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.
92 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
94 PP/02 Harmonised TLS library version reporting; only show if debugging.
95 Layout now matches that introduced for other libraries in 4.74 PP/03.
97 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
99 PP/04 New "dns_use_edns0" global option.
101 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
104 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
105 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
107 TK/01 Updated PolarSSL code to 0.14.2.
108 Bugzilla 1097. Patch from Andreas Metzler.
110 PP/07 Catch divide-by-zero in ${eval:...}.
113 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
116 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
117 format-string attack -- SECURITY: remote arbitrary code execution.
119 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
120 time unintentionally subject to list matching rules, letting the header
121 cause arbitrary Exim lookups (of items which can occur in lists, *not*
122 arbitrary string expansion). This allowed for information disclosure.
124 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
125 INT_MIN/-1 -- value coerced to INT_MAX.
131 NM/01 Workround for PCRE version dependency in version reporting
134 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
135 This fixes portability to compilers other than gcc, notably
136 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
138 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
139 makefiles for portability to HP-UX and POSIX correctness.
141 PP/01 Permit LOOKUP_foo enabling on the make command-line.
142 Also via indented variable definition in the Makefile.
143 (Debugging by Oliver Heesakkers).
145 PP/02 Restore caching of spamd results with expanded spamd_address.
146 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
148 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
149 Improves build reliability. Fix from: Frank Elsner
151 NM/02 Fix wide character breakage in the rfc2047 coding
152 Fixes bug 1064. Patch from Andrey N. Oktyabrski
154 NM/03 Allow underscore in dnslist lookups
155 Fixes bug 1026. Patch from Graeme Fowler
157 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
158 Code patches from Adam Ciarcinski of NetBSD.
160 NM/04 Fixed exiqgrep to cope with mailq missing size issue
163 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
164 is logged, to avoid truncation. Patch from John Horne.
166 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
167 Patch from Jakob Hirsch.
169 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
170 SQL string expansion failure details.
171 Patch from Andrey Oktyabrski.
173 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
174 Patch from Simon Arlott.
176 PP/09 New lookups functionality failed to compile on old gcc which rejects
177 extern declarations in function scope.
178 Patch from Oliver Fleischmann
180 PP/10 Use sig_atomic_t for flags set from signal handlers.
181 Check getgroups() return and improve debugging.
182 Fixed developed for diagnosis in bug 927 (which turned out to be
185 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
186 Patch from Mark Zealey.
188 PP/12 Bugzilla 1056: Improved spamd server selection.
189 Patch from Mark Zealey.
191 PP/13 Bugzilla 1086: Deal with maildir quota file races.
192 Based on patch from Heiko Schlittermann.
194 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
195 Patch from Uwe Doering, sign-off by Michael Haardt.
197 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
198 variable declaration deep within a block. Bug and patch from
201 PP/15 lookups-Makefile IRIX compatibilty coercion.
203 PP/16 Make DISABLE_DKIM build knob functional.
205 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
206 Patch by Simon Arlott
208 TF/03 Fix valgrind.h portability to C89 compilers that do not support
209 variable argument macros. Our copy now differs from upstream.
215 TF/01 Failure to get a lock on a hints database can have serious
216 consequences so log it to the panic log.
218 TF/02 Log LMTP confirmation messages in the same way as SMTP,
219 controlled using the smtp_confirmation log selector.
221 TF/03 Include the error message when we fail to unlink a spool file.
223 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
224 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
225 for maintaining out-of-tree patches for some time.
227 PP/01 Bugzilla 139: Documentation and portability issues.
228 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
229 Handle per-OS dynamic-module compilation flags.
231 PP/02 Let /dev/null have normal permissions.
232 The 4.73 fixes were a little too stringent and complained about the
233 permissions on /dev/null. Exempt it from some checks.
234 Reported by Andreas M. Kirchwitz.
236 PP/03 Report version information for many libraries, including
237 Exim version information for dynamically loaded libraries. Created
238 version.h, now support a version extension string for distributors
239 who patch heavily. Dynamic module ABI change.
241 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
242 privilege escalation vulnerability whereby the Exim run-time user
243 can cause root to append content of the attacker's choosing to
246 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
249 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
250 If dropping privileges for untrusted macros, we disabled normal logging
251 on the basis that it would fail; for the Exim run-time user, this is not
252 the case, and it resulted in successful deliveries going unlogged.
253 Fixed. Reported by Andreas Metzler.
259 PP/01 Date: & Message-Id: revert to normally being appended to a message,
260 only prepend for the Resent-* case. Fixes regression introduced in
261 Exim 4.70 by NM/22 for Bugzilla 607.
263 PP/02 Include check_rfc2047_length in configure.default because we're seeing
264 increasing numbers of administrators be bitten by this.
266 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
268 PP/03 Bugzilla 994: added openssl_options main configuration option.
270 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
272 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
274 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
276 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
277 without a peer certificate, leading to a segfault because of an
278 assumption that peers always have certificates. Be a little more
279 paranoid. Problem reported by Martin Tscholak.
281 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
282 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
283 NB: ClamAV planning to remove STREAM in "middle of 2010".
284 CL also introduces -bmalware, various -d+acl logging additions and
285 more caution in buffer sizes.
287 PP/09 Implemented reverse_ip expansion operator.
289 PP/10 Bugzilla 937: provide a "debug" ACL control.
291 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
293 PP/12 Bugzilla 973: Implement --version.
295 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
297 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
299 PP/15 Bugzilla 816: support multiple condition rules on Routers.
301 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
302 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
303 ignore trailing whitespace.
305 JJ/02 prevent non-panic DKIM error from being sent to paniclog
307 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
310 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
311 Notification from Dr Andrew Aitchison.
313 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
314 ExtendedDetectionInfo response format.
315 Notification from John Horne.
317 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
320 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
321 XSL and documented dependency on system catalogs, with examples of how
324 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
327 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
328 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
329 configuration file which is writeable by the Exim user or group.
331 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
332 of configuration files to cover files specified with the -C option if
333 they are going to be used with root privileges, not just the default
336 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
337 option (effectively making it always true).
339 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
340 files to be used while preserving root privileges.
342 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
343 that rogue child processes cannot use them.
345 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
346 run-time user, instead of root.
348 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
349 Exim run-time user without dropping privileges.
351 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
352 result string, instead of calling string_vformat() twice with the same
355 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
356 for other users. Others should always drop root privileges if they use
357 -C on the command line, even for a whitelisted configure file.
359 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
361 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
367 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
368 $data_path, and $header_path variables; fixed documentation bugs and
371 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
372 exipick to access non-standard spools, including the "frozen" queue
375 NM/01 Bugzilla 965: Support mysql stored procedures.
376 Patch from Alain Williams
378 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
380 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
381 Patch from Andreas Metzler
383 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
384 Patch from Kirill Miazine
386 NM/05 Bugzilla 671: Added umask to procmail example.
388 JJ/03 installed exipick 20100323.0, fixing doc bug
390 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
391 directory. Notification and patch from Dan Rosenberg.
393 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
395 TK/02 Improve log output when DKIM signing operation fails.
397 MH/01 Treat the transport option dkim_domain as a colon separated
398 list, not as a single string, and sign the message with each element,
399 omitting multiple occurences of the same signer.
401 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
402 Bugzilla 985, 986. Patch by Simon Arlott
404 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
405 Patch by Simon Arlott
407 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
408 MBX locking. Notification from Dan Rosenberg.
414 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
416 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
418 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
420 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
422 NM/04 Bugzilla 915: Flag broken perl installation during build.
428 TK/01 Added patch by Johannes Berg that expands the main option
429 "spamd_address" if it starts with a dollar sign.
431 TK/02 Write list of recipients to X-Envelope-Sender header when building
432 the mbox-format spool file for content scanning (suggested by Jakob
435 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
436 (http://www.dcc-servers.net/) support via dccifd. Activated by
437 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
439 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
440 by Mark Daniel Reidel <mr@df.eu>.
442 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
443 When building exim an external PCRE library is now needed -
444 PCRE is a system library on the majority of modern systems.
445 See entry on PCRE_LIBS in EDITME file.
447 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
448 conversation. Added nologin parameter to request.
449 Patch contributed by Kirill Miazine.
451 TF/01 Do not log submission mode rewrites if they do not change the address.
453 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
455 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
456 log files in place. Contributed by Roberto Lima.
458 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
460 TF/03 Bugzilla 615: When checking the local_parts router precondition
461 after a local_part_suffix or local_part_prefix option, Exim now
462 does not use the address's named list lookup cache, since this
463 contains cached lookups for the whole local part.
465 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
466 Robert Millan. Documentation is in experimental-spec.txt.
468 TF/04 Bugzilla 668: Fix parallel build (make -j).
470 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
472 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
473 Patch provided by Jan Srzednicki.
475 TF/05 Leading white space used to be stripped from $spam_report which
476 wrecked the formatting. Now it is preserved.
478 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
479 that they are available at delivery time.
481 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
483 TF/08 TLS error reporting now respects the incoming_interface and
484 incoming_port log selectors.
486 TF/09 Produce a more useful error message if an SMTP transport's hosts
487 setting expands to an empty string.
489 NM/06 Bugzilla 744: EXPN did not work under TLS.
490 Patch provided by Phil Pennock.
492 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
493 Patch provided by Richard Godbee.
495 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
496 acl_smtp_notquit, added index entry.
498 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
499 Patch provided by Eugene Bujak.
501 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
502 accept(). Patch provided by Maxim Dounin.
504 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
505 Patch provided by Phil Pennock.
507 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
509 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
510 Patch provided by Brad "anomie" Jorsch.
512 NM/14 Bugzilla 622: Added timeout setting to transport filter.
513 Patch provided by Dean Brooks.
515 TK/05 Add native DKIM support (does not depend on external libraries).
517 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
518 Patch provided by Graeme Fowler.
520 NM/16 Bugzilla 851: Documentation example syntax fix.
522 NM/17 Changed NOTICE file to remove references to embedded PCRE.
524 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
527 NM/19 Bugzilla 745: TLS version reporting.
528 Patch provided by Phil Pennock.
530 NM/20 Bugzilla 167: bool: condition support.
531 Patch provided by Phil Pennock.
533 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
534 clients. Patch provided by Phil Pennock.
536 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
537 Patch provided by Brad "anomie" Jorsch.
539 NM/23 Bugzilla 687: Fix misparses in eximstats.
540 Patch provided by Heiko Schlittermann.
542 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
543 Patch provided by Heiko Schlittermann.
545 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
546 plus update to original patch.
