on the safe side, I
3359 have made the code more robust (and fixed the comments that describe what
3362 33. Some experimental protocols are using DNS PTR records for new purposes. The
3363 keys for these records are domain names, not reversed IP addresses. The
3364 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3365 leaves it alone. Component reversal etc. now happens only for IP addresses.
3368 34. The host_aton() function is supposed to be passed a string that is known
3369 to be a valid IP address. However, in the case of IPv6 addresses, it was
3370 not checking this. This is a hostage to fortune. Exim now panics and dies
3371 if the condition is not met. A case was found where this could be provoked
3372 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3373 components; fortuitously, this particular loophole had already been fixed
3374 by change 4.50/55 or 4.44/33 above.
3376 If there are any other similar loopholes, the new check in host_aton()
3377 itself should stop them being exploited. The report I received stated that
3378 data on the command line could provoke the exploit when Exim was running as
3379 exim, but did not say which command line option was involved. All I could
3380 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3381 running as the user.
3384 35. There was a buffer overflow vulnerability in the SPA authentication code
3385 (which came originally from the Samba project). I have added a test to the
3386 spa_base64_to_bits() function which I hope fixes it.
3389 36. The daemon start-up calls getloadavg() while still root for those OS that
3390 need the first call to be done as root, but it missed one case: when
3391 deliver_queue_load_max is set with deliver_drop_privilege. This is
3392 necessary for the benefit of the queue runner, because there is no re-exec
3393 when deliver_drop_privilege is set.
3395 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3396 was in use, was not putting the data itself into the right store pool;
3397 consequently, it could be overwritten for a subsequent message in the same
3398 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3401 38. Sometimes the final signoff response after QUIT could fail to get
3402 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3403 < 0 before doing a fflush(). This bug looks as though it goes back to the
3404 introduction of TLS in release 3.20, but "sometimes" must have been rare
3405 because the tests only now provoked it.
3411 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3412 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3413 Now it returns an integer. A number of calls were still expecting a T/F
3414 return. Fortuitously, in all cases, the tests worked in OK situations,
3415 which is the norm. However, things would have gone wrong on any write
3416 failures on the smtp file descriptor. This function is used when sending
3417 messages over SMTP and also when doing verify callouts.
3419 2. When Exim is called to do synchronous delivery of a locally submitted
3420 message (the -odf or -odi options), it no longer closes stderr before doing
3423 3. Implemented the mua_wrapper option.
3425 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3427 5. Implemented the functions header_remove(), header_testname(),
3428 header_add_at_position(), and receive_remove_recipient(), and exported them
3431 6. If an ACL "warn" statement specified the addition of headers, Exim already
3432 inserted X-ACL-Warn: at the start if there was no header name. However, it
3433 was not making this test for the second and subsequent header lines if
3434 there were newlines in the string. This meant that an invalid header could
3435 be inserted if Exim was badly configured.
3437 7. Allow an ACL "warn" statement to add header lines at the start or after all
3438 the Received: headers, as well as at the end.
3440 8. Added the rcpt_4xx retry error code.
3442 9. Added postmaster_mailfrom=xxx to callout verification option.
3444 10. Added mailfrom=xxxx to the callout verification option, for verify=
3447 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3448 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3450 12. Inserted some casts to stop certain compilers warning when using pointer
3451 differences as field lengths or precisions in printf-type calls (mostly
3452 affecting debugging statements).
3454 13. Added optional readline() support for -be (dynamically loaded).
3456 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3457 same clock tick as a message's arrival, so that its received time was the
3458 same as the "first fail" time on the retry record, and that message
3459 remained on the queue past the ultimate address timeout, every queue runner
3460 would try a delivery (because it was past the ultimate address timeout) but
3461 after another failure, the ultimate address timeout, which should have then
3462 bounced the address, did not kick in. This was a "< instead of <=" error;
3463 in most cases the first failure would have been in the next clock tick
3464 after the received time, and all would be well.
3466 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3467 being recognized when the domain list was tested by the match_domain
3468 condition in an expansion string.
3470 16. Added the ${str2b64: operator.
3472 17. Exim was always calling setrlimit() to set a large limit for the number of
3473 processes, without checking whether the existing limit was already
3474 adequate. (It did check for the limit on file descriptors.) Furthermore,
3475 errors from getrlimit() and setrlimit() were being ignored. Now they are
3476 logged to the main and panic logs, but Exim does carry on, to try to do its
3477 job under whatever limits there are.
3479 18. Imported PCRE 5.0.
3481 19. Trivial typo in log message " temporarily refused connection" (the leading
3484 20. If the log selector return_path_on_delivery was set and an address was
3485 redirected to /dev/null, the delivery process crashed because it assumed
3486 that a return path would always be set for a "successful" delivery. In this
3487 case, the whole delivery is bypassed as an optimization, and therefore no
3490 21. Internal re-arrangement: the function for sending a challenge and reading
3491 a response while authentication was assuming a zero-terminated challenge
3492 string. It's now changed to take a pointer and a length, to allow for
3493 binary data in such strings.
3495 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3497 23. Exim was not respecting finduser_retries when seeking the login of the
3498 uid under which it was called; it was always trying 10 times. (The default
3499 setting of finduser_retries is zero.) Also, it was sleeping after the final
3500 failure, which is pointless.
3502 24. Implemented tls_on_connect_ports.
3504 25. Implemented acl_smtp_predata.
3506 26. If the domain in control=submission is set empty, Exim assumes that the
3507 authenticated id is a complete email address when it generates From: or
3508 Sender: header lines.
3510 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3511 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3512 chown and chgrp in /bin and hostname in /usr/bin.
3514 28. Exim was keeping the "process log" file open after each use, just as it
3515 does for the main log. This opens the possibility of it remaining open for
3516 long periods when the USR1 signal hits a daemon. Occasional processlog
3517 errors were reported, that could have been caused by this. Anyway, it seems
3518 much more sensible not to leave this file open at all, so that is what now
3521 29. The long-running daemon process does not normally write to the log once it
3522 has entered its main loop, and it closes the log before doing so. This is
3523 so that log files can straightforwardly be renamed and moved. However,
3524 there are a couple of unusual error situations where the daemon does write
3525 log entries, and I had neglected to close the log afterwards.
3527 30. The text of an SMTP error response that was received during a remote
3528 delivery was being truncated at 512 bytes. This is too short for some of
3529 the long messages that one sometimes sees. I've increased the limit to
3532 31. It is now possible to make retry rules that apply only when a message has a
3533 specific sender, in particular, an empty sender.
3535 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3536 it possible to be selective about when SMTP synchronization is enforced.
3538 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3540 32. Implemented hosts_connection_nolog.
3542 33. Added an ACL for QUIT.
3544 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3547 35. Added mailbox_size and mailbox_filecount to appendfile.
3549 36. Added control = no_multiline_responses to ACLs.
3551 37. There was a bug in the logic of the code that waits for the clock to tick
3552 in the case where the clock went backwards by a substantial amount such
3553 that the microsecond fraction of "now" was more than the microsecond
3554 fraction of "then" (but the whole seconds number was less).
3556 38. Added support for the libradius Radius client library this is found on
3557 FreeBSD (previously only the radiusclient library was supported).
3563 1. When certain lookups returned multiple values in the form name=value, the
3564 quoting of the values was not always being done properly. Specifically:
3565 (a) If the value started with a double quote, but contained no whitespace,
3567 (b) If the value contained whitespace other than a space character (i.e.
3568 tabs or newlines or carriage returns) it was not quoted.
3569 This fix has been applied to the mysql and pgsql lookups by writing a
3570 separate quoting function and calling it from the lookup code. The fix
3571 should probably also be applied to nisplus, ibase and oracle lookups, but
3572 since I cannot test any of those, I have not disturbed their existing code.
3574 2. A hit in the callout cache for a specific address caused a log line with no
3575 reason for rejecting RCPT. Now it says "Previous (cached) callout
3576 verification failure".
3578 3. There was an off-by-one bug in the queryprogram router. An over-long
3579 return line was truncated at 256 instead of 255 characters, thereby
3580 overflowing its buffer with the terminating zero. As well as fixing this, I
3581 have increased the buffer size to 1024 (and made a note to document this).
3583 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3584 when Exim is waiting for an SMTP response from a remote server, Exim
3585 restarts its select() call on the socket, thereby resetting its timeout.
3586 This is not a problem when such interrupts are rare. Somebody set up a cron
3587 job to run exiwhat every 2 minutes, which is less than the normal select()
3588 timeout (5 or 10 minutes). This meant that the select() timeout never
3589 kicked in because it was always reset. I have fixed this by comparing the
3590 time when an interrupt arrives with the time at the start of the first call
3591 to select(). If more time than the timeout has elapsed, the interrupt is
3592 treated as a timeout.
3594 5. Some internal re-factoring in preparation for the addition of Sieve
3595 extensions (by MH). In particular, the "personal" test is moved to a
3596 separate function, and given an option for scanning Cc: and Bcc: (which is
3597 not set for Exim filters).
3599 6. When Exim created an email address using the login of the caller as the
3600 local part (e.g. when creating a From: or Sender: header line), it was not
3601 quoting the local part when it contained special characters such as @.
3603 7. Installed new OpenBSD configuration files.
3605 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3606 try to make them clearer.
3608 9. Callout options, other than the timeout value, were being ignored when
3609 verifying sender addresses in header lines. For example, when using
3611 verify = header_sender/callout=no_cache
3613 the cache was (incorrectly) being used.
3615 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3616 only the Cygwin environment.
3618 11. When return_path_on_delivery was set as a log selector, if different remote
3619 addresses in the same message used different return paths and parallel
3620 remote delivery occurred, the wrong values would sometimes be logged.
3621 (Whenever a remote delivery process finished, the return path value from
3622 the most recently started remote delivery process was logged.)
3624 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3625 lines when AUTH and/or TLS are in use. This is the "received protocol"
3626 field. Exim used to use "asmtp" for authenticated SMTP, without any
3627 indication (in the protocol name) for TLS use. Now it follows the RFC and
3628 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3629 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3630 names appear in log lines as well as in Received: header lines.
3632 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3633 extensions, and comparison tests, and to fix some bugs.
3635 14. Changes to the "personal" filter test:
3637 (1) The test was buggy in that it was just doing the equivalent of
3638 "contains" tests on header lines. For example, if a user's address was
3639 anne@some.where, the "personal" test would incorrectly be true for
3641 To: susanne@some.where
3643 This test is now done by extracting each address from the header in turn,
3644 and checking the entire address. Other tests that are part of "personal"
3645 are now done using regular expressions (for example, to check local parts
3646 of addresses in From: header lines).
3648 (2) The list of non-personal local parts in From: addresses has been
3649 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3650 taken from the Sieve specification recommendations.
3652 (3) If the message contains any header line starting with "List-" it is
3653 treated as non-personal.
3655 (4) The test for "circular" in the Subject: header line has been removed
3656 because it now seems ill-conceived.
3658 15. Minor typos in src/EDITME comments corrected.
3660 16. Installed latest exipick from John Jetmore.
3662 17. If headers_add on a router specified a text string that was too long for
3663 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3664 of string_sprintf() is now avoided.
3666 18. $message_body_size was not set (it was always zero) when running the DATA
3667 ACL and the local_scan() function.
3669 19. For the "mail" command in an Exim filter, no default was being set for
3670 the once_repeat time, causing a random time value to be used if "once" was
3671 specified. (If the value happened to be <= 0, no repeat happened.) The
3672 default is now 0s, meaning "never repeat". The "vacation" command was OK
3673 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3674 (I found it when inspecting the code).
3676 20. There is now an overall timeout for performing a callout verification. It
3677 defaults to 4 times the callout timeout, which applies to individual SMTP
3678 commands during the callout. The overall timeout applies when there is more
3679 than one host that can be tried. The timeout is checked before trying the
3680 next host. This prevents very long delays if there are a large number of
3681 hosts and all are timing out (e.g. when the network connections are timing
3682 out). The value of the overall timeout can be changed by specifying an
3683 additional sub-option for "callout", called "maxwait". For example:
3685 verify = sender/callout=5s,maxwait=20s
3687 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3688 to the end before writing, but this should make it even safer).
3690 22. Exim was forgetting that it had advertised PIPELINING for the second and
3691 subsequent messages on an SMTP connection. It was also not resetting its
3692 memory on STARTTLS and an internal HELO.
3694 23. When Exim logs an SMTP synchronization error within a session, it now
3695 records whether PIPELINING has been advertised or not.
3697 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3698 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3699 rather than long int.
3701 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3703 26. Added the never_mail option to autoreply.
3709 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3710 crash if the getsockname() call failed; this can happen if a connection is
3711 closed very soon after it is established. The problem was simply in the
3712 order in which certain operations were done, causing Exim to try to write
3713 to the SMTP stream before it had set up the file descriptor. The bug has
3714 been fixed by making things happen in the correct order.
3720 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3721 before the connection was closed, thus losing the rejection response.
3723 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3724 some early Solaris releases, but causes trouble in current releases where
3725 socklen_t is defined.
3727 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3730 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3733 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3734 that releasing the top of it at the end releases what was used for sub-
3735 expansions (unless the block got too big). However, discard this block if
3736 the first thing is a variable or header, so that we can use its block when
3737 it is dynamic (useful for very large $message_headers, for example).
3739 6. Lookups now cache *every* query, not just the most recent. A new, separate
3740 store pool is used for this. It can be recovered when all lookup caches are
3741 flushed. Lookups now release memory at the end of their result strings.
3742 This has involved some general refactoring of the lookup sources.
3744 7. Some code has been added to the store_xxx() functions to reduce the amount
3745 of flapping under certain conditions.
3747 8. log_incoming_interface used to affect only the <= reception log lines. Now
3748 it causes the local interface and port to be added to several more SMTP log
3749 lines, for example "SMTP connection from", and rejection lines.
3751 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3753 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3755 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3756 could be overwritten at the end of the current message (or the start of a
3757 new message if it was set in a HELO ACL). The value is now preserved for
3758 the duration of the SMTP connection.
3760 12. If a transport had a headers_rewrite setting, and a matching header line
3761 contained an unqualified address, that address was qualified, even if it
3762 did not match any rewriting rules. The underlying bug was that the values
3763 of the flags that permit the existence of unqualified sender and recipient
3764 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3765 non-local messages, and by -bnq for local messages) were not being
3766 preserved with the message after it was received.
3768 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3769 "next input=" as part of the text comprising the host identity instead of
3770 the correct text. The code was using the same buffer for two different
3771 strings. However, depending on which order the printing function evaluated
3772 its arguments, the bug did not always show up. Under Linux, for example, my
3773 test suite worked just fine.
3775 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3776 doesn't work with some older versions of Perl. It has been changed to "my",
3777 which in any case is probably the better facility to use.
3779 15. A really picky compiler found some instances of statements for creating
3780 error messages that either had too many or two few arguments for the format
3783 16. The size of the buffer for calls to the DNS resolver has been increased
3784 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3785 for addresses that have a lot of PTR records. This alleviates a problem; it
3786 does not fully solve it.
3788 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3789 buffer now truncates the list and logs the incident, which is the same
3790 action as happens when Exim is looking up a host name and its aliases.
3791 Previously in this situation something unpredictable would happen;
3792 sometimes it was "internal error: store_reset failed".
3794 18. If a server dropped the connection unexpectedly when an Exim client was
3795 using GnuTLS and trying to read a response, the client delivery process
3796 crashed while trying to generate an error log message.
3798 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3799 string, for example:
3801 warn message = H1: something\nH2: something
3803 the text was added as a single header line from Exim's point of view
3804 though it ended up OK in the delivered message. However, searching for the
3805 second and subsequent header lines using $h_h2: did not work. This has been
3806 fixed. Similarly, if a system filter added multiple headers in this way,
3807 the routers could not see them.
3809 20. Expanded the error message when iplsearch is called with an invalid key to
3810 suggest using net-iplsearch in a host list.
3812 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3813 ACLs are no longer actually imposed (and a message to that effect is
3816 22. If a "gecos" field in a passwd entry contained escaped characters, in
3817 particular, if it contained a \" sequence, Exim got it wrong when building
3818 a From: or a Sender: header from that name. A second bug also caused
3819 incorrect handling when an unquoted " was present following a character
3820 that needed quoting.
3822 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3823 was not being matched caselessly.
3825 24. Arranged for all hyphens in the exim.8 source to be escaped with
3828 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3829 properly. Recipient callout cache records were still being keyed to include
3830 the sender, even when use_sender was set false. This led to far more
3831 callouts that were necessary. The sender is no longer included in the key
3832 when use_sender is false.
3834 26. Added "control = submission" modifier to ACLs.
3836 27. Added the ${base62d: operator to decode base 62 numbers.
3838 28. dnsdb lookups can now access SRV records.
3840 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3841 the configuration file.
3843 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3844 (-v) mode. This makes the output for a verbose queue run more intelligible.
3846 31. Added a use_postmaster feature to recipient callouts.
3848 32. Added the $body_zerocount variable, containing the number of binary zero
3849 bytes in the message body.
3851 33. The time of last modification of the "new" subdirectory is now used as the
3852 "mailbox time last read" when there is a quota error for a maildir
3855 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3857 35. Added +ignore_unknown as a special item in host lists.
3859 36. Code for decoding IPv6 addresses in host lists is now included, even if
3860 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3861 address was recognized as an IP address, but was then not correctly decoded
3862 into binary, causing unexpected and incorrect effects when compared with
3869 1. Very minor rewording of debugging text in manualroute to say "list of
3870 hosts" instead of "hostlist".
3872 2. If verify=header_syntax was set, and a header line with an unqualified
3873 address (no domain) and a large number of spaces between the end of the
3874 name and the colon was received, the reception process suffered a buffer
3875 overflow, and (when I tested it) crashed. This was caused by some obsolete
3876 code that should have been removed. The fix is to remove it!
3878 3. When running in the test harness, delay a bit after writing a bounce
3879 message to get a bit more predictability in the log output.
3881 4. Added a call to search_tidyup() just before forking a reception process. In
3882 theory, someone could use a lookup in the expansion of smtp_accept_max_
3883 per_host which, without the tidyup, could leave open a database connection.