548 NM/26 Bugzilla 799: Documentation correction for ratelimit.
550 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
551 Patch provided by David Brownlee.
553 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
555 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
557 NM/30 Bugzilla 888: TLS documentation bugfixes.
559 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
561 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
562 Unlike the original bugzilla I have changed all shell scripts in src tree.
564 NM/33 Bugzilla 898: Transport filter timeout fix.
565 Patch by Todd Rinaldo.
567 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
568 Patch by Serge Demonchaux.
570 NM/35 Bugzilla 39: Base64 decode bug fixes.
571 Patch by Jakob Hirsch.
573 NM/36 Bugzilla 909: Correct connect() call in dcc code.
575 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
577 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
579 NM/39 Bugzilla 911: Fixed MakeLinks build script.
585 TK/01 Add preliminary DKIM support. Currently requires a forked version of
586 ALT-N's libdkim that I have put here:
587 http://duncanthrax.net/exim-experimental/
589 Note to Michael Haardt: I had to rename some vars in sieve.c. They
590 were called 'true' and it seems that C99 defines that as a reserved
591 keyword to be used with 'bool' variable types. That means you could
592 not include C99-style headers which use bools without triggering
593 build errors in sieve.c.
595 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
596 as mailq or other aliases. Changed the --help handling significantly
597 to do whats expected. exim_usage() emits usage/help information.
599 SC/01 Added the -bylocaldomain option to eximstats.
601 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
603 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
605 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
611 PH/01 Another patch from the Sieve maintainer.
613 PH/02 When an IPv6 address is converted to a string for single-key lookup
614 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
615 dots are used instead of colons so that keys in lsearch files need not
616 contain colons. This was done some time before quoting was made available
617 in lsearch files. However, iplsearch files do require colons in IPv6 keys
618 (notated using the quote facility) so as to distinguish them from IPv4
619 keys. This meant that lookups for IP addresses in host lists did not work
620 for iplsearch lookups.
622 This has been fixed by arranging for IPv6 addresses to be expressed with
623 colons if the lookup type is iplsearch. This is not incompatible, because
624 previously such lookups could never work.
626 The situation is now rather anomolous, since one *can* have colons in
627 ordinary lsearch keys. However, making the change in all cases is
628 incompatible and would probably break a number of configurations.
630 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
633 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
634 conversion specification without a maximum field width, thereby enabling
635 a rogue spamd server to cause a buffer overflow. While nobody in their
636 right mind would setup Exim to query an untrusted spamd server, an
637 attacker that gains access to a server running spamd could potentially
638 exploit this vulnerability to run arbitrary code as the Exim user.
640 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
641 $primary_hostname instead of what libspf2 thinks the hosts name is.
643 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
644 a directory entry by the name of the lookup key. Previously, if a
645 symlink pointed to a non-existing file or a file in a directory that
646 Exim lacked permissions to read, a lookup for a key matching that
647 symlink would fail. Now it is enough that a matching directory entry
648 exists, symlink or not. (Bugzilla 503.)
650 PH/03 The body_linecount and body_zerocount variables are now exported in the
653 PH/04 Added the $dnslist_matched variable.
655 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
656 This means they are set thereafter only if the connection becomes
659 PH/06 Added the client_condition to authenticators so that some can be skipped
660 by clients under certain conditions.
662 PH/07 The error message for a badly-placed control=no_multiline_responses left
663 "_responses" off the end of the name.
665 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
667 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
668 (without spaces) instead of just copying the configuration text.
670 PH/10 Added the /noupdate option to the ratelimit ACL condition.
672 PH/11 Added $max_received_linelength.
674 PH/12 Added +ignore_defer and +include_defer to host lists.
676 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
677 way in which PCRE > 7.0 is built.
679 PH/14 Implemented queue_only_load_latch.
681 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
682 MAIL command. The effect was to mangle the value on 64-bit systems.
684 PH/16 Another patch from the Sieve maintainer.
686 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
688 PH/18 If a system quota error occurred while trying to create the file for
689 a maildir delivery, the message "Mailbox is full" was not appended to the
690 bounce if the delivery eventually timed out. Change 4.67/27 below applied
691 only to a quota excession during the actual writing of the file.
693 PH/19 It seems that peer DN values may contain newlines (and other non-printing
694 characters?) which causes problems in log lines. The DN values are now
695 passed through string_printing() before being added to log lines.
697 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
698 and InterBase are left for another time.)
700 PH/21 Added message_body_newlines option.
702 PH/22 Guard against possible overflow in moan_check_errorcopy().
704 PH/23 POSIX allows open() to be a macro; guard against that.
706 PH/24 If the recipient of an error message contained an @ in the local part
707 (suitably quoted, of course), incorrect values were put in $domain and
708 $local_part during the evaluation of errors_copy.
714 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
715 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
718 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
719 issue a MAIL command.
721 PH/02 In an ACL statement such as
723 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
725 if a client was not listed at all, or was listed with a value other than
726 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
727 the condition was not true (as it should be), so access was not denied.
728 The bug was that the ! inversion was incorrectly passed on to the second
729 item. This has been fixed.
731 PH/03 Added additional dnslists conditions == and =& which are different from
732 = and & when the dns lookup returns more than one IP address.
734 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
735 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
737 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
738 FSYNC, which compiles an option called disable_fsync that allows for
739 bypassing fsync(). The documentation is heavily laced with warnings.
741 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
743 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
744 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
745 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
746 including adding "make clean"; (3) Added -fPIC when compiling the test
747 dynamically loaded module, to get rid of a warning.
749 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
750 message fails, move_frozen_messages = true and ignore_bounce_errors_after
751 = 0s. The bug is otherwise harmless.
753 PH/07 There was a bug in the dovecot authenticator such that the value of
754 $auth1 could be overwritten, and so not correctly preserved, after a
755 successful authentication. This usually meant that the value preserved by
756 the server_setid option was incorrect.
758 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
760 PH/09 Installed PCRE release 7.0.
762 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
763 run for batched SMTP input. It is now run at the start of every message
764 in the batch. While fixing this I discovered that the process information
765 (output by running exiwhat) was not always getting set for -bs and -bS
766 input. This is fixed, and it now also says "batched" for BSMTP.
768 PH/11 Added control=no_pipelining.
770 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
771 patch, slightly modified), and move the expansion of helo_data till after
772 the connection is made in the smtp transport (so it can use these
775 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
777 PH/14 Added log_selector = +pid.
779 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
781 PH/16 Add ${if forany and ${if forall.
783 PH/17 Added dsn_from option to vary the From: line in DSNs.
785 PH/18 Flush SMTP output before performing a callout, unless control =
786 no_callout_flush is set.
788 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
789 was true (the default) a successful delivery failed to delete the retry
790 item, thus causing premature timeout of the address. The bug is now
793 PH/20 Added hosts_avoid_pipelining to the smtp transport.
795 PH/21 Long custom messages for fakedefer and fakereject are now split up
796 into multiline reponses in the same way that messages for "deny" and
797 other ACL rejections are.
799 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
800 with slight modification.
802 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
803 draft, changing the syntax and factoring some duplicate code".
805 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
806 for deliveries of the second and subsequent messages over the same SMTP
809 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
810 ${reduce, with only minor "tidies".
812 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
814 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
815 expansion side effects.
817 PH/27 When a message times out after an over-quota error from an Exim-imposed
818 quota, the bounce message says "mailbox is full". This message was not
819 being given when it was a system quota that was exceeded. It now should
822 MH/03 Made $recipients available in local_scan(). local_scan() already has
823 better access to the recipient list through recipients_list[], but
824 $recipients can be useful in postmaster-provided expansion strings.
826 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
827 in the case of a MAIL command with additional options following the
828 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
829 were accidentally chopped off.
831 PH/29 SMTP synchronization checks are implemented when a command is read -
832 there is a check that no more input is waiting when there shouldn't be
833 any. However, for some commands, a delay in an ACL can mean that it is
834 some time before the response is written. In this time, more input might
835 arrive, invalidly. So now there are extra checks after an ACL has run for
836 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
837 pipelining has not been advertised.
839 PH/30 MH's patch to allow iscntrl() characters to be list separators.
841 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
842 returned in the SMTP response when smtp_return_error_details was false.
845 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
846 instead of the C I/O that was originally supplied, because problems were
849 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
850 Exim which did not show up earlier: it was assuming that a call to
851 SSL_CTX_set_info_callback() might give an error value. In fact, there is
852 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
853 was a macro that became an assignment, so it seemed to work. This has
854 changed to a proper function call with a void return, hence the compile
855 error. Exim's code has been fixed.
857 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
860 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
862 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
864 PH/37 If a message is not accepted after it has had an id assigned (e.g.
865 because it turns out to be too big or there is a timeout) there is no
866 "Completed" line in the log. When some messages of this type were
867 selected by exigrep, they were listed as "not completed". Others were
868 picked up by some special patterns. I have improved the selection
869 criteria to be more general.
871 PH/38 The host_find_failed option in the manualroute router can now be set
872 to "ignore", to completely ignore a host whose IP address cannot be
873 found. If all hosts are ignored, the behaviour is controlled by the new
874 host_all_ignored option.
876 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
877 homing or because of multiple MX records with /mx) generated more than
878 one IP address, and the following item turned out to be the local host,
879 all the secondary addresses of the first item were incorrectly removed
880 from the list, along with the local host and any following hosts (which
881 is what is supposed to happen).
883 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
884 whoever called Exim into the -H file. In the case of the daemon it was
885 behaving confusingly. When first started, it used values for whoever
886 started the daemon, but after a SIGHUP it used the Exim user (because it
887 calls itself on a restart). I have changed the code so that it now always
890 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
891 message are rejected with the same error (e.g. no authentication or bad
892 sender address), and a DATA command is nevertheless sent (as can happen
893 with PIPELINING or a stupid MUA), the error message that was given to the
894 RCPT commands is included in the rejection of the DATA command. This is
895 intended to be helpful for MUAs that show only the final error to their
898 PH/42 Another patch from the Sieve maintainer.
900 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
901 Eximstats - Fixed some broken HTML links and added missing column headers
903 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
904 columns for Rejects, Temp Rejects, Ham, and Spam rows.
906 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
908 PH/43 Yet another patch from the Sieve maintainer.
910 PH/44 I found a way to check for a TCP/IP connection going away before sending
911 the response to the final '.' that terminates a message, but only in the
912 case where the client has not sent further data following the '.'
913 (unfortunately, this is allowed). However, in many cases there won't be
914 any further data because there won't be any more messages to send. A call
915 to select() can be used: if it shows that the input is "ready", there is
916 either input waiting, or the socket has been closed. An attempt to read
917 the next input character can distinguish the two cases. Previously, Exim
918 would have sent an OK response which the client would never have see.
919 This could lead to message repetition. This fix should cure that, at
920 least in a lot of common cases.
922 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
923 advertised in response to EHLO.
929 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
930 fixed by 4.65/MH/01 (is this a record?) are fixed:
932 (i) An empty string was always treated as zero by the numeric comparison
933 operators. This behaviour has been restored.
935 (ii) It is documented that the numeric comparison operators always treat
936 their arguments as decimal numbers. This was broken in that numbers
937 starting with 0 were being interpreted as octal.
939 While fixing these problems I realized that there was another issue that
940 hadn't been noticed. Values of message_size_limit (both the global option
941 and the transport option) were treated as octal if they started with 0.
942 The documentation was vague. These values are now always treated as
943 decimal, and I will make that clear in the documentation.
949 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
950 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
953 MH/01 Don't check that the operands of numeric comparison operators are
954 integers when their expansion is in "skipping" mode (fixes bug
955 introduced by 4.64-PH/07).