3885 5. Added the variables $recipient_data and $sender_data which get set from a
3886 lookup success in an ACL "recipients" or "senders" condition, or a router
3887 "senders" option, similar to $domain_data and $local_part_data.
3889 6. Moved the writing of debug_print from before to after the "senders" test
3892 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3893 problems for message scanning, either using a data ACL, or using
3894 local_scan() because the Received: header was not generated till after they
3895 were called (in order to set the time as the time of reception completion).
3896 I have revised the way this works. The header is now generated after the
3897 body is received, but before the ACL or local_scan() are called. After they
3898 are run, the timestamp in the header is updated.
3904 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3905 before starting a queue runner without re-exec. This happened only when
3906 deliver_drop_privilege was set or when the Exim user was set to root. The
3907 effect of the bug was that timeouts during subsequent deliveries caused
3908 crashes instead of being properly handled. The handler is now left at its
3909 default (and expected) setting.
3911 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3912 message, again when deliver_drop_privilege is set or Exim is run as root.
3913 The bug described in (1) was not present in this case, but the tidying up
3914 of the other signals was missing. I have made the two cases consistent.
3916 3. The ignore_target_hosts setting on a manualroute router was being ignored
3917 for hosts that were looked up using the /MX notation.
3919 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3922 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3923 operated on the sender address. After changing the $sender_address to <>
3924 for the sender address verify, Exim was re-instated it as the original
3925 (before rewriting) address, but remembering that it had rewritten it, so it
3926 wasn't rewriting it again. This bug also had the effect of breaking the
3927 sender address verification caching when the sender address was rewritten.
3929 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3930 This has been changed so that if the ip literal address matches
3931 ignore_target_hosts, the router declines.
3933 7. Added expansion conditions match_domain, match_address, and match_local_
3934 part (NOT match_host).
3936 8. The placeholder for the Received: header didn't have a length field set.
3938 9. Added code to Exim itself and to exim_lock to test for a specific race
3939 condition that could lead to file corruption when using MBX delivery. The
3940 issue is with the lockfile that is created in /tmp. If this file is removed
3941 after a process has opened it but before that process has acquired a lock,
3942 there is the potential for a second process to recreate the file and also
3943 acquire a lock. This could lead to two Exim processes writing to the file
3944 at the same time. The added code performs the same test as UW imapd; it
3945 checks after acquiring the lock that its file descriptor still refers to
3946 the same named file.
3948 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3949 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3952 11. Added the smtp_active_hostname option. If used, this will typically be made
3953 to depend on the incoming interface address. Because $interface_address is
3954 not set up until the daemon has forked a reception process, error responses
3955 that can happen earlier (such as "too many connections") no longer contain
3958 12. If an expansion in a condition on a "warn" statement fails because a lookup
3959 defers, the "warn" statement is abandoned, and the next ACL statement is
3960 processed. Previously this caused the whole ACL to be aborted.
3962 13. Added the iplsearch lookup type.
3964 14. Added ident_timeout as a log selector.
3966 15. Added tls_certificate_verified as a log selector.
3968 16. Added a global option tls_require_ciphers (compare the smtp transport
3969 option of the same name). This controls incoming TLS connections.
3971 17. I finally figured out how to make tls_require_ciphers do a similar thing
3972 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3973 before starting the TLS session.
3975 18. Tabs are now shown as \t in -bP output.
3977 19. If the log selector return_path_on_delivery was set, Exim crashed when
3978 bouncing a message because it had too many Received: header lines.
3980 20. If two routers both had headers_remove settings, and the first one included
3981 a superfluous trailing colon, the final name in the first list and the
3982 first name in the second list were incorrectly joined into one item (with a
3983 colon in the middle).
3989 1. Added -C and -D options to the exinext utility, mainly to make it easier
3990 to include in the automated testing, but these could be helpful when
3991 multiple configurations are in use.
3993 2. The exinext utility was not formatting the output nicely when there was
3994 an alternate port involved in the retry record key, nor when there was a
3995 message id as well (for retries that were specific to a specific message
3996 and a specific host). It was also confused by IPv6 addresses, because of
3997 the additional colons they contain. I have fixed the IPv4 problem, and
3998 patched it up to do a reasonable job for IPv6.
4000 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4001 delivery, the log line now contains "pipelined" if PIPELINING was used.
4003 4. An SMTP transport process used to panic and die if the bind() call to set
4004 an explicit outgoing interface failed. This has been changed; it is now
4005 treated in the same way as a connect() failure.
4007 5. A reference to $sender_host_name in the part of a conditional expansion
4008 that was being skipped was still causing a DNS lookup. This no longer
4011 6. The def: expansion condition was not recognizing references to header lines
4012 that used bh_ and bheader_.
4014 7. Added the _cache feature to named lists.
4016 8. The code for checking quota_filecount in the appendfile transport was
4017 allowing one more file than it should have been.
4019 9. For compatibility with Sendmail, the command line option
4027 and sets the incoming protocol and host name (for trusted callers). The
4028 host name and its colon can be omitted when only the protocol is to be set.
4029 Note the Exim already has two private options, -pd and -ps, that refer to
4030 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4031 "s", but I don't think that's a major issue.
4033 10. A number of refactoring changes to the code, none of which should affect
4036 (a) The number of logging options was getting close to filling up the
4037 32-bit word that was used as a bit map. I have split them into two classes:
4038 those that are passed in the argument to log_write(), and those that are
4039 only ever tested independently outside of that function. These are now in
4040 separate 32-bit words, so there is plenty of room for expansion again.
4041 There is no change in the user interface or the logging behaviour.
4043 (b) When building, for example, log lines, the code previously used a
4044 macro that called string_cat() twice, in order to add two strings. This is
4045 not really sufficiently general. Furthermore, there was one instance where
4046 it was actually wrong because one of the argument was used twice, and in
4047 one call a function was used. (As it happened, calling the function twice
4048 did not affect the overall behaviour.) The macro has been replaced by a
4049 function that can join an arbitrary number of extra strings onto a growing
4052 (c) The code for expansion conditions now uses a table and a binary chop
4053 instead of a serial search (which was left over from when there were very
4054 few conditions). Also, it now recognizes conditions like "pam" even when
4055 the relevant support is not compiled in: a suitably worded error message is
4056 given if an attempt is made to use such a condition.
4058 11. Added ${time_interval:xxxxx}.
4060 12. A bug was causing one of the ddress fields not to be passed back correctly
4061 from remote delivery subprocesses. The field in question was not being
4062 subsequently used, so this caused to problems in practice.
4064 13. Added new log selectors queue_time and deliver_time.
4066 14. Might have fixed a bug in maildirsizefile handling that threw up
4067 "unexpected character" debug warnings, and recalculated the data
4068 unnecessarily. In any case, I expanded the warning message to give more
4071 15. Added the message "Restricted characters in address" to the statements in
4072 the default ACL that block characters like @ and % in local parts.
4074 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4075 Three changes have been made:
4077 (a) There was a serious bug; a negative response to MAIL caused the whole
4078 recipient domain to be cached as invalid, thereby blocking all messages
4079 to all local parts at the same domain, from all senders. This bug has
4080 been fixed. The domain is no longer cached after a negative response to
4081 MAIL if the sender used is not empty.
4083 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4086 (c) A new callout option, "use_sender" has been added for people who want
4087 the modified behaviour.
4093 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4096 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4097 indeed breaks things for older releases.
4099 3. Added additional logging to the case where there is a problem reading data
4100 from a filter that is running in a subprocess using a pipe, in order to
4101 try to track down a specific problem.
4103 4. Testing facility fudge: when running in the test harness and attempting
4104 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4105 getting "No route to host". Convert this to a timeout.
4107 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4110 6. Some OS don't have socklen_t but use size_t instead. This affects the
4111 fifth argument of getsockopt() amongst other things. This is now
4112 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4113 can be set for individual OS. I have set it for SunOS5, OSF1, and
4114 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4115 some earlier ones do not.
4117 7. Change 4.30/15 was not doing the test caselessly.
4119 8. The standard form for an IPv6 address literal was being rejected by address
4120 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4121 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4122 this, as well as the form without the "IPv6" on the front (but only when
4123 address literals are enabled, of course).
4125 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4127 10. Exim crashed if a message with an empty sender address specified by -f
4128 encountered a router with an errors_to setting. This could be provoked only
4129 by a command such as
4133 where an empty string was supplied; "<>" did not hit this bug.
4135 11. Installed PCRE release 4.5.
4137 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4138 remained set. It is now erased.
4140 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4141 times from message ids (which are base 36 rather than the normal 62).
4143 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4144 were being counted as actual protocol errors, and logged if the log
4145 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4146 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4147 and DATA following a set of rejected RCPTs do not count as protocol errors.
4148 In other words, Exim assumes they were pipelined, though this may not
4149 actually be the case. Of course, in all cases the client gets an
4150 appropriate error code.
4152 15. If a lookup fails in an ACL condition, a message about the failure may
4153 be available; it is used if testing the ACL cannot continue, because most
4154 such messages specify what the cause of the deferral is. However, some
4155 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4156 that caused an old message to be retained and used if a later statement
4157 caused a defer, replacing the real cause of the deferral.