957 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
958 child addresses, Exim now panics and dies. Previously, because the count
959 is held in a short int, deliveries were likely to be lost. As such a
960 large number of recipients for a single message is ridiculous
961 (performance will be very, very poor), I have chosen to impose a limit
962 rather than extend the field.
968 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
969 leftover -K file (the existence of which was triggered by #402).
970 While we were at it, introduced process PID as part of the -K
971 filename. This should rule out race conditions when creating
974 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
975 processing considerably. Previous code took too long for large mails,
976 triggering a timeout which in turn triggers #401.
978 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
979 in the DK code in transports.c. sendfile() is not really portable,
980 hence the _LINUX specificness.
982 TF/01 In the add_headers option to the mail command in an Exim filter,
983 there was a bug that Exim would claim a syntax error in any
984 header after the first one which had an odd number of characters
987 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
988 callout verification, Exim cached a "reject" for the entire domain. This
989 is correct for most verifications, but it is not correct for a recipient
990 verification with use_sender or use_postmaster set, because in that case
991 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
992 case of MAIL FROM:<> rejection from other early rejections (e.g.
993 rejection of HELO). When verifying a recipient using a non-null MAIL
994 address, the cache is ignored if it shows MAIL FROM:<> rejection.
995 Whatever the result of the callout, the value of the domain cache is
996 left unchanged (for any other kind of callout, getting as far as trying
997 RCPT means that the domain itself is ok).
999 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1002 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1003 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1004 session, the daemon ignores ECONNECT errors and logs others; it now
1005 ignores EPIPE as well.
1007 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1008 (quoted-printable decoding).
1010 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1011 later the small subsequent patch to fix an introduced bug.
1013 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1015 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1017 PH/08 An error is now given if message_size_limit is specified negative.
1019 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1020 to be given (somewhat) arbitrary names.
1022 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1025 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1026 miscellaneous code fixes
1028 PH/10 Added the log_reject_target ACL modifier to specify where to log
1031 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1032 hostname. This is wrong, because it relates to the incoming message (and
1033 probably the interface on which it is arriving) and not to the outgoing
1034 callout (which could be using a different interface). This has been
1035 changed to use the value of the helo_data option from the smtp transport
1036 instead - this is what is used when a message is actually being sent. If
1037 there is no remote transport (possible with a router that sets up host
1038 addresses), $smtp_active_hostname is used.
1040 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1041 tweaks were necessary in order to get it to work (see also 21 below):
1042 (a) The code assumed that strncpy() returns a negative number on buffer
1043 overflow, which isn't the case. Replaced with Exim's string_format()
1045 (b) There were several signed/unsigned issues. I just did the minimum
1046 hacking in of casts. There is scope for a larger refactoring.
1047 (c) The code used strcasecmp() which is not a standard C function.
1048 Replaced with Exim's strcmpic() function.
1049 (d) The code set only $1; it now sets $auth1 as well.
1050 (e) A simple test gave the error "authentication client didn't specify
1051 service in request". It would seem that Dovecot has changed its
1052 interface. Fortunately there's a specification; I followed it and
1053 changed what the client sends and it appears to be working now.
1055 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1058 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1059 address A is aliased to B and C, where B exists and C does not. Without
1060 -v the output is "A verified" because verification stops after a
1061 successful redirection if more than one address is generated. However,
1062 with -v the child addresses are also verified. Exim was outputting "A
1063 failed to verify" and then showing the successful verification for C,
1064 with its parentage. It now outputs "B failed to verify", showing B's
1065 parentage before showing the successful verification of C.
1067 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1068 look up a TXT record in a specific list after matching in a combined
1071 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1072 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1073 they consult the DNS. I had assumed they would set it the way they
1074 wanted; and indeed my experiments on Linux seem to show that in some
1075 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1076 To be on the safe side, however, I have now made the interface to
1077 host_find_byname() similar to host_find_bydns(), with an argument
1078 containing the DNS resolver options. The host_find_byname() function now
1079 sets these options at its start, just as host_find_bydns() does. The smtp
1080 transport options dns_qualify_single and dns_search_parents are passed to
1081 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1082 of host_find_byname() use the default settings of RES_DEFNAMES
1083 (qualify_single) but not RES_DNSRCH (search_parents).
1085 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1086 spool_read_header() do less string testing, by means of a preliminary
1087 switch on the second character of optional "-foo" lines. (This is
1088 overdue, caused by the large number of possibilities that now exist.
1089 Originally there were few.) While I was there, I also converted the
1090 str(n)cmp tests so they don't re-test the leading "-" and the first
1091 character, in the hope this might squeeze out yet more improvement.
1093 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1094 flag allowing group syntax was set by the header_syntax check but not
1095 turned off, possible causing trouble later; (2) The flag was not being
1096 set at all for the header_verify test, causing "group"-style headers to
1097 be rejected. I have now set it in this case, and also caused header_
1098 verify to ignore an empty address taken from a group. While doing this, I
1099 came across some other cases where the code for allowing group syntax
1100 while scanning a header line wasn't quite right (mostly, not resetting
1101 the flag correctly in the right place). These bugs could have caused
1102 trouble for malformed header lines. I hope it is now all correct.
1104 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1105 with the "reply" argument non-NULL. The code, however (which originally
1106 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1107 but it didn't always do it. This confused somebody who was copying the
1108 code for some other use. I have removed all the tests.
1110 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1111 feature that was used to support insecure browsers during the U.S. crypto
1112 embargo. It requires special client support, and Exim is probably the
1113 only MTA that supported it -- and would never use it because real RSA is
1114 always available. This code has been removed, because it had the bad
1115 effect of slowing Exim down by computing (never used) parameters for the
1116 RSA_EXPORT functionality.
1118 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1119 authenticator to fail if there's a tab character in the incoming data
1120 (there should never be unless someone is messing about, as it's supposed
1121 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1122 if the connection is using TLS or if the remote IP is the same as the
1123 local IP, and the "valid-client-cert option" if a client certificate has
1126 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1127 authenticators. This can be used for authorization after authentication
1128 succeeds. (In the case of plaintext, it servers for both authentication
1131 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1132 if any retry times were supplied.
1134 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1135 connection, where there is no client IP address to check. In this
1136 situation, the verify now always succeeds.
1138 PH/25 Applied John Jetmore's -Mset patch.
1140 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1142 PH/27 In a string expansion for a processed (not raw) header when multiple
1143 headers of the same name were present, leading whitespace was being
1144 removed from all of them, but trailing whitespace was being removed only
1145 from the last one. Now trailing whitespace is removed from each header
1146 before concatenation. Completely empty headers in a concatenation (as
1147 before) are ignored.
1149 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1150 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1152 PH/29 [Removed. This was a change that I later backed out, and forgot to
1153 correct the ChangeLog entry (that I had efficiently created) before
1154 committing the later change.]
1156 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1157 address errors (4xx response to RCPT) over the same connection as other
1158 messages routed to the same hosts. Such deliveries are always "forced",
1159 so retry times are not inspected. This resulted in far too many retries
1160 for the affected addresses. The effect occurred only when there were more
1161 hosts than the hosts_max_try setting in the smtp transport when it had
1162 the 4xx errors. Those hosts that it had tried were not added to the list
1163 of hosts for which the message was waiting, so if all were tried, there
1164 was no problem. Two fixes have been applied:
1166 (i) If there are any address or message errors in an SMTP delivery, none
1167 of the hosts (tried or untried) are now added to the list of hosts
1168 for which the message is waiting, so the message should not be a
1169 candidate for sending over the same connection that was used for a
1170 successful delivery of some other message. This seems entirely
1171 reasonable: after all the message is NOT "waiting for some host".
1172 This is so "obvious" that I'm not sure why it wasn't done
1173 previously. Hope I haven't missed anything, but it can't do any
1174 harm, as the worst effect is to miss an optimization.
1176 (ii) If, despite (i), such a delivery is accidentally attempted, the
1177 routing retry time is respected, so at least it doesn't keep
1178 hammering the server.
1180 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1181 in ${readsocket because some servers need this prod.
1183 PH/32 Added some extra debug output when updating a wait-xxx database.
1185 PH/33 The hint "could be header name not terminated by colon", which has been
1186 given for certain expansion errors for a long time, was not being given
1187 for the ${if def:h_colon_omitted{... case.
1189 PH/34 The spec says: "With one important exception, whenever a domain list is
1190 being scanned, $domain contains the subject domain." There was at least
1191 one case where this was not true.
1193 PH/35 The error "getsockname() failed: connection reset by peer" was being
1194 written to the panic log as well as the main log, but it isn't really
1195 panic-worthy as it just means the connection died rather early on. I have
1196 removed the panic log writing for the ECONNRESET error when getsockname()
1199 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1200 runs only) independently of the message's sender address. This meant
1201 that, if the 4xx error was in fact related to the sender, a different
1202 message to the same recipient with a different sender could confuse
1203 things. In particualar, this can happen when sending to a greylisting
1204 server, but other circumstances could also provoke similar problems.
1205 I have changed the default so that the retry time for these errors is now
1206 based a combination of the sender and recipient addresses. This change
1207 can be overridden by setting address_retry_include_sender=false in the
1210 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1211 remote server are returned as part of bounce messages. This was not
1212 happening for LMTP over a pipe (the lmtp transport), but now it is the
1213 same for both kinds of LMTP.
1215 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1216 in header lines that were in fact CNAMEs. This is no longer the case.
1218 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1219 and queue runs started by the daemon processed all messages. This has
1220 been fixed so that -R and -S can now usefully be given with -q<time>.
1222 PH/40 Import PCRE release 6.7 (fixes some bugs).
1224 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1226 PH/42 Give an error if -q is specified more than once.
1228 PH/43 Renamed the variables $interface_address and $interface_port as
1229 $received_ip_address and $received_port, to make it clear that these
1230 values apply to message reception, and not to the outgoing interface when
1231 a message is delivered. (The old names remain recognized, of course.)
1233 PH/44 There was no timeout on the connect() call when using a Unix domain
1234 socket in the ${readsocket expansion. There now is.
1236 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1237 be meaningful with "accept".
1239 SC/01 Eximstats V1.43
1240 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1242 SC/02 Eximstats V1.44
1243 Use a glob alias rather than an array ref in the generated
1244 parser. This improves both readability and performance.
1246 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1247 Collect SpamAssassin and rejection statistics.
1248 Don't display local sender or destination tables unless
1249 there is data to show.
1250 Added average volumes into the top table text output.
1252 SC/04 Eximstats V1.46
1253 Collect data on the number of addresses (recipients)
1254 as well as the number of messages.
1256 SC/05 Eximstats V1.47
1257 Added 'Message too big' to the list of mail rejection
1258 reasons (thanks to Marco Gaiarin).
1260 SC/06 Eximstats V1.48
1261 Mainlog lines which have GMT offsets and are too short to
1262 have a flag are now skipped.