4159 16. If an IP address had so many PTR records that the DNS lookup buffer
4160 was not large enough to hold them, Exim could crash while trying to process
4161 the truncated data. It now detects and logs this case.
4163 17. Further to 4.21/58, another change has been made: if (and only if) the
4164 first line of a message (the first header line) ends with CRLF, a bare LF
4165 in a subsequent header line has a space inserted after it, so as not to
4166 terminate the header.
4168 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4169 unnecessarily, used atoi() instead of strtol(), and didn't check the
4170 termination when getting file sizes from file names by regex.
4172 19. Completely re-implemented the support for maildirsize files, in the light
4173 of a number of problems with the previous contributed implementation
4174 (4.30/29). In particular:
4176 . If the quota is zero, the maildirsize file is maintained, but no quota is
4179 . If the maildir directory does not exist, it is created before any attempt
4180 to write a maildirsize file.
4182 . The quota value in the file is just a cache; if the quota is changed in
4183 the transport, the new value overrides.
4185 . A regular expression is available for excluding directories from the
4188 20. The autoreply transport checks the characters in options that define the
4189 message's headers; it allows continued headers, but it was checking with
4190 isspace() after an embedded newline instead of explicitly looking for a
4193 21. If all the "regular" hosts to which an address was routed had passed their
4194 expiry times, and had not reached their retry times, the address was
4195 bounced, even if fallback hosts were defined. Now Exim should go on to try
4198 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4199 equivalent code in the SMTP transport. Some hosts send humungous responses
4200 to HELO/EHLO, more than 1024 it seems.
4202 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4203 gives compiler warnings in some environments. I've now done it "properly",
4206 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4207 (because of problems with the built-in one) was declared to return uschar *
4208 instead of char *, causing compiler failure.
4210 25. Fixed a file descriptor leak when processing alias/forward files.
4212 26. Fixed a minor format string issue in dbfn.c.
4214 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4216 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4217 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4220 29. When a sender address is verified, it is cached, to save repeating the test
4221 when there is more than one recipient in a message. However, when the
4222 verification involves a callout, it is possible for different callout
4223 options to be set for different recipients. It is too complicated to keep
4224 track of this in the cache, so now Exim always runs a verification when a
4225 callout is required, relying on the callout cache for the optimization.
4226 The overhead is duplication of the address routing, but this should not be
4229 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4230 to be verified with callout=postmaster, and the main callout worked but the
4231 postmaster check failed, the verification correctly failed. However, if a
4232 subsequent RCPT command asked for sender verification *without* the
4233 postmaster check, incorrect caching caused this verification also to fail,
4236 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4237 it was not caching the DNS options (qualify_single, search_parents) that
4238 were used when the lookup failed. A subsequent lookup with different
4239 options therefore always gave the same answer, though there were cases
4240 where it should not have. (Example: a "domains = !$mx_any" option on a
4241 dnslookup router: the "domains" option is always processed without any
4242 widening, but the router might have qualify_single set.) Now Exim uses the
4243 cached value only when the same options are set.
4245 32. Added John Jetmore's "exipick" utility to the distribution.
4247 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4248 than a timeout (e.g. a certificate is required, and is not provided), an
4249 Exim server now closes the connection immediately. Previously it waited for
4250 the client to close - but if the client is SSL, it seems that they each
4251 wait for each other, leading to a delay before one of them times out.
4253 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4254 maintained 0.8.x compatibility because I don't think many are using it, and
4255 it is clearly obsolete.
4257 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4260 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4261 client certificate was expired. A simple patch fixes this, though I don't
4262 understand the full logic of why the verify callback is called multiple
4265 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4266 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4267 which causes problems with some clients (such as the Certicom SSL Plus
4268 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4269 disables the coutermeasure allowing Eudora to connect."
4271 38. Exim was not checking that a write() to a log file succeeded. This could
4272 lead to Bad Things if a log got too big, in particular if it hit a file
4273 size limit. Exim now panics and dies if it cannot write to a log file, just
4274 as it does if it cannot open a log file.
4276 39. Modified OS/Makefile-Linux so that it now contains
4278 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4280 The two -D definitions ensure that Exim is compiled with large file
4281 support, which makes it possible to handle log files that are bigger than
4284 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4285 instance) a domain was checked against a named list that involved a lookup,
4286 causing $domain_data to be set, then another domain was checked against the
4287 same list, then the first domain was re-checked, the value of $domain_data
4288 after the final check could be wrong. In particular, if the second check
4289 failed, it could be set empty. This bug probably also applied to
4292 41. The strip_trailing_dot option was not being applied to the address given
4293 with the -f command-line option.
4295 42. The code for reading a message's header from the spool was incrementing
4296 $received_count, but never initializing it. This meant that the value was
4297 incorrect (doubled) while delivering a message in the same process in which
4298 it was received. In the most common configuration of Exim, this never
4299 happens - a fresh exec is done - but it can happen when
4300 deliver_drop_privilege is set.
4302 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4303 it now includes up to 150 characters of the unexpected data in the log
4306 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4307 and building data strings. The size of both of these buffers was 10 000
4308 bytes - far larger than anybody would *ever* want, thought I. Needless to
4309 say, somebody hit the limit. I have increased the maximum line length to
4310 20 000 and the maximum data length of concatenated lines to 100 000. I have
4311 also fixed two bugs, because there was no checking on these buffers. Tsk,
4312 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4313 buffer is too small.
4315 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4316 lsearch lookups. Now it does.
4318 46. When parsing a route_list item in a manualroute router, a fixed-length
4319 buffer was used for the list of hosts. I made this 1024 bytes long,
4320 thinking that nobody would ever have a list of hosts that long. Wrong.
4321 Somebody had a whole pile of complicated expansion conditions, and the
4322 string was silently truncated, leading to an expansion error. It turns out
4323 that it is easier to change to an unlimited length (owing to other changes
4324 that have happened since this code was originally written) than to build
4325 structure for giving a limitation error. The length of the item that
4326 expands into the list of hosts is now unlimited.
4328 47. The lsearch lookup could not handle data where the length of text line was
4329 more than 4095 characters. Such lines were truncated, leading to shortened
4330 data being returned. It should now handle lines of any length.
4332 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4333 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4336 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4337 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4338 no longer happens in comment lines. A list of replacements is now placed
4339 at the head of all of the source files, except those whose only change is
4340 to replace PERL_COMMAND in the very first #! line.
4342 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4343 messages on the queue, with a bottom-up merge sort, using code contributed
4344 by Michael Haardt. This should make operations like -bp somewhat faster on
4345 large queues. It won't affect queue runners, except when queue_run_in_order
4348 51. Installed eximstats 1.31 in the distribution.
4350 52. Added support for SRV lookups to the dnslookup router.
4352 53. If an ACL referred to $message_body or $message_body_end, the value was not
4353 reset for any messages that followed in the same SMTP session.
4355 54. The store-handling optimization for building very long strings was not
4356 differentiating between the different store pools. I don't think this
4357 actually made any difference in practice, but I've tidied it.
4359 55. While running the routers to verify a sender address, $sender_address
4360 was still set to the sender address. This is wrong, because when routing to
4361 send a bounce to the sender, it would be empty. Therefore, I have changed
4362 it so that, while verifying a sender address, $sender_address is set to <>.
4363 (There is no change to what happens when verifying a recipient address.)
4365 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4366 target A or AAAA records (if not already returned) without resetting the
4367 qualify_single or search_parents options of the DNS resolver. These are
4368 inappropriate in this case because the targets of MX and SRV records must
4369 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4370 target that, when qualified, matched something in the local domain. These
4371 two options are now turned off when doing these lookups.
4373 57. It seems that at least some releases of Reiserfs (which does not have the
4374 concept of a fixed number of inodes) returns zero and not -1 for the
4375 number of available inodes. This interacted badly with check_spool_inodes,
4376 which assumed that -1 was the "no such thing" setting. What I have done is
4377 to check that the total number of inodes is greater than zero before doing
4378 the test of how many are available.
4380 58. When a "warn" ACL statement has a log_message modifier, the message is
4381 remembered, and not repeated. This is to avoid a lot of repetition when a
4382 message has many recipients that cause the same warning to be written.
4383 Howewer, Exim was preserving the list of already written lines for an
4384 entire SMTP session, which doesn't seem right. The memory is now reset if a
4385 new message is started.
4387 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4388 debugging output unless log_rewrite was also set.
4390 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4391 of (int)(handle) when we know that handle contains (void *)(-1).
4393 61. The Exim daemon panic-logs an error return when it closes the incoming
4394 connection. However "connection reset by peer" seems to be common, and
4395 isn't really an error worthy of noting specially, so that particular error
4398 62. When Exim is trying to find all the local interfaces, it used to panic and
4399 die if the ioctl to get the interface flags failed. However, it seems that
4400 on at least one OS (Solaris 9) it is possible to have an interface that is
4401 included in the list of interfaces, but for which you get a failure error
4402 for this call. This happens when the interface is not "plumbed" into a
4403 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4404 failure of the "get flags" call assumes that the interface is down.
4406 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4407 makes life easier for people who are doing arithmetic on fields extracted
4408 from dates, where you often get leading zeros that should not be
4409 interpreted as octal.
4411 64. Added qualify_domain to the redirect router, to override the global
4414 65. If a pathologically long header line contained very many addresses (the
4415 report of this problem mentioned 10 000) and each of them was rewritten,
4416 Exim could use up a very large amount of memory. (It kept on making new
4417 copies of the header line as it rewrote, and never released the old ones.)
4418 At the expense of a bit more processing, the header rewriting function has
4419 been changed so that it no longer eats memory in this way.
4421 66. The generation of the Received: header has been moved from the time that a
4422 message starts to be received, to the time that it finishes. The timestamp
4423 in the Received: header should now be very close to that of the <= log
4424 line. There are two side-effects of this change:
4426 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4427 logged header lines no longer include the local Received: line, because
4428 it has not yet been created. The same applies to a copy of the message
4429 that is returned to a non-SMTP sender when a message is rejected.
4431 (b) When a filter file is tested using -bf, no additional Received: header
4432 is added to the test message. After some thought, I decided that this
4435 This change does not affect the value of $received_for. It is still set
4436 after address rewriting, but before local_scan() is called.
4438 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4440 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4441 gave an unhelpful panic error message, and a defer error. I have managed to
4442 change this behaviour so that it now rejects any supplied certificate,
4443 which seems right, as the list of acceptable certificates is empty.
4445 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4446 gave an unhelpful defer error. I have not managed to make this reject any
4447 supplied certificates, but the error message it gives is "no certificate
4448 supplied", which is not helpful.
4450 70. exigrep's output now also includes lines that are not associated with any
4451 message, but which match the given pattern. Implemented by a patch from
4452 Martin Sluka, which also tidied up the Perl a bit.
4454 71. Recipient callout verification, like sender verification, was using <> in
4455 the MAIL FROM command. This isn't really the right thing, since the actual
4456 sender may affect whether the remote host accepts the recipient or not. I
4457 have changed it to use the actual sender in the callout; this means that
4458 the cache record is now keyed on a recipient/sender pair, not just the
4459 recipient address. There doesn't seem to be a real danger of callout loops,
4460 since a callout by the remote host to check the sender would use <>.
4461 [SEE ABOVE: changed after hitting problems.]
4463 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4464 temporary errors. However, in the case of such a code being given after
4465 the end of a data transmission (i.e. after ".") Exim was failing to write
4466 a retry record for the message. (Yes, there was some broken host that was
4467 actually sending 8xx at this point.)
4469 73. An unknown lookup type in a host list could cause Exim to panic-die when
4470 the list was checked. (An example that provoked this was putting <; in the
4471 middle of a list instead of at the start.) If this happened during a DATA
4472 ACL check, a -D file could be left lying around. This kind of configuration
4473 error no longer causes Exim to die; instead it causes a defer errror. The
4474 incident is still logged to the main and panic logs.
4476 74. Buglet left over from Exim 3 conversion. The message "too many messages
4477 in one connection" was written to the rejectlog but not the mainlog, except
4478 when address rewriting (yes!) was being logged.
4480 75. Added write_rejectlog option.
4482 76. When a system filter was run not as root (that is, when system_filter_user
4483 was set), the values of the $n variables were not being returned to the
4484 main process; thus, they were not subsequently available in the $sn
4487 77. Added +return_path_on_delivery log selector.
4489 78. A connection timeout was being treated differently from recipients deferred
4490 when testing hosts_max_try with a message that was older than the host's
4491 retry timeout. (The host should not be counted, thus allowing all hosts to
4492 be tried at least once before bouncing.) This may have been the cause of an
4493 occasionally reported bug whereby a message would remain on the queue
4494 longer than the retry timeout, but would be bounced if a delivery was
4495 forced. I say "may" because I never totally pinned down the problem;
4496 setting up timeout/retry tests is difficult. See also the next item.
4498 79. The ultimate address timeout was not being applied to errors that involved
4499 a combination of host plus message (for example, a timeout on a MAIL
4500 command). When an address resolved to a number of possible hosts, and they
4501 were not all tried for each delivery (e.g. because of hosts_max_try), a
4502 message could remain on the queue longer than the retry timeout.
4504 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4507 81. Fixed an obscure SMTP outgoing bug which required at least the following
4508 conditions: (a) there was another message waiting for the same server;
4509 (b) the server returned 5xx to all RCPT commands in the first message so
4510 that the message was not completed; (c) the server dropped the connection
4511 or gave a negative response to the RSET that Exim sends to abort the
4512 transaction. The observed case was a dropped connection after DATA that had
4513 been sent in pipelining mode. That is, the server had advertised PIPELINING
4514 but was not implementing it correctly. The effect of the bug was incorrect
4515 behaviour, such as trying another host, and this could lead to a crash.
4521 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4522 and daemon.c were passed as pointers to ints; they should have been
4523 pointers to socklen_t variables (which are typically unsigned ints).
4525 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4528 3. Fixed a really odd bug that affected only the testing scheme; patching a
4529 certain fixed string in the binary changed the value of another string that
4530 happened to be identical to the end of the original first string.
4532 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4533 name", it returns that address as the IP address. On some operating
4534 systems (e.g. Solaris), it also passes back the IP address string as the
4535 "host name". However, on others (e.g. Linux), it passes back an empty
4536 string. Exim wasn't checking for this, and was changing the host name to an
4537 empty string, assuming it had been canonicized.
4539 5. Although rare, it is permitted to have more than one PTR record for a given
4540 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4541 all the names associated with an address, because they do in Solaris.
4542 However, it seems that they do not in Linux for data that comes from the
4543 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4544 I found this out when I moved to a new Linux workstation and tried to run
4545 the Exim test suite.
4547 To get round this problem I have changed the code so that it now does its
4548 own call to the DNS to look up PTR records when searching for a host name.
4549 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4550 addresses that are only in /etc/hosts are still found.
4552 This behaviour is, however, controlled by an option called host_lookup_
4553 order, which defaults to "bydns:byaddr". If people want to use the other
4554 order, or indeed, just use one or the other means of lookup, they can
4555 specify it in this variable.
4557 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4558 some operating systems, this comes back from gethostbyaddr() as an empty
4559 string, and this is what Exim used to test for. However, it seems that in
4560 other systems, "." is yielded. Exim now tests for this case too.
4562 7. The values of check_spool_space and check_log_space are now held internally
4563 as a number of kilobytes instead of an absolute number of bytes. If a
4564 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4565 kilobyte. This means that much larger values can be stored.
4567 8. Exim monitor: an attempt to get the action menu when not actually pointing
4568 at a message produces an empty menu entitled "No message selected". This
4569 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4570 no entries in it ("Shell widget menu has zero width and/or height"). So I
4571 have added a single, blank menu entry in this case.
4573 9. Added ${quote_local_part.
4575 10. MIME decoding is now applied to the contents of Subject: header lines when
4578 11. Now that a reference to $sender_host_address automatically causes a reverse
4579 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4580 host lookup before query-style lookups in lists that might use this
4581 variable. This has therefore been abolished, and the "net-" prefix is no
4582 longer necessary for query-style lookups.
4584 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4585 have been a typo for LFLAGS, so it has been changed.
4587 13. The install script calls Exim with "-C /dev/null" in order to find the
4588 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4589 to be output. Howeve, since Exim outputs its version number before the
4590 error, it didn't break the script. It just looked ugly. I fixed this by
4591 always allowing "-C /dev/null" if the caller is root.
4593 14. Ignore overlarge ACL variable number when reading spool file - insurance
4594 against a later release with more variables having written the file.
4596 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4597 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4598 this, as well as the form without the "IPv6" on the front.
4600 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4601 OS/Makefile-Darwin file.
4603 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4604 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4606 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4607 variable. However, some OS use ints for inodes. I've added cast to long int
4608 to get rid of the compiler warning.
4610 19. I had forgotten to lock out "/../" in configuration file names when
4611 ALT_CONFIG_PREFIX was set.
4613 20. Routers used for verification do not need to specify transports. However,
4614 if such a router generated a host list, and callout was configured, Exim
4615 crashed, because it could not find a port number from the (non-existent)
4616 transport. It now assumes port 25 in this circumstance.
4618 21. Added the -t option to exigrep.
4620 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4621 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4622 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4623 them is set, LOOKUP_LSEARCH is forced.
4625 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4627 24. Added sender and host information to the "rejected by local_scan()" log
4628 line; previously there was no indication of these.
4630 25. Added .include_if_exists.
4632 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4633 Linux. It turns out that not all file systems support this. Apparently some
4634 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4635 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4636 not supported on the file descriptor, is now ignored when Exim is trying to
4637 sync a directory. This applies only to Linux.
4639 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4641 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4642 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4645 29. Added support for maildirsize files from supplied patch (modified a bit).
4647 30. The use of :fail: followed by an empty string could lead Exim to respond to
4648 sender verification failures with (e.g.):
4650 550 Verification failed for <xxx>
4651 550 Sender verify failed
4653 where the first response line was missing the '-' that indicates it is not
4654 the final line of the response.
4656 31. The loop for finding the name of the user that called Exim had a hardwired
4657 limit of 10; it now uses the value of finduser_retries, which is used for
4658 all other user lookups.