1264 SC/07 Eximstats V1.49 (Alain Williams)
1265 Added the -emptyok flag.
1267 SC/08 Eximstats V1.50
1268 Fixes for obtaining the IP address from reject messages.
1270 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1271 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1272 whitesspace changes from 4.64-PH/27
1274 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1277 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1278 are found, allow negative numbers in numeric criteria)
1280 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1282 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1283 to match changes made in 4.64-PH/43
1285 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1287 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1288 "domain" is actually an IP address, and force a failure. This locks out
1289 those revolvers/nameservers that support "A-for-A" lookups, in
1290 contravention of the specifications.
1292 PH/48 When a host name was looked up from an IP address, and the subsequent
1293 forward lookup of the name timed out, the host name was left in
1294 $sender_host_name, contrary to the specification.
1296 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1297 restricted to single-key lookups, Exim was not diagnosing an error if
1298 * or *@ was used with a query-style lookup.
1300 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1302 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1303 long ago, but noone interested enough thought of it. Let's just say that
1304 the "1.1" means that there are some new functions that weren't there at
1305 some point in the past.
1307 PH/51 Error processing for expansion failure of helo_data from an smtp
1308 transport during callout processing was broken.
1310 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1311 tested/used via the -bh/-bhc/-bs options.
1313 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1314 bug, fixed in subsequent PCRE releases).
1316 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1317 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1319 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1325 SC/01 Use a glob alias rather than an array ref in eximstats generated
1326 parser. This improves both readability and performance.
1328 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1329 Don't display local sender or destination tables in eximstats unless
1330 there is data to show.
1331 Added average volumes into the eximstats top table text output.
1333 SC/03 Collect data on the number of addresses (recipients) as well
1334 as the number of messages in eximstats.
1336 TF/01 Correct an error in the documentation for the redirect router. Exim
1337 does (usually) call initgroups() when daemonizing.
1339 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1340 with consistent privilege compared to when running as a daemon.
1342 TF/03 Note in the spec that $authenticated_id is not set for local
1343 submissions from trusted users.
1345 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1346 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1348 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1349 by adding some example configuration directives to the default
1350 configuration file. A little bit of work is required to uncomment the
1351 directives and define how usernames and passwords are checked, but
1352 there is now a framework to start from.
1354 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1355 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1356 without this. I don't know how relevant this is to other LDAP libraries.
1358 PH/02 Add the verb name to the "unknown ACL verb" error.
1360 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1362 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1364 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1365 directory not expanded when it should be if an expanded home directory
1366 was set for the address (which is overridden by the transport).
1368 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1371 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1372 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1373 because it is too late at that time, and has no effect.
1375 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1376 security issue with \' (bugzilla #107). I could not use the
1377 PQescapeStringConn() function, because it needs a PGconn value as one of
1380 PH/08 When testing addresses using -bt, indicate those final addresses that
1381 are duplicates that would not cause an additional delivery. At least one
1382 person was confused, thinking that -bt output corresponded to deliveries.
1383 (Suppressing duplicates isn't a good idea as you lose the information
1384 about possibly different redirections that led to the duplicates.)
1386 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1387 systems where poll() doesn't work, in particular OS X.
1389 PH/10 Added more information to debugging output for retry time not reached.
1391 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1392 operations in malware.c.
1394 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1397 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1398 syslog_duplication set false, log lines that would normally be written
1399 both the the main log and to the reject log were not written to syslog at
1402 PH/14 In the default configuration, change the use of "message" in ACL warn
1403 statements to "add_header".
1405 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1406 not followed by a command (e.g. "seen endif").
1408 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1409 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1412 PH/17 Added extra conditions to the default value of delay_warning_condition
1416 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1417 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1418 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1421 The Auto-Submitted: and various List- headers are standardised, whereas I
1422 don't think Precedence: ever was.
1424 PH/18 Refactored debugging code in route_finduser() to show more information,
1425 in particular, the error code if getpwnam() issues one.
1427 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1428 This is apparently needed in addition to the PH/07 change above to avoid
1429 any possible encoding problems.
1431 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1432 but not after initializing Perl.
1434 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1435 output them only if debugging. By default they are written stderr,
1436 apparently, which is not desirable.
1438 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1441 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1444 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1446 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1447 authenticated or an ident call has been made. Suppress the default
1448 values for $authenticated_id and $authenticated_sender (but permit -oMai
1449 and -oMas) when testing with -bh.
1451 PH/24 Re-jigged the order of the tests in the default configuration so that the
1452 tests for valid domains and recipients precede the DNS black list and CSA
1453 tests, on the grounds that those ones are more expensive.
1455 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1456 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1457 This bug exists in every version of Exim that I still have, right back to
1460 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1461 However, an attempt to turn on case-sensitivity in a regex key by
1462 including (?-i) didn't work because the subject string was already
1463 lowercased, and the effects were non-intuitive. It turns out that a
1464 one-line patch can be used to allow (?-i) to work as expected.
1470 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1471 other effects) broke the use of negated acl sub-conditions.
1473 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1476 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1477 "Deny" causes Exim to reject the incoming connection with a 554 error.
1478 Unfortunately, if there is a major crisis, such as a disk failure,
1479 tcp-wrappers gives "deny", whereas what one would like would be some
1480 kind of temporary error. A kludge has been added to help with this.
1481 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1482 554 error is used if errno is still zero or contains ENOENT (which occurs
1483 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1486 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1488 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1489 errors. Otherwise a message that provokes a temporary error (when other
1490 messages do not) can cause a whole host to time out.
1492 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1493 addresses were routed directly to files or pipes from a redirect router.
1494 File deliveries just didn't batch; pipe deliveries might have suffered
1497 PH/06 A failure to get a lock for a hints database would erroneously always say
1498 "Failed to get write lock", even when it was really a read lock.
1500 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1501 of 0600. This has been changed to use the value of the lockfile_mode
1502 option (which defaults to 0600).
1504 PH/08 Applied small patch from the Sieve maintainer.
1506 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1507 folder from quota calculations, a direct delivery into this folder messed
1508 up the contents of the maildirsize file. This was because the regex was
1509 used only to exclude .Trash (or whatever) when the size of the mailbox
1510 was calculated. There was no check that a delivery was happening into an
1511 excluded directory. This bug has been fixed by ignoring all quota
1512 processing for deliveries into excluded directories.
1514 PH/10 Added the maildirfolder_create_regex option to appendfile.
1520 PH/01 The code for finding all the local interface addresses on a FreeBSD
1521 system running IPv6 was broken. This may well have applied to all BSD
1522 systems, as well as to others that have similar system calls. The broken
1523 code found IPv4 interfaces correctly, but gave incorrect values for the
1524 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1525 that it would not match correctly against @[] and not recognize the IPv6
1528 PH/02 The ipliteral router was not recognizing addresses of the form user@
1529 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1531 PH/03 Added disable_ipv6.
1533 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1534 decoded form, because it is most often used to construct To: headers
1535 lines in autoreplies, and the decoded form may well be syntactically
1536 invalid. However, $reply_address has leading white space removed, and all
1537 newlines turned into spaces so that the autoreply transport does not
1540 PH/05 If group was specified without a user on a router, and no group or user
1541 was specified on a transport, the group from the router was ignored.
1543 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1544 for visible compile-time settings that can be used to change these
1545 numbers, for those that want even more. Backwards compatibility with old
1546 spool files has been maintained. However, going back to a previous Exim
1547 release will lost any variables that are in spool files.
1549 PH/07 Two small changes when running in the test harness: increase delay when
1550 passing a TCP/IP connection to a new process, in case the original
1551 process has to generate a bounce, and remove special handling of
1552 127.0.0.2 (sic), which is no longer necessary.
1554 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1555 be the same on different OS.
1557 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1560 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1561 whether --show-vars was specified or not
1563 JJ/02 exipick: Added support for new ACL variable spool format introduced
1566 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1567 syntactically invalid From: or Reply-to: line, and a filter used this to
1568 generate an autoreply, and therefore failed to obtain an address for the
1569 autoreply, Exim could try to deliver to a non-existent relative file
1570 name, causing unrelated and misleading errors. What now happens is that
1571 it logs this as a hard delivery error, but does not attempt to create a
1574 PH/11 The exinext utility has a -C option for testing purposes, but although
1575 the given file was scanned by exinext itself; it wasn't being passed on
1576 when Exim was called.
1578 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1579 an end-of-file indication when reading a command response.
1581 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1582 compiled. In many other places in Exim, IPv6 addresses are always
1583 recognized, so I have changed this. It also means that IPv4 domain
1584 literals of the form [IPV4:n.n.n.n] are now always recognized.
1586 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1587 used if the router is not running as root, for example, when verifying at
1588 ACL time, or when using -bh. The debugging output from this situation was
1589 non-existent - all you got was a failure to exec. I have made two
1592 (a) Failures to set uid/gid, the current directory, or a process leader
1593 in a subprocess such as that created by queryprogram now generate
1594 suitable debugging ouput when -d is set.
1596 (b) The queryprogram router detects when it is not running as root,
1597 outputs suitable debugging information if -d is set, and then runs
1598 the subprocess without attempting to change uid/gid.
1600 PH/15 Minor change to Makefile for building test_host (undocumented testing
1603 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1604 additional section of a DNS packet that returns MX or SRV records.
1605 Instead, it always explicitly searches for A/AAAA records. This avoids
1606 major problems that occur when a DNS server includes only records of one
1607 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1608 fixed another bug: if SRV records were looked up and the corresponding
1609 address records were *not* found in the additional section, the port
1610 values from the SRV records were lost.
1612 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1613 using the correct key (the original address) when searching the retry
1614 rules in order to find which one to use for generating the retry hint.
1616 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1617 adding the default one. Similarly, if it contains a Reply-To: header, the
1618 errors_reply_to option, if set, is not used.
1620 PH/19 When calculating a retry time, Exim used to measure the "time since
1621 failure" by looking at the "first failed" field in the retry record. Now
1622 it does not use this if it is later than than the arrival time of the
1623 message. Instead it uses the arrival time. This makes for better
1624 behaviour in cases where some deliveries succeed, thus re-setting the
1625 "first failed" field. An example is a quota failure for a huge message
1626 when small messages continue to be delivered. Without this change, the
1627 "time since failure" will always be short, possible causing more frequent
1628 delivery attempts for the huge message than are intended.
1629 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1631 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1632 $1, $2, $3) because the numerical variables can be reset during some
1633 expansion items (e.g. "match"), thereby losing the authentication data.
1635 PH/21 Make -bV show the size of off_t variables so that the test suite can
1636 decide whether to run tests for quotas > 2G.
1638 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1639 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1640 filecount value is greater than 2G or if a quota value is greater than 2G
1641 on a system where the size of off_t is not greater than 4, a panic error
1644 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1645 never match. The debug and -bh output now contains an explicit error
1646 message indicating a malformed IPv4 address or mask.
1648 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1649 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1650 PH/23 above applies.
1652 PH/25 Do not write to syslog when running in the test harness. The only
1653 occasion when this arises is a failure to open the main or panic logs
1654 (for which there is an explicit test).