4660 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4663 33. Exim was neglecting to zero errno before one call of strtol() when
4664 expanding a string and expecting an integer value. On some systems this
4665 resulted in spurious "integer overflow" errors. Also, it was casting the
4666 result into an int without checking.
4668 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4669 did not work. The code looks as if it has *never* worked, though it appears
4670 to have been documented since at least releast 1.62. I have made it work.
4672 35. The "timeout_DNS" error in retry rules, also documented since at least
4673 1.62, also never worked. As it isn't clear exactly what this means, and
4674 clearly it isn't a major issue, I have abolished the feature by treating it
4675 as "timeout", and writing a warning to the main and panic logs.
4677 36. The display of retry rules for -brt wasn't always showing the error code
4680 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4681 timeout_connect_A, timeout_connect_MX.
4683 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4684 to the empty sender.
4686 39. The daemon was not analysing the content of -oX till after it had closed
4687 stderr and disconnected from the controlling terminal. This meant that any
4688 syntax errors were only noted on the panic log, and the return code from
4689 the command was 0. By re-arranging the code a little, I've made the
4690 decoding happen first, so such errors now appear on stderr, and the return
4691 code is 1. However, the actual setting up of the sockets still happens in
4692 the disconnected process, so errors there are still only recorded on the
4695 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4696 connections (as happens on some IP stacks) was logged at start up time as
4697 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4698 it from "IPv6 and IPv4", which means that two separate sockets are being
4701 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4702 says whether AF_INET or AF_INET6 was passed as an argument.
4704 42. Exiwhat output was messed up when time zones were included in log
4707 43. Exiwhat now gives more information about the daemon's listening ports,
4708 and whether -tls-on-connect was used.
4710 44. The "port" option of the smtp transport is now expanded.
4712 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4713 silently ignored. Now an error message is written to the main and panic
4716 46. There's a new ACL modifier called "logwrite" which writes to a log file
4717 as soon as it is encountered.
4719 47. Added $local_user_uid and $local_user_gid at routing time.
4721 48. Exim crashed when trying to verify a sender address that was being
4724 49. Exim was recognizing only a space character after ".include". It now also
4725 recognizes a tab character.
4727 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4728 extracting the relevant information from the specification. The man page no
4729 longer contains scrambled data for the -d option, and I've added a section
4730 at the front about calling Exim under different names.
4732 51. Added "extra_headers" argument to the "mail" command in filter files.
4734 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4737 53. Installed eximstats 1.29.
4739 54. Added transport_filter_timeout as a generic transport option.
4741 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4742 Cc: header lines. This was required by RFC 822, but it not required by RFC
4745 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4746 incoming messages that did not have them. Now it does so only if the
4747 message originates locally, that is, if there is no associated remote host
4748 address. When Resent- header lines are present, this applies to the Resent-
4749 lines rather than the non-Resent- lines.
4751 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4752 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4754 58. Messages for configuration errors now include the name of the main
4755 configuration file - useful now that there may be more than one file in a
4756 list (.included file names were always shown).
4758 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4759 for those rare installations that do not start the daemon as root or run it
4760 setuid root. I've cut out the call to initgroups() if the daemon is not
4763 60. The Exim user and group can now be bound into the binary as text strings
4764 that are looked up at the start of Exim's processing.
4766 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4768 62. Added $mailstore_basename variable.
4770 63. Installed patch to sieve.c from Michael Haardt.
4772 64. When Exim failed to open the panic log after failing to open the main log,
4773 the original message it was trying to log was written to stderr and debug
4774 output, but if they were not available (the usual case in production), it
4775 was lost. Now it is written to syslog before the two lines that record the
4776 failures to open the logs.
4778 65. Users' Exim filters run in subprocesses under the user's uid. It is
4779 possible for a "deliver" command or an alias in a "personal" command to
4780 provoke an address rewrite. If logging of address rewriting is configured,
4781 this fails because the process is not running as root or exim. There may be
4782 a better way of dealing with this, but for the moment (because 4.30 needs
4783 to be released), I have disabled address rewrite logging when running a
4784 filter in a non-root, non-exim process.
4790 1. The buildconfig auxiliary program wasn't quoting the value set for
4791 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4792 not defined. This bug was masked in 4.22 by the effect that was fixed in
4795 2. Some messages that were rejected after a message id was allocated were
4796 shown as "incomplete" by exigrep. It no longer does this for messages that
4797 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4799 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4800 have allow_domain_literals set, the ID did not get logged in the <= line.
4801 Domain literals are now always recognized in Message-ID: header lines.
4803 4. The first argument for a ${extract expansion item is the key name or field
4804 number. Leading and trailing spaces in this item were not being ignored,
4805 causing some misleading effects.
4807 5. When deliver_drop_privilege was set, single queue runner processes started
4808 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4809 same command in the process it spins off) were not dropping privilege.
4811 6. When the daemon running as "exim" started a queue runner, it always
4812 re-executed Exim in the spun-off process. This is a waste of effort when
4813 deliver_drop_privilege is set. The new process now just calls the
4814 queue-runner function directly.
4820 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4823 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4824 ignored. Though the use of -oP was forcing the writing of a pid file, it
4825 was always written to the default place.
4827 3. If the message "no IP address found for host xxxx" is generated during
4828 incoming verification, it is now followed by identification of the incoming
4829 connection (so you can more easily find what provoked it).
4831 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4833 5. Added some features to "harden" Exim a bit more against certain attacks:
4835 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4836 be put in Local/Makefile. This is like the never_users runtime option,
4837 but it cannot be overridden. The default setting is "root".
4839 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4840 prefix string with which any file named in a -C command line option
4843 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4844 is retained for -C and -D only if the caller of Exim is root. Without
4845 it, the exim user may also use -C and -D and retain privilege.
4847 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4848 command line option is disabled.
4850 6. Macro names set by the -D option must start with an upper case letter, just
4851 like macro names defined in the configuration file.
4853 7. Added "dereference=" facility to LDAP.
4855 8. Two instances of the typo "uknown" in the source files are fixed.
4857 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4858 the Configure-Makefile script screwed up while processing it.
4860 10. Incorporated PCRE 4.4.
4862 11. The SMTP synchronization check was not operating right at the start of an
4863 SMTP session. For example, it could not catch a HELO sent before the client
4864 waited for the greeting. There is now a check for outstanding input at the
4865 point when the greeting is written. Because of the duplex, asynchronous
4866 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4867 way, but not yet received, when the check is performed.
4869 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4870 on TCP/IP sockets, because this apparently causes some broken clients to
4873 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4874 unchanged) from the Cygwin maintainer.
4876 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4877 was supported instead of testing for mbx. Effectively a typo.
4879 15. The spa authenticator server code was not checking that the input it
4880 received was valid base64.
4882 16. The debug output line for the "set" modifier in ACLs was not showing the
4883 name of the variable that was being set.
4885 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4887 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4888 lookup on its own. Something else was needed to trigger the lookup. For
4889 example, a match in host_lookup or the need for a host name in a host list.
4890 Now, if $sender_host_name is referenced and the host name has not yet been
4891 looked up, a lookup is performed. If the lookup fails, the variable remains
4892 empty, and $host_lookup_failed is set to "1".
4894 19. Added "eqi" as a case-independent comparison operator.
4896 20. The saslauthd authentication condition could segfault if neither service
4897 nor realm was specified.
4899 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4900 error message that was logged was misleading, and incoming SMTP
4901 connections were dropped. The message is now more accurate, and temporary
4902 errors are given to SMTP connections.
4904 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4905 (except RSET) with a 421 error, until QUIT is received. However, it was
4906 failing to send a response to QUIT.
4908 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4909 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4910 ACL is now run after the helo_try_verify_hosts code.
4912 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4913 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4914 case-independent, so other case variants are also recognized). Apparently
4915 some systems use these upper case variants.
4917 25. If more than two messages were waiting for the same host, and a transport
4918 filter was specified for the transport, Exim sent two messages over the
4919 same TCP/IP connection, and then failed with "socket operation on non-
4920 socket" when it tried to send the third.
4922 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4924 27. The extern definition of crypt16() in expand.c was not being excluded when
4925 the OS had its own crypt16() function.
4927 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4928 as a preferred synonym for return_size_limit, both as an option and as an
4931 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4933 30. Changed the default configuration ACL to relax the local part checking rule
4934 for addresses that are not in any local domains. For these addresses,
4935 slashes and pipe symbols are allowed within local parts, but the sequence
4936 /../ is explicitly forbidden.
4938 31. SPA server authentication was not clearing the challenge buffer before
4941 32. log_message in a "warn" ACL statement was writing to the reject log as
4942 well as to the main log, which contradicts the documentation and doesn't
4943 seem right (because no rejection is happening). So I have stopped it.
4945 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4946 However, I am unable to do any testing of this.
4948 34. Fixed an infelicity in the appendfile transport. When checking directories
4949 for a mailbox, to see if any needed to be created, it was accidentally
4950 using path names with one or more superfluous leading slashes; tracing
4951 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4953 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4954 contents are added to the log line that is written for every discarded
4955 recipient. (Previously a log_message setting was ignored.)
4957 36. The ${quote: operator now quotes the string if it is empty.
4959 37. The install script runs exim in order to find its version number. If for
4960 some reason other than non-existence or emptiness, which it checks, it
4961 could not run './exim', it was installing it with an empty version number,
4962 i.e. as "exim-". This error state is now caught, and the installation is
4965 38. An argument was missing from the function that creates an error message
4966 when Exim fails to connect to the socket for saslauthd authentication.
4967 This could cause Exim to crash, or give a corrupted message.
4969 39. Added isip, isip4, and isip6 to ${if conditions.
4971 40. The ACL variables $acl_xx are now saved with the message, and can be
4972 accessed later in routers, transports, and filters.
4974 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4975 strings in the file are not string-expanded.
4977 42. If a MAIL command specified a SIZE value that was too large to fit into an
4978 int variable, the check against message_size_limit failed. Such values are
4979 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4980 day this will have to be increased, but I don't think I want to be around
4981 when emails are that large.
4988 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4989 iconv() is not standard in FreeBSD.
4991 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4992 IPv6 enabled. The observed symptom was a segmentation fault on return from
4993 the function os_common_find_running_interfaces() in src/os.c.
4995 3. In the check_special_case() function in daemon.c I had used "errno" as an
4996 argument name, which causes warnings on some systems. This was basically a
4997 typo, since it was named "eno" in the comments!
4999 4. The code that waits for the clock to tick (at a resolution of some fraction
5000 of a second) so as to ensure message-id uniqueness was always waiting for
5001 at least one whole tick, when it could have waited for less. [This is
5002 almost certainly not relevant at current processor speeds, where it is
5003 unlikely to ever wait at all. But we try to future-proof.]
5005 5. The function that sleeps for a time interval that includes fractions of a
5006 second contained a race. It did not block SIGALRM between setting the
5007 timer, and suspending (a couple of lines later). If the interval was short
5008 and the sigsuspend() was delayed until after it had expired, the suspension
5009 never ended. On busy systems this could lead to processes getting stuck for
5012 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5013 process, before it forks any delivery processes. The open lookup caching
5014 mechanism meant that the open file or database connection was passed into
5015 the delivery process. The problem was that delivery processes always tidy
5016 up cached lookup data. This could cause a problem for the next delivery
5017 process started by the queue runner, because the external queue runner
5018 process does not know about the closure. So the next delivery process
5019 still has data in the lookup cache. In the case of a file lookup, there was
5020 no problem because closing a file descriptor in a subprocess doesn't affect
5021 the parent. However, if the lookup was caching a connection to a database,
5022 the connection was closed, and the second delivery process was likely to
5023 see errors such as "PGSQL: query failed: server closed the connection
5024 unexpectedly". The problem has been fixed by closing all cached lookups
5025 in a queue runner before running a delivery process.
5027 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5028 seem to have the "const" qualifier which it has on other OS. I've
5031 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5032 *of the same type* (client or server) with the same public name that an
5033 error should be diagnosed.
5035 9. When Exim looked up a host name for an IP address, but failed to find the
5036 original IP address when looking up the host name (a safety check), it
5037 output the message "<ip address> does not match any IP for NULL", which was
5038 confusing, to say the least. The bug was that the host name should have
5039 appeared instead of "NULL".
5041 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5042 user that is built into the binary, and the -C or -D options is used, root
5043 privilege is dropped before the configuration file is read. In addition,
5044 logging is switched to stderr instead of the normal log files. If the
5045 configuration then re-defines the Exim user, the unprivileged environment
5046 is probably not what is expected, so Exim logs a panic warning message (but
5049 However, if deliver_drop_privilege is set, the unprivileged state may well
5050 be exactly what is intended, so the warning has been cut out in that case,
5051 and Exim is allowed to try to write to its normal log files.
5057 1. smtp_return_error_details was not giving details for temporary sender
5058 or receiver verification errors.
5060 2. Diagnose a configuration error if two authenticators have the same public
5063 3. Exim used not to create the message log file for a message until the first
5064 delivery attempt. This could be confusing when incoming messages were held
5065 for policy or load reasons. The message log file is now created at the time
5066 the message is received, and an initial "Received" line is written to it.
5068 4. The automatically generated man page for command line options had a minor
5069 bug that caused no ill effects; however, a more serious problem was that
5070 the procedure for building the man page automatically didn't always
5071 operate. Consequently, release 4.20 contains an out-of-date version. This
5072 shouldn't happen again.
5074 5. When building Exim with embedded Perl support, the script that builds the
5075 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5076 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5078 6. The freeze_tell option was not being used for messages that were frozen on
5079 arrival, either by an ACL or by local_scan().
5081 7. Added the smtp_incomplete_transaction log selector.
5083 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5084 was accepting AUTH without a new EHLO.
5086 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5087 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5090 10. Logging of TCP/IP connections (when configured) now happens in the main
5091 daemon process instead of the child process, so that the TCP/IP connection
5092 count is more accurate (but it can never be perfect).
5094 11. The use of "drop" in a nested ACL was not being handled correctly in the
5095 outer ACL. Now, if condition failure induced by the nested "drop" causes
5096 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5097 or "require"), the connection is dropped.
5099 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5100 that yield "discard" can now be used with an "accept" or a "discard" verb,
5101 but an error is generated for any others (because I can't see a useful way
5102 to define what should happen).
5104 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5105 are now processed in the same way as lines in the Exim configuration file.
5106 In particular, continuation lines are supported.
5108 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5110 15. Added -ti meaning -t -i.
5112 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5113 domains, and warn by logging if others are found.
5115 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5116 returned from GIFCONF when Exim is trying to find the list of interfaces on
5117 a host. The code in os.c has been modified to copy each ifreq to an aligned
5118 structure in all cases.
5120 Also, in some cases, the returned ifreq's were being copied to a 'struct
5121 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5122 means the last couple of bytes of an IPv6 address could be chopped if the
5123 ifreq contained only a normal sockaddr (14 bytes storage).
5125 18. Named domain lists were not supported in the hosts_treat_as_local option.
5126 An entry such as +xxxx was not recognized, and was treated as a literal
5129 19. Ensure that header lines added by a DATA ACL are included in the reject log
5130 if the ACL subsequently rejects the message.
5132 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5133 MD5 (which is deprecated).
5135 21. When testing a filter file using -bf, Exim was writing a message when it
5136 took the sender from a "From " line in the message, but it was not doing so
5137 when it took $return_path from a Return-Path: header line. It now does.
5139 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5140 with a valid header line field name (a series of printing characters
5141 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5143 23. Changed "disc" in the source to "disk" to conform to the documentation and
5144 the book and for uniformity.
5146 24. Ignore Sendmail's -Ooption=value command line item.
5148 25. When execve() failed while trying to run a command in a pipe transport,
5149 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5150 could be confused with a return value of 69 from the command itself. This
5151 has been changed to 127, the value the shell returns if it is asked to run
5152 a non-existent command. The wording for the related log line suggests a
5153 non-existent command as the problem.
5155 26. If received_header_text expands to an empty string, do not add a Received:
5156 header line to the message. (Well, it adds a token one on the spool, but
5157 marks it "old" so that it doesn't get used or transmitted.)
5159 27. Installed eximstats 1.28 (addition of -nt option).
5161 28. There was no check for failure on the call to getsockname() in the daemon
5162 code. This can fail if there is a shortage of resources on the system, with
5163 ENOMEM, for example. A temporary error is now given on failure.
5165 29. Contrary to the C standard, it seems that in some environments, the
5166 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5167 program. Exim now does this explicitly; it affects the formatting of
5168 timestamps using strftime().
5170 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5171 complaints. I've now initialized all the variables, to avoid this.
5173 32. Header lines added by a system filter were not being "seen" during
5174 transport-time rewrites.
5176 33. The info_callback() function passed to OpenSSL is set up with type void
5177 (*)(SSL *, int, int), as described somewhere. However, when calling the
5178 function (actually a macro) that sets it up, the type void(*)() is
5179 expected. I've put in a cast to prevent warnings from picky compilers.
5181 34. If a DNS black list lookup found a CNAME record, but there were no A
5182 records associated with the domain it pointed at, Exim crashed.
5184 35. If a DNS black list lookup returned more than one A record, Exim ignored
5185 all but the first. It now scans all returned addresses if a particular IP
5186 value is being sought. In this situation, the contents of the
5187 $dnslist_value variable are a list of all the addresses, separated by a
5190 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5191 to accept a host name and all its aliases if the forward lookup for any of
5192 them yielded the IP address of the incoming connection. Now it accepts only
5193 those names whose forward lookup yields the correct IP address. Any other
5194 names are discarded. This closes a loophole whereby a rogue DNS
5195 administrator could create reverse DNS records to break through a
5196 wildcarded host restriction in an ACL.
5198 37. If a user filter or a system filter that ran in a subprocess used any of
5199 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5200 the wrong values were passed to the pipe command ($thisaddress had the
5201 value of $0, $0 had the value of $1, etc). This bug was introduced by
5202 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5204 38. Improved the line breaking for long SMTP error messages from ACLs.