1656 PH/26 Added the /no_tell option to "control=freeze".
1658 PH/27 If a host name lookup failed very early in a connection, for example, if
1659 the IP address matched host_lookup and the reverse lookup yielded a name
1660 that did not have a forward lookup, an error message of the form "no IP
1661 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1662 could be logged. Now it outputs the IP address instead of "NULL".
1664 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1665 allows the sender and the authenticated sender to be set when
1666 submitting a message from within Exim. Since child_open_exim() is
1667 documented for local_scan(), the new function should be too.
1669 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1670 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1671 results in an empty string is now treated as unset.
1673 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1675 PH/31 Added sender_verify_fail logging option.
1677 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1678 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1679 tidied the source and removed it altogether.
1681 PH/33 When a queue run was abandoned because the load average was too high, a
1682 log line was always written; now it is written only if the queue_run log
1683 selector is set. In addition, the log line for abandonment now contains
1684 information about the queue run such as the pid. This is always present
1685 in "start" and "stop" lines but was omitted from the "abandon" line.
1687 PH/34 Omit spaces between a header name and the colon in the error message that
1688 is given when verify = headers_syntax fails (if there are lots of them,
1689 the message gets confusing).
1691 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1692 names, as there are now some PTR records that contain slashes. This check
1693 is only to protect against broken name servers that fall over on strange
1694 characters, so the fact that it applies to all lookups doesn't matter.
1696 PH/36 Now that the new test suite is complete, we can remove some of the
1697 special code in Exim that was needed for the old test suite. For example,
1698 sorting DNS records because real resolvers return them in an arbitrary
1699 order. The new test suite's fake resolver always returns records in the
1702 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1703 bounces) except when queue_only is set, to avoid logging races between
1704 the different processes.
1706 PH/38 Panic-die if .include specifies a non-absolute path.
1708 PH/39 A tweak to the "H" retry rule from its user.
1710 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1711 a label. They prevented compilation on older perls.
1713 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1714 a warning to be raised on newish perls.
1716 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1717 on queue. Changes to match documented behaviour of showing count of
1718 messages matching specified criteria.
1720 PH/40 Changed the default ident timeout from 30s to 5s.
1722 PH/41 Added support for the use of login_cap features, on those BSD systems
1723 that have them, for controlling the resources used by pipe deliveries.
1725 PH/42 The content-scanning code uses fopen() to create files in which to put
1726 message data. Previously it was not paying any attention to the mode of
1727 the files. Exim runs with umask(0) because the rest of the code creates
1728 files with open(), and sets the required mode explicitly. Thus, these
1729 files were ending up world-writeable. This was not a big issue, because,
1730 being within the spool directory, they were not world-accessible. I have
1731 created a function called modefopen, which takes an additional mode
1732 argument. It sets umask(777), creates the file, chmods it to the required
1733 mode, then resets the umask. All the relevant calls to fopen() in the
1734 content scanning code have been changed to use this function.
1736 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1737 to 24 hours. This avoids potential overflow problems when processing G
1738 and H retry rules. I suspect nobody ever tinkers with this value.
1740 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1742 PH/45 When the plaintext authenticator is running as a client, the server's
1743 challenges are checked to ensure they are valid base64 strings. By
1744 default, the authentication attempt is cancelled if an invalid string is
1745 received. Setting client_ignore_invalid_base64 true ignores these errors.
1746 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1747 they are received. Thus, the responses can be made to depend on the
1748 challenges. If an invalid string is ignored, an empty string is placed in
1751 PH/46 Messages that are created by the autoreply transport now contains a
1752 References: header, in accordance with RFCs 2822 and 3834.
1754 PH/47 Added authenticated_sender_force to the smtp transport.
1756 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1758 PH/49 Installed latest patch from the Sieve maintainer.
1760 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1761 was also set, the appendfile transport was unnecessarily scanning a
1762 directory of message files (e.g. for maildir delivery) to find the count
1763 of files (along with the size), even though it did not need this
1764 information. It now does the scan only if it needs to find either the
1765 size of the count of files.
1767 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1769 PH/52 Two bugs concerned with error handling when the smtp transport is
1772 (i) Exim was not creating retry information for temporary errors given
1773 for individual recipients after the DATA command when the smtp transport
1774 was used in LMTP mode. This meant that they could be retried too
1775 frequently, and not timed out correctly.
1777 (ii) Exim was setting the flag that allows error details to be returned
1778 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1779 recipients that were returned after the DATA command.
1781 PH/53 This is related to PH/52, but is more general: for any failing address,
1782 when detailed error information was permitted to be returned to the
1783 sender, but the error was temporary, then after the final timeout, only
1784 "retry timeout exceeded" was returned. Now it returns the full error as
1785 well as "retry timeout exceeded".
1787 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1788 do this, and (what is worse) MTAs that accept it.
1790 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1791 will now be deprecated.
1793 PH/56 New os.c-cygwin from the Cygwin maintainer.
1795 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1796 formats (previously only available in exim formats via -bpr, -bpru,
1797 and -bpra. Now also available in native and exiqgrep formats)
1799 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1800 with very large, slow to parse queues
1802 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1804 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1806 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1807 responses to authentication challenges, though it was showing the
1808 challenges; (ii) I've removed the CR characters from the debug output for
1811 PH/58 Allow for the insertion of a newline as well as a space when a string
1812 is turned into more than one encoded-word during RFC 2047 encoding. The
1813 Sieve code now uses this.
1815 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1816 data_4xx, lost_connection, tls_required.
1818 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1819 message was being sent as an SMTP response.
1821 PH/61 Add -l and -k options to exicyclog.
1823 PH/62 When verifying, if an address was redirected to one new address, so that
1824 verification continued, and the new address failed or deferred after
1825 having set something in $address_data, the value of $address_data was not
1826 passed back to the ACL. This was different to the case when no
1827 redirection occurred. The value is now passed back in both cases.
1829 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1830 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1831 use login_cap.h, so on its own it isn't the distinguishing feature. The
1832 new name refers directly to the setclassresources() function.
1834 PH/65 Added configuration files for NetBSD3.
1836 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1838 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1839 is preferred over IPv4.
1841 PH/68 The bounce_return_message and bounce_return_body options were not being
1842 honoured for bounces generated during the reception of non-SMTP messages.
1843 In particular, this applied to messages rejected by the ACL. This bug has
1844 been fixed. However, if bounce_return_message is true and bounce_return_
1845 body is false, the headers that are returned for a non-SMTP message
1846 include only those that have been read before the error was detected.
1847 (In the case of an ACL rejection, they have all been read.)
1849 PH/69 The HTML version of the specification is now built in a directory called
1850 spec_html instead of spec.html, because the latter looks like a path with
1851 a MIME-type, and this confuses some software.
1853 PH/70 Catch two compiler warnings in sieve.c.
1855 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1856 function verify_get_ident() calls ip_connect() to connect a socket, but
1857 if the "connect()" function timed out, ip_connect() used to close the
1858 socket. However, verify_get_ident() also closes the socket later, and in
1859 between Exim writes to the log, which may get opened at this point. When
1860 the socket was closed in ip_connect(), the log could get the same file
1861 descriptor number as the socket. This naturally causes chaos. The fix is
1862 not to close the socket in ip_connect(); the socket should be closed by
1863 the function that creates it. There was only one place in the code where
1864 this was missing, in the iplookup router, which I don't think anybody now
1865 uses, but I've fixed it anyway.
1867 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1868 well as to direct DNS lookups. Otherwise the handling of names in host
1869 lists is inconsistent and therefore confusing.
1875 PH/01 Two changes to the default runtime configuration:
1877 (1) Move the checks for relay_from_hosts and authenticated clients from
1878 after to before the (commented out) DNS black list checks.
1880 (2) Add control=submission to the relay_from_hosts and authenticated
1881 clients checks, on the grounds that messages accepted by these
1882 statements are most likely to be submissions.
1884 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1886 (1) Generate an error if the third argument for the ${prvs expansion is
1889 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1892 (3) Reset the variables that are obtained from the first argument of
1893 ${prvscheck and used in the second argument before leaving the code,
1894 because their memory is reclaimed, so using them afterwards may do
1897 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1898 one (it's much easier than Tom thought :-).
1900 (5) Because of (4), we can now allow for the use of $prvscheck_result
1901 inside the third argument.
1903 PH/03 For some reason, the default setting of PATH when running a command from
1904 a pipe transport was just "/usr/bin". I have changed it to
1907 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1908 anything to be listed in the output from -bV.
1910 PH/05 When a filter generated an autoreply, the entire To: header line was
1911 quoted in the delivery log line, like this:
1913 => >A.N.Other <ano@some.domain> <original@ddress> ...
1915 This has been changed so that it extracts the operative address. There
1916 may be more than one such address. If so, they are comma-separated, like
1919 => >ano@some.domain,ona@other.domain <original@ddress> ...
1921 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1922 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1923 not being looked up in the rDNS to get a host name, Exim was showing the
1924 IP address twice in Received: lines, even though the IP addresses were
1925 identical. For example:
1927 Received: from [1.2.3.4] (helo=[1.2.3.4])
1929 However, if the real host name was known, it was omitting the HELO data
1930 if it matched the actual IP address. This has been tidied up so that it
1931 doesn't show the same IP address twice.
1933 PH/07 When both +timestamp and +memory debugging was on, the value given by
1934 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1935 called by the debug printing, thereby overwriting the timestamp buffer.
1936 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1938 PH/08 When the original message was included in an autoreply transport, it
1939 always said "this is a copy of the message, including all the headers",
1940 even if body_only or headers_only was set. It now gives an appropriate
1943 PH/09 Applied a patch from the Sieve maintainer which:
1945 o fixes some comments
1946 o adds the (disabled) notify extension core
1947 o adds some debug output for the result of if/elsif tests
1948 o points to the current vacation draft in the documentation
1949 and documents the missing references header update
1953 o fixes a bug in processing the envelope test (when testing
1954 multiple envelope elements, the last element determinted the
1957 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1958 Electronic Mail") by including:
1960 Auto-submitted: auto-generated
1962 in the messages that it generates (bounce messages and others, such as
1963 warnings). In the case of bounce messages for non-SMTP mesages, there was
1964 also a typo: it was using "Auto_submitted" (underscore instead of
1965 hyphen). Since every message generated by Exim is necessarily in response
1966 to another message, thes have all been changed to:
1968 Auto-Submitted: auto-replied
1970 in accordance with these statements in the RFC:
1972 The auto-replied keyword:
1974 - SHOULD be used on messages sent in direct response to another
1975 message by an automatic process,
1977 - MUST NOT be used on manually-generated messages,
1979 - MAY be used on Delivery Status Notifications (DSNs) and Message
1980 Disposition Notifications (MDNs),
1982 - MUST NOT be used on messages generated by automatic or periodic
1983 processes, except for messages which are automatic responses to
1986 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1987 to the default Received: header definition.
1989 PH/12 Added log selector acl_warn_skipped (default on).
1991 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1992 variables because (a) they are in the wrong storage pool and (b) even if
1993 they were copied, it wouldn't work properly because of the caching.
1995 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1996 checking when decoding. Apparently there are clients that generate
1997 overlong encoded strings. Why am I not surprised?
1999 PH/15 If the first argument of "${if match_address" was not empty, but did not
2000 contain an "@" character, Exim crashed. Now it writes a panic log message
2001 and treats the condition as false.
2003 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2005 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2006 extension "envelope-auth". The code is finished and in agreement with
2007 other implementations, but there is no documentation so far and in fact,
2008 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2009 not changing the active code.
2011 Print executed "if" and "elsif" statements when debugging is used. This
2012 helps a great deal to understand what a filter does.
2014 Document more things not specified clearly in RFC3028. I had all this
2015 sorted out, when out of a sudden new issues came to my mind. Oops."
2017 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2020 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2021 canonical form (as documented). However, after a host name lookup from
2022 the IP address, check_host() was doing a simple string comparison with
2023 addresses acquired from the DNS when checking that the found name did
2024 have the original IP as one of its addresses. Since any found IPv6
2025 addresses are likely to be in abbreviated form, the comparison could
2026 fail. Luckily, there already exists a function for doing the comparison
2027 by converting both addresses to binary, so now that is used instead of
2028 the text comparison.
2030 PH/20 There was another similar case to PH/19, when a complete host name was
2031 given in a host list; looking up its IP address could give an abbreviated
2032 form, whereas the current host's name might or might not be abbreviated.
2033 The same fix has been applied.
2039 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2040 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2043 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2044 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2046 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2048 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2049 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2050 Also, for "Auto-Submitted", treat anything other than "no" as
2051 non-personal, in accordance with RFC 3834. (Previously it treated
2052 anything starting "auto-" as non-personal.)
2054 TF/01 The control=submission/name=... option had a problem with syntax
2055 errors if the name included a slash character. The /name= option
2056 now slurps the rest of the string, so it can include any characters
2057 but it must come last in the list of options (after /sender_retain
2060 PH/05 Some modifications to the interface to the fake nameserver for the new
2068 TK/01 Added the "success_on_redirect" address verification option. See
2069 NewStuff for rationale and an example.
2071 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2073 PH/02 Patch to exigrep to allow it to work on syslog lines.
2075 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2076 fread() to skip over the body file's header line, because in Cygwin the
2077 header line is locked and is inaccessible.
2079 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2080 co-exist for some time) to make it clear that it is the Exim ID that is
2081 referenced, not the Message-ID: header line.
2083 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2084 string_format() function, because snprintf() does not exist on all
2087 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2088 ${stat: expansion item.
2090 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2091 protocol synchronization error", to keep the pedants happy.
2093 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2094 well as for IRIX systems, when gcc is being used. See the host.c source
2097 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2099 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2102 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2103 transport and to the smtp transport in LMTP mode.
2105 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2107 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2108 run a filter in a subprocess. This could lead to confusion in subsequent
2109 lookups in the parent process. There should also be a search_tidyup() at
2110 the end of the subprocess.
2112 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2113 only if the host matched helo_try_verify_hosts, which caused the
2114 verification to occur when the EHLO/HELO command was issued. The ACL just
2115 tested the remembered result. Now, if a previous verification attempt has
2116 not happened, "verify = helo" does it there and then.
2118 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2120 TK/03 Fix log output including CR from clamd.
2122 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2123 exist provoked a memory error which could cause a segfault.
2125 PH/15 Installed PCRE 6.2
2127 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2129 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2130 of the problem. Specifically, suggested +O2 rather than +O1 for the
2133 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2135 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2136 record, the port was not correctly logged when the outgoing_port log
2137 selector was set (it logged the transort's default port).
2139 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2140 fallback_hosts, and "hosts" in the smtp transport.
2142 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2143 host errors such as "Connection refused".
2145 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2146 authentication with radiusclient 0.4.9:
2148 - Error returned from rc_read_config was caught wrongly
2149 - Username/password not passed on to radius server due to wrong length.
2151 The presumption is that some radiusclient API changes for 4.51/PH/17
2152 were not taken care of correctly. The code is still untested by me (my
2153 Linux distribution still has 0.3.2 of radiusclient), but it was
2154 contributed by a Radius user.
2156 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2157 expanding the "port" option of the smtp transport.
2159 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2160 while reading a MIME header. Thanks to Tom Hughes for a patch.
2162 PH/24 Include config.h inside local_scan.h so that configuration settings are
2165 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2166 that in an ACL for RCPT (for example), you can examine exactly what was
2169 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2170 commands, but it was not correctly comparing the address with the actual
2171 client host address. Thus, it would show the EHLO address in Received:
2172 header lines when this was not necessary.
2174 PH/27 Added the % operator to ${eval:}.
2176 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2177 it should be ignoring failures (because with -C, for example, it has lost
2178 privilege). It wasn't ignoring creation failures other than "already
2181 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2184 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2185 because an input line was too long, either on its own, or by virtue of
2186 too many continuations, the temporary file was not being removed, and the
2187 return code was incorrect.
2189 PH/31 Missing "BOOL" in function definition in filtertest.c.
2191 PH/32 Applied Sieve patches from the maintainer.
2193 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2195 PH/33 Added "verify = not_blind".
2197 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2198 Local/Makefile (with some defaults set). These are used in built scripts
2199 such as exicyclog, but they have never been used in the exim_install
2200 script (though there are many overriding facilities there). I have
2201 arranged that the exim_install script now takes note of these two
2204 PH/35 Installed configuration files for Dragonfly.
2206 PH/36 When a locally submitted message by a trusted user did not contain a
2207 From: header, and the sender address was obtained from -f or from an SMTP
2208 MAIL command, and the trusted user did not use -F to supply a sender
2209 name, $originator_name was incorrectly used when constructing a From:
2210 header. Furthermore, $originator_name was used for submission mode
2211 messages from external hosts without From: headers in a similar way,
2212 which is clearly wrong.
2214 PH/37 Added control=suppress_local_fixups.
2216 PH/38 When log_selector = +received_sender was set, and the addition of the
2217 sender made the log line's construction buffer exactly full, or one byte
2218 less than full, an overflow happened when the terminating "\n" was
2221 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2222 when the result of a list match is failure because a DNS lookup failed.
2224 PH/40 RM_COMMAND is now used in the building process.
2226 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2227 the "build-* directories that it finds.
2229 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2230 address in a domain literal was a prefix of an interface address.
2232 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2233 when verifying a sender address, unless rewrite_headers is false.
2235 PH/44 Wrote a long comment about why errors_to addresses are verified as
2236 recipients, not senders.
2238 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2239 the ratelimit ACL was added.
2241 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2243 PH/46 Added extra information about PostgreSQL errors to the error string.
2245 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2246 avoiding the need to install special zones in a real server. This is
2247 backwards compatible; if it can't find the fake resolver, it drops back.
2248 Thus, both old and new test suites can be run.
2250 TF/02 Added util/ratelimit.pl
2252 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2253 clock is set back in time.
2255 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2256 Candler <B.Candler@pobox.com>.
2258 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2259 OK for addresses that are the result of redirections.
2261 PH/48 A number of further additions for the benefit of the new test suite,
2262 including a fake gethostbyname() that interfaces to the fake DNS resolver
2265 TF/06 The fix for widen_domains has also been applied to qualify_single and
2266 search_parents which are the other dnslookup options that can cause
2269 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2272 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2274 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2275 service" if it encountered an MX record with an empty target hostname.
2276 The message is now "an MX or SRV record indicated no SMTP service".
2278 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2279 if the DNS of the sending site is misconfigured. This is quite a
2280 common situation. This change restores the behaviour of treating a
2281 helo verification defer as a failure.
2283 PH/51 If self=fail was set on a router, the bounce message did not include the
2284 actual error message.
2290 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2292 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2293 command itself ended in error, the underlying message about the transport
2294 filter timeout was being overwritten with the pipe command error. Now the
2295 underlying error message should be appended to the second error message.
2297 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2299 PH/02 Reduce the amount of output that "make" produces by default. Full output
2300 can still be requested.
2302 PH/03 The warning log line about a condition test deferring for a "warn" verb
2303 was being output only once per connection, rather than after each
2304 occurrence (because it was using the same function as for successful
2305 "warn" verbs). This seems wrong, so I have changed it.
2307 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2308 it should not have, which might have caused a crash in the right
2309 circumstances, but probably never did.
2311 PH/04 Installed a modified version of Tony Finch's patch to make submission
2312 mode fix the return path as well as the Sender: header line, and to
2313 add a /name= option so that you can make the user's friendly name appear
2316 TF/03 Added the control = fakedefer ACL modifier.
2318 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2319 Mark Lowes for thorough testing.
2321 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2323 TK/03 Merged latest SRS patch from Miles Wilton.
2325 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2326 with the definition in sysexits.h (which is #included earlier).
2327 Fortunately, Exim does not actually use EX_OK. The code used to try to
2328 preserve the sysexits.h value, by assumimg that macro definitions were
2329 scanned for macro replacements. I have been disabused of this notion,
2330 so now the code just undefines EX_OK before #including unistd.h.
2332 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2333 in the smtp transport. When a block could not be written in a single
2334 write() function, the timeout was being re-applied to each part-write.
2335 This seems wrong - if the receiver was accepting one byte at a time it
2336 would take for ever. The timeout is now adjusted when this happens. It
2337 doesn't have to be particularly precise.
2339 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2340 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2342 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2343 without a domain if the check to <postmaster@domain> fails.
2345 SC/01 Eximstats: added -xls and the ability to specify output files
2346 (patch written by Frank Heydlauf).
2348 SC/02 Eximstats: use FileHandles for outputing results.
2350 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2352 SC/04 Eximstats: fixed display of large numbers with -nvr option
2354 SC/05 Eximstats: fixed merging of reports with empty tables.
2356 SC/06 Eximstats: added the -include_original_destination flag
2358 SC/07 Eximstats: removed tabs and trailing whitespace.
2360 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2362 TK/06 MBOX spool code: Add real "From " MBOX separator line
2363 so the .eml file is really in mbox format (even though
2364 most programs do not really care). Patch from Alex Miller.
2366 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2367 The latter is generated from $received_to and is only set if the
2368 message has one envelope recipient. SA can use these headers,
2369 obviously out-of-the-box. Patch from Alex Miller.
2371 PH/08 The ${def test on a variable was returning false if the variable's
2372 value was "0", contrary to what the specification has always said!
2373 The result should be true unless the variable is empty.
2375 PH/09 The syntax error of a character other than { following "${if
2376 def:variable_name" (after optional whitespace) was not being diagnosed.
2377 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2378 accidental colon was present, for example, could give incorrect results.
2380 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2381 result is used (not including appendfile, where other changes are about
2384 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2385 This involved changing a lot of size variables from int to off_t. It
2386 should work with maildirs and everything.
2388 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2389 spamd dying while we are connected to it.
2391 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2394 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2395 function for BDB 4.3.
2397 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2399 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2400 the sources for PCRE have been re-arranged and more files are now
2403 PH/15 The code I had for printing potentially long long variables in PH/11
2404 above was not the best (it lost precision). The length of off_t variables
2405 is now inspected at build time, and an appropriate printing format (%ld
2406 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2407 to be "long long int" or "long int". This is needed for the internal
2408 formatting function string_vformat().
2410 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2411 the configuration file to be ":syslog", then the script "guesses" where
2412 the logs files are, rather than using the compiled in default. In our
2413 case the guess is not the same as the compiled default, so the script
2414 suddenly stopped working when I started to use syslog. The patch checks
2415 to see if log_file_path is "". If so, it attempts to read it from exim
2416 with no configuration file to get the compiled in version, before it
2417 falls back to the previous guessing code."
2419 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2420 implementing BATV in an Exim configuration. See NewStuff for the gory
2423 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2424 Makefile that are specific to HP-UX.
2426 PH/18 If the "use_postmaster" option was set for a recipient callout together
2427 with the "random" option, the postmaster address was used as the MAIL
2428 FROM address for the random test, but not for the subsequent recipient
2429 test. It is now used for both.
2431 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2432 patch removes a few documentation additions to RFC 3028, because the
2433 latest draft now contains them. It adds the new en;ascii-case comparator
2434 and a new error check for 8bit text in MIME parts. Comparator and
2435 require names are now matched exactly. I enabled the subaddress
2436 extension, but it is not well tested yet (read: it works for me)."
2438 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2439 rework some of the code of TK/09 above to avoid the hardwired use of
2440 "%lld" and "long long". Replaced the call to snprintf() with a call to
2443 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2444 records point to non-existent hosts", "retry timeout exceeded", and
2445 "retry time not reached for any host after a long failure period".
2447 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2448 experimental DomainKeys support:
2450 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2451 (2) On an error such as an illegally used "control", the wrong name for
2452 the control was given.
2454 These problems did NOT occur unless DomainKeys support was compiled.
2456 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2458 PH/24 Added ${if match_ip condition.
2460 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2461 databases so that it will be absolutely obvious if a crash occurs in the
2462 DB library. This is a regular occurrence (often caused by mis-matched
2465 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2466 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2467 picked up on a user's system that detects such things. There doesn't seem
2468 to be a gcc warning option for this - only an attribute that has to be
2469 put on the function's prototype. It seems that in Fedora Core 4 they have
2470 set this on a number of new functions. No doubt there will be more in due
2473 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2474 specify a transport. However, if an address that was verified by such a
2475 router was the subject of a callout, Exim crashed because it tried to
2476 read the rcpt_include_affixes from the non-existent transport. Now it
2477 just assumes that the setting of that option is false. This bug was
2478 introduced by 4.51/PH/31.
2480 PH/28 Changed -d+all to exclude +memory, because that information is very
2481 rarely of interest, but it makes the output a lot bigger. People tend to
2482 do -d+all out of habit.
2484 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2485 code in os-type was giving problems when libc.so lives in lib64, like on
2488 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2489 aren't the modern standard, and it seems that some systems' include files
2490 don't always have them. Exim was already checking for some of the newer
2491 ones like T_AAAA, and defining it itself. I've added checks for all the
2492 record types that Exim uses.
2494 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2495 not automatically generating a new one, as it is supposed to. This
2496 prevented TLS from working. If the file did exist, but contained invalid
2497 data, a new version was generated, as expected. It was only the case of a
2498 non-existent file that was broken.
2500 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2501 with a change in libdomainkeys > 0.64.
2503 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2504 from DNS. If the selector record carries the flag, it now has
2505 precedence over the domain-wide flag.
2507 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2509 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2510 the use of a port name that isn't defined in /etc/services) occurred, the
2511 message was deferred as in a normal delivery, and thus remained on the
2512 spool, instead of being failed because of the mua_wrapper setting. This
2513 is now fixed, and I tidied up some of the mua_wrapper messages at the
2516 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2517 the messages in a hash of arrays rather than using individual hashes.
2518 This is a bit cleaner and results in dramatic memory savings, albeit
2519 at a slight CPU cost.
2521 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2522 as requested by Marc Sherman.
2524 SC/10 Eximstats: added histograms for user specified patterns as requested
2527 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2529 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2530 fopen() in the content-scanning modules that did not already have it.
2536 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2537 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2539 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2541 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2543 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2544 to test Sieve filters that use "vacation".
2546 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2547 that changes the way the GnuTLS parameters are stored in the cache file.
2548 The new format can be generated externally. For backward compatibility,
2549 if the data in the cache doesn't make sense, Exim assumes it has read an
2550 old-format file, and it generates new data and writes a new file. This
2551 means that you can't go back to an older release without removing the
2554 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2555 work if there are any delivery delays because "one_time" forces the
2556 parent to be marked "delivered", so its unseen clone is never tried
2557 again. For this reason, Exim now forbids the simultaneous setting of
2560 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2561 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2562 entry if you want to know the details. The fix, however, neglected to
2563 consider the case when local delivery batching is involved. The test for
2564 "previously delivered" was not happening when checking to see if an
2565 address could be batched with a previous (undelivered) one; under
2566 certain circumstances this could lead to multiple deliveries to the same
2569 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2570 in its include files, and this causes problems building Exim.
2572 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2573 header_syntax) but Exim was just ignoring anything given after a slash.
2574 In particular, this caused confusion with an attempt to use "verify =
2575 reverse_host_lookup/defer_ok". An error is now given when options are
2576 supplied for verify items that do not have them. (Maybe reverse_host_
2577 lookup should have a defer_ok option, but that's a different point.)
2579 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2580 defined by RFC 821) to 2048, because there were problems with some AUTH
2581 commands, and RFC 1869 says the size should be increased for extended
2582 SMTP commands that take arguments.
2584 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2587 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2588 "unknown" error; now it says that the functionality isn't in the binary.
2590 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2591 an address' error message when a string expansion fails (syntax or
2592 whatever). Otherwise the password may appear in the log. Following change
2593 PH/42 below, there is no longer a chance of it appearing in a bounce
2596 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2598 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2599 of its addresses was ever tried. (Bugzilla bug #2.)
2601 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2602 the result incorrectly in the debug output. (It correctly added a newline
2603 to what was transported.)
2605 TF/01 Added $received_time.
2607 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2608 commented out examples of how to interface to a virus scanner and to
2609 SpamAssassin. Also added commented examples of av_scanner and
2610 spamd_address settings.
2612 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2613 and controls are allowed in which ACLs. There were a couple of minor
2614 errors. Some of the entries in the conditions table (which is a table of
2615 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2616 negation of where the condition IS allowed.
2618 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2620 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2621 header file does not have a version number, so I've had to invent a new
2622 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2623 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2624 radiusclient), but it was contributed by a Radius user.
2626 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2627 files or directories, for OpenSSL.
2629 PH/19 When an Exim process that is running as root has to create an Exim log
2630 file, it does so in a subprocess that runs as exim:exim so as to get the
2631 ownership right at creation (otherwise, other Exim processes might see
2632 the file with the wrong ownership). There was no test for failure of this
2633 fork() call, which would lead to the process getting stuck as it waited
2634 for a non-existent subprocess. Forks do occasionally fail when resources
2635 run out. I reviewed all the other calls to fork(); they all seem to check
2638 PH/20 When checking for unexpected SMTP input at connect time (before writing
2639 the banner), Exim was not dealing correctly with a non-positive return
2640 from the read() function. If the client had disconnected by this time,
2641 the result was a log entry for a synchronization error with an empty
2642 string after "input=" when read() returned zero. If read() returned -1
2643 (an event I could not check), uninitialized data bytes were printed.
2644 There were reports of junk text (parts of files, etc) appearing after
2647 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2649 PH/22 Added support for macro redefinition, and (re)definition in between
2650 driver and ACL definitions.
2652 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2653 forgetting to use the resulting value; it was using the unexpanded value.
2655 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2656 hadn't been configured. The fix is from Juergen Kreileder, who
2657 understands it better than I do:
2659 "Here's what I see happening with three configured cyrus_sasl
2660 authenticators configured (plain, login, cram-md5):
2662 On startup auth_cyrus_sasl_init() gets called for each of these.
2663 This means three calls to sasl_listmech() without a specified mech_list.
2664 => SASL tests which mechs of all available mechs actually work
2665 => three warnings about OTP not working
2666 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2668 With the patch, sasl_listmech() also gets called three times. But now
2669 SASL's mech_list option is set to the server_mech specified in the the
2670 authenticator. Or in other words, the answer from sasl_listmech()
2671 gets limited to just the mech you're testing for (which is different
2673 => the return list contains just 'plain' or 'login', 'cram-md5' or
2674 nothing depending on the value of ob->server_mech.
2676 I've just tested the patch: Authentication still works fine,
2677 unavailable mechs specified in the exim configuration are still
2678 caught, and the auth.log warnings about OTP are gone."
2680 PH/25 When debugging is enabled, the contents of the command line are added
2681 to the debugging output, even when log_selector=+arguments is not
2684 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2685 answer is "GNU", and only if the return is "GNU/something" is the answer
2688 PH/27 $acl_verify_message is now set immediately after the failure of a
2689 verification in an ACL, and so is available in subsequent modifiers. In
2690 particular, the message can be preserved by coding like this:
2692 warn !verify = sender
2693 set acl_m0 = $acl_verify_message
2695 Previously, $acl_verify_message was set only while expanding "message"
2696 and "log_message" when a very denied access.
2698 PH/28 Modified OS/os.c-Linux with
2700 -#ifndef OS_LOAD_AVERAGE
2701 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2703 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2704 nomenclature these days.)
2706 PH/29 Installed patch from the Sieve maintainer that adds the options
2707 sieve_useraddress and sieve_subaddress to the redirect router.
2709 PH/30 In these circumstances:
2710 . Two addresses routed to the same list of hosts;
2711 . First host does not offer TLS;
2712 . First host accepts first address;
2713 . First host gives temporary error to second address;
2714 . Second host offers TLS and a TLS session is established;
2715 . Second host accepts second address.
2716 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2717 and peerdn, if requested) that were in fact used only for the second
2720 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2721 attention to any local part prefix or suffix that was matched by the
2722 router that accepted the address. It now behaves in the same way as it
2723 does for delivery: the affixes are removed from the local part unless
2724 rcpt_include_affixes is set on the transport.
2726 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2727 timeout during the DATA phase of an incoming message.
2729 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2730 applied a patch sanctioned by the Sieve maintainer.
2732 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2733 the uid or gid is negative. A case of a negative gid caused this to be
2734 noticed. The fix allows for either to be negative.
2736 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2737 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2740 PH/36 The change PH/12 above was broken. Fixed it.
2742 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2743 the grounds that routing the same address twice would always produce the
2744 same answer. This might have been true once, but it is certainly no
2745 longer true now. Routing a child address may depend on the previous
2746 routing that produced that child. Some complicated redirection strategies
2747 went wrong when messages had multiple recipients, and made Exim's
2748 behaviour dependent on the order in which the addresses were given.
2750 I have moved the duplicate checking until after the routing is complete.
2751 Exim scans the addresses that are assigned to local and remote
2752 transports, and removes any duplicates. This means that more work will be
2753 done, as duplicates will always all be routed, but duplicates are
2754 presumably rare, so I don't expect this is of any significance.
2756 For deliveries to pipes, files, and autoreplies, the duplicate checking
2757 still happens during the routing process, since they are not going to be
2760 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2761 It corrects a timeout issue with spamd. This is Ian's comment: "The
2762 background is that sometimes spamd either never reads data from a
2763 connection it has accepted, or it never writes response data. The exiscan
2764 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2765 blindly assumes that writes won't block so it may never time out."
2767 PH/39 Allow G after quota size as well as K and M.
2769 PH/40 The value set for $authenticated_id in an authenticator may not contain
2770 binary zeroes or newlines because the value is written to log lines and
2771 to spool files. There was no check on this. Now the value is run through
2772 the string_printing() function so that such characters are converted to
2773 printable escape sequences.
2775 PH/41 $message_linecount is a new variable that contains the total number of
2776 lines in the message. Compare $body_linecount, which is the count for the
2779 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2780 bounce and delay warning messages, except in certain special cases, which
2783 (a) An SMTP error message from a remote host;
2784 (b) A message specified in a :fail: redirection;
2785 (c) A message specified in a "fail" command in a system filter;
2786 (d) A message specified in a FAIL return from the queryprogram router;
2787 (e) A message specified by the cannot_route_message router option.
2789 In these cases only, Exim does include the error details in bounce and
2790 warning messages. There are also a few cases where bland messages such
2791 as "unrouteable address" or "local delivery error" are given.
2793 PH/43 $value is now also set for the "else" part of a ${run expansion.
2795 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2796 being worked on, but at least Exim now implements the latest version to
2799 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2800 process to complete was treated as a delivery failure, a timeout while
2801 writing the message to the pipe was logged, but erroneously treated as a
2802 successful delivery. Such timeouts include transport filter timeouts. For
2803 consistency with the overall process timeout, these timeouts are now
2804 treated as errors, giving rise to delivery failures by default. However,
2805 there is now a new Boolean option for the pipe transport called
2806 timeout_defer, which, if set TRUE, converts the failures into defers for
2807 both kinds of timeout. A transport filter timeout is now identified in
2810 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2811 systems where "make" and "gmake" are different, calling "gmake" at top
2812 level broke things. I've arranged for the value of $(MAKE) to be passed
2813 from the Makefile to this script so that it can call the same version of
2817 A note about Exim versions 4.44 and 4.50
2818 ----------------------------------------
2820 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2821 changes of various kinds. As a consequence, a big documentation update was
2822 needed. This delayed the release for rather longer than seemed good, especially
2823 in the light of a couple of (minor) security issues. Therefore, the changes
2824 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2825 release. So 4.44 and 4.50 are in effect two different branches that both start
2828 I have left the 4.50 change log unchanged; it contains all the changes since
2829 4.43. The change log for 4.44 is below; many of its items are identical to
2830 those for 4.50. This seems to be the most sensible way to preserve the
2831 historical information.
2837 1. Minor wording change to the doc/README.SIEVE file.
2839 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2840 computation of the current number of files was incorrect.
2842 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2843 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2846 4. Give more explanation in the error message when the command for a transport
2847 filter fails to execute.
2849 5. There are several places where Exim runs a non-Exim command in a
2850 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2851 was being done only for the command run by the queryprogram router. It is
2852 now done for all such subprocesses. The other cases are: ${run, transport
2853 filters, and the commands run by the lmtp and pipe transports.
2855 6. Added CONFIGURE_GROUP build-time option.
2857 7. Some older OS have a limit of 256 on the maximum number of file
2858 descriptors. Exim was using setrlimit() to set 1000 as a large value
2859 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2860 systems. I've change it so that if it can't get 1000, it tries for 256.
2862 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2863 was an oversight, and furthermore, ever since the addition of extra
2864 controls (e.g. 4.43/32), the checks on when to allow different forms of
2865 "control" were broken. There should now be diagnostics for all cases when a
2866 control that does not make sense is encountered.
2868 9. Added the /retain_sender option to "control=submission".
2870 10. $recipients is now available in the predata ACL (oversight).
2872 11. Tidy the search cache before the fork to do a delivery from a message
2873 received from the command line. Otherwise the child will trigger a lookup
2874 failure and thereby defer the delivery if it tries to use (for example) a
2875 cached ldap connection that the parent has called unbind on.
2877 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2878 of $address_data from the recipient verification was clobbered by the
2879 sender verification.
2881 13. The value of address_data from a sender verification is now available in
2882 $sender_address_data in subsequent conditions in the ACL statement.
2884 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2886 15. Added a new option "connect=<time>" to callout options, to set a different
2889 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2890 was its contents. (It was OK if the option was not defined at all.)
2892 17. A "Completed" log line is now written for messages that are removed from
2893 the spool by the -Mrm option.
2895 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2896 information about exactly what failed.
2898 19. Added -dd to debug only the daemon process.
2900 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2901 handles timeouts, both on the server side and network timeouts. Renamed the
2902 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2904 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2905 It is now set to "smtps".
2907 22. $host_address is now set to the target address during the checking of
2908 ignore_target_hosts.
2910 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2911 being passed; this would have caused $sender_host_name to have been used if
2912 matching the list had actually called for a host name (not very likely,
2913 since this list is usually IP addresses). A host name is now passed as
2916 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2917 code that specifies a non-restarting handler (typically sigaction() in
2918 modern systems) in an attempt to fix a rare and obscure crash bug.
2920 25. Narrowed the window for a race in the daemon that could cause it to ignore
2921 SIGCHLD signals. This is not a major problem, because they are used only to
2922 wake it up if nothing else does.
2924 26. A malformed maildirsize file could cause Exim to calculate negative values
2925 for the mailbox size or file count. Odd effects could occur as a result.
2926 The maildirsize information is now recalculated if the size or filecount
2929 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2930 support for a long time. Removed HAVE_SYS_VFS_H.
2932 28. Installed the latest version of exipick from John Jetmore.
2934 29. In an address list, if the pattern was not a regular expression, an empty
2935 subject address (from a bounce message) matched only if the pattern was an
2936 empty string. Non-empty patterns were not even tested. This was the wrong
2937 because it is perfectly reasonable to use an empty address as part of a
2938 database query. An empty address is now tested by patterns that are
2939 lookups. However, all the other forms of pattern expect the subject to
2940 contain a local part and a domain, and therefore, for them, an empty
2941 address still always fails if the pattern is not itself empty.
2943 30. Exim went into a mad DNS loop when attempting to do a callout where the
2944 host was specified on an smtp transport, and looking it up yielded more
2945 than one IP address.
2947 31. Re-factored the code for checking spool and log partition space into a
2948 function that finds that data and another that does the check. The former
2949 is then used to implement four new variables: $spool_space, $log_space,
2950 $spool_inodes, and $log_inodes.
2952 32. The RFC2047 encoding function was originally intended for short strings
2953 such as real names; it was not keeping to the 75-character limit for
2954 encoded words that the RFC imposes. It now respects the limit, and
2955 generates multiple encoded words if necessary. To be on the safe side, I
2956 have increased the buffer size for the ${rfc2047: expansion operator from
2959 33. It is now permitted to omit both strings after an "if" condition; if the
2960 condition is true, the result is "true". As before, when the second string
2961 is omitted, a false condition yields an empty string. This makes it less
2962 cumbersome to write custom ACL and router conditions.
2964 34. Failure to deliver a bounce message always caused it to be frozen, even if
2965 there was an errors_to setting on the router. The errors_to setting is now
2968 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2969 canonical form (fully expanded) before being placed in
2970 $sender_host_address.
2972 36. The table in the code that translates DNS record types into text (T_A to
2973 "A" for instance) was missing entries for NS and CNAME. It is just possible
2974 that this could have caused confusion if both these types were looked up
2975 for the same domain, because the text type is used as part of Exim's
2976 per-process caching. But the chance of anyone hitting this buglet seems
2979 37. The dnsdb lookup has been extended in a number of ways.
2981 (1) There is a new type, "zns", which walks up the domain tree until it
2982 finds some nameserver records. It should be used with care.
2984 (2) There is a new type, "mxh", which is like "mx" except that it returns
2985 just the host names, not the priorities.
2987 (3) It is now possible to give a list of domains (or IP addresses) to be
2988 looked up. The behaviour when one of the lookups defers can be
2989 controlled by a keyword.
2991 (4) It is now possible to specify the separator character for use when
2992 multiple records are returned.
2994 38. The dnslists ACL condition has been extended: it is now possible to supply
2995 a list of IP addresses and/or domains to be looked up in a particular DNS
2998 39. Added log_selector=+queue_time_overall.
3000 40. When running the queue in the test harness, wait just a tad after forking a
3001 delivery process, to get repeatability of debugging output.
3003 41. Include certificate and key file names in error message when GnuTLS fails
3004 to set them up, because the GnuTLS error message doesn't include the name
3005 of the failing file when there is a problem reading it.
3007 42. Allow both -bf and -bF in the same test run.
3009 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3011 44. The "Exiscan patch" is now merged into the mainline Exim source.
3013 45. Sometimes the final signoff response after QUIT could fail to get
3014 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3015 < 0 before doing a fflush(). This bug looks as though it goes back to the
3016 introduction of TLS in release 3.20, but "sometimes" must have been rare
3017 because the tests only now provoked it.
3019 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3020 (this can affect the format of dates).
3022 47. exim_tidydb, when checking for the continued existence of a message for
3023 which it has found a message-specific retry record, was not finding
3024 messages that were in split spool directories. Consequently, it was
3025 deleting retry records that should have stayed in existence.
3027 48. Steve fixed some bugs in eximstats.
3029 49. The SPA authentication driver was not abandoning authentication and moving
3030 on to the next authenticator when an expansion was forced to fail,
3031 contradicting the general specification for all authenticators. Instead it
3032 was generating a temporary error. It now behaves as specified.
3034 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3035 (the order specifies the preference for clients). The order is now AES256,
3036 AES128, 3DES, ARCFOUR128.
3038 51. Small patch to Sieve code - explicitly set From: when generating an
3041 52. Exim crashed if a remote delivery caused a very long error message to be
3042 recorded - for instance if somebody sent an entire SpamAssassin report back
3043 as a large number of 550 error lines. This bug was coincidentally fixed by
3044 increasing the size of one of Exim's internal buffers (big_buffer) that
3045 happened as part of the Exiscan merge. However, to be on the safe side, I
3046 have made the code more robust (and fixed the comments that describe what
3049 53. Now that there can be additional text after "Completed" in log lines (if
3050 the queue_time_overall log selector is set), a one-byte patch to exigrep
3051 was needed to allow it to recognize "Completed" as not the last thing in
3054 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3055 patch that reportedly fixes this has been added. I am not expert enough to
3056 create a test for it. This is what the patch creator wrote:
3058 "I found a little strange behaviour of ldap code when working with
3059 Windows 2003 AD Domain, where users was placed in more than one
3060 Organization Units. When I tried to give exim partial DN, the exit code
3061 