5205 Previously, if there was no break point between 40 and 75 characters, Exim
5206 left the rest of the message alone. Two changes have been made: (a) I've
5207 reduced the minimum length to 35 characters; (b) if it can't find a break
5208 point between 35 and 75 characters, it looks ahead and uses the first one
5209 that it finds. This may give the occasional overlong line, but at least the
5210 remaining text gets split now.
5212 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5213 file descriptors might be low, and that setting 1000 would always raise it.
5214 It turns out that in some environments, the limit is already over 1000 and
5215 that lowering it causes trouble. So now Exim takes care not to decrease it.
5217 40. When delivering a message, the value of $return_path is set to $sender_
5218 address at the start of routing (routers may change the value). By an
5219 oversight, this default was not being set up when an address was tested by
5220 -bt or -bv, which affected the outcome if any router or filter referred to
5223 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5224 log only when "message" or "log_message" are set. However, if one of the
5225 conditions was an address verification, or a call to a nested ACL, the
5226 messages generated by the underlying test were being passed through. This
5227 no longer happens. The underlying message is available in $acl_verify_
5228 message for both "message" and "log_message" expansions, so it can be
5229 passed through if needed.
5231 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5232 new expansion $bh_ to give the encoded byte string without charset
5233 translation. Translation happens only if iconv() is available; HAVE_ICONV
5234 indicates this at build time. HEADERS_CHARSET gives the charset to
5235 translate to; headers_charset can change it in the configuration, and
5236 "headers charset" can change it in an individual filter file.
5238 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5239 that creates RFC 2047 encoded "words" labels them as that charset instead
5240 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5241 expansion operator; (ii) when Exim creates a From: line for a local
5242 message; (iii) when a header line is rewritten to include a "phrase" part.
5244 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5245 buggy, causing it to skip the first lines of messages whose message ID
5246 ended in 'D'. This would not have bitten before Exim release 4.14, because
5247 message IDs were unlikely to end in 'D' before then. The effect was to have
5248 incorrect size information for certain domains.
5250 45. #include "config.h" was missing at the start of the crypt16.c module. This
5251 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5254 46. If there was a timeout during a "random" callout check, Exim treated it as
5255 a failure of the random address, and carried on sending RSET and the real
5256 address. If the delay was just some slowness somewhere, the response to the
5257 original RCPT would be taken as a response to RSET and so on, causing
5258 mayhem of various kinds.
5260 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5261 when I implemented it. It didn't allow for the fact that some option values
5262 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5263 the right test for positive values.
5265 48. Domain names in DNS records are case-independent. Exim always looks them up
5266 in lower case. Some resolvers return domain names in exactly the case they
5267 appear in the zone file, that is, they may contain uppercase letters. Not
5268 all resolvers do this - some return always lower case. Exim was treating a
5269 change of case by a resolver as a change of domain, similar to a widening
5270 of a domain abbreviation. This triggered its re-routing code and so it was
5271 trying to route what was effectively the same domain again. This normally
5272 caused routing to fail (because the router wouldn't handle the domain
5273 twice). Now Exim checks for this case specially, and just changes the
5274 casing of the domain that it ultimately uses when it transmits the message
5277 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5280 50. If a filter generated a file delivery with a non-absolute name (possible if
5281 no home directory exists for the router), the forbid_file option was not
5284 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5285 the existing equality matching.
5287 52. Exim was using ints instead of ino_t variables in some places where it was
5288 dealing with inode numbers.
5290 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5291 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5292 called TMPDIR, and if it finds it is different, it changes its value.
5294 54. The smtp_printf() function is now made available to local_scan() so
5295 additional output lines can be written before returning. There is also an
5296 smtp_fflush() function to enable the detection of a dropped connection.
5297 The variables smtp_input and smtp_batched_input are exported to
5300 55. Changed the default runtime configuration: the message "Unknown user"
5301 has been removed from the ACL, and instead placed on the localuser router,
5302 using the cannot_route_message feature. This means that any verification
5303 failures that generate their own messages won't get overridden. Similarly,
5304 the "Unrouteable address" message that was in the ACL for unverifiable
5305 relay addresses has also been removed.
5307 56. Added hosts_avoid_esmtp to the smtp transport.
5309 57. The exicyclog script was not checking for the esoteric option
5310 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5311 will work only if exicyclog is run under the appropriate euid.
5313 58. Following a discussion on the list, the rules by which Exim recognises line
5314 endings on incoming messages have been changed. The -dropcr and drop_cr
5315 options are now no-ops, retained only for backwards compatibility. The
5316 following line terminators are recognized: LF CRLF CR. However, special
5317 processing applies to CR:
5319 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5320 nor a local message in the state where . is a terminator.
5322 (ii) If a bare CR is encountered in a header line, an extra space is added
5323 after the line terminator so as not to end the header. The reasoning
5324 behind this is that bare CRs in header lines are most likely either
5325 to be mistakes, or people trying to play silly games.
5327 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5328 was being incorrectly given as 18 bytes larger than it should have been.
5329 This is a VOB (very old bug).
5331 60. This may never have affected anything current, but just in case it has:
5332 When the local host is found other than at the start of a list of hosts,
5333 the local host, those with the same MX, and any that follow, are discarded.
5334 When the list in question was part of a longer list of hosts, the following
5335 hosts (not currently being processed) were also being discarded. This no
5336 longer happens. I'm not sure if this situation could ever has previously
5339 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5342 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5343 header. This is something that is recommended in a new Internet Draft, and
5344 is something that is documented as being done by Sendmail. There are two
5345 possible values. For messages generated by the autoreply transport, Exim
5348 Auto-Submitted: auto-replied
5350 whereas for all other generated messages (e.g. bounces) it adds
5352 Auto-Submitted: auto-generated
5354 63. The "personal" condition in filters now includes a test for the
5355 Auto-Submitted: header. If it contains the string "auto-" the message it
5356 not considered personal.
5358 64. Added rcpt_include_affixes as a generic transport option.
5360 65. Added queue_only_override (default true).
5362 66. Added the syslog_duplication option.
5364 67. If what should have been the first header line of a message consisted of
5365 a space followed by a colon, Exim was mis-interpreting it as a header line.
5366 It isn't of course - it is syntactically invalid and should therefore be
5367 treated as the start of the message body. The misbehaviour could have
5368 caused a number of strange effects, including loss of data in subsequent
5369 header lines, and spool format errors.
5371 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5372 client host had authenticated. This control can now be exercised by an ACL
5373 for more flexibility.
5375 69. By default, callouts do not happen when testing with -bh. There is now a
5376 variant, -bhc, which does actually run the callout code, including
5377 consulting and updating the callout cache.
5379 70. Added support for saslauthd authentication, courtesy of Alexander
5382 71. If statvfs() failed on the spool or log directories while checking their
5383 size for availability, Exim confusingly gave the error "space shortage".
5384 Furthermore, in debugging mode it crashed with a floating point exception.
5385 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5386 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5387 serious problem, Exim now writes to the main and panic logs when this
5388 happens, with details of the failure. It then refuses to accept the
5389 incoming message, giving the message "spool directory problem" or "log
5390 directory problem" with a 421 code for SMTP messages.
5392 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5393 0, 1, and 2 exist, because some OS complain for execs without them (see
5394 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5395 descriptors. However, the code omitted to check that the open succeeded,
5396 causing mysterious errors if for some reason the permissions on /dev/null
5397 got screwed. Now Exim writes a message to the main and panic logs, and
5398 bombs out if it can't open /dev/null.
5400 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5401 interact so that it is all more flexible. It is supposed to remain
5402 backwards compatible. Also added extra_local_interfaces.
5404 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5405 to bomb out with an assertion failure - to the client this appears as a
5406 connection drop. This problem occurs in the part of the code that was taken
5407 from the Samba project. Fortunately, the assertion is in a very simple
5408 function, so I have fixed this by reproducing the function inline in the
5409 one place where it is called, and arranging for authentication to fail
5410 instead of killing the process with assert().
5412 75. The SPA client code was not working when the server requested OEM rather
5413 than Unicode encoding.
5415 76. Added code to make require_files with a specific uid setting more usable in
5416 the case where statting the file as root fails - usually a non-root-mounted
5417 NFS file system. When this happens and the failure is EACCES, Exim now
5418 forks a subprocess and does the per-uid checking as the relevant uid.
5420 77. Added process_log_path.
5422 78. If log_file_path was not explicitly set, a setting of check_log_space or
5423 check_log_inodes was ignored.
5425 79. If a space check for the spool or log partitions fails, the incident is now
5426 logged. Of course, in the latter case the data may get lost...
5428 80. Added the %p formatting code to string_format() so that it can be used to
5429 print addresses in debug_print(). Adjusted all the address printing in the
5430 debugging in store.c to use %p rather than %d.
5432 81. There was a concern that a line of code in smtp_in.c could overflow a
5433 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5434 initially expressed, the concern was not well-founded, because trailing
5435 spaces are removed early. However, if the trailing spaces were followed by
5436 a NULL, they did not get removed, so the overflow was possible. Two fixes
5439 (a) I re-wrote the offending code in a cleaner fashion.
5440 (b) If an incoming SMTP command contains a NULL character, it is rejected
5443 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5444 runs initgroups(), so that if the Exim user is in any additional groups,
5445 they will be used during message reception.
5451 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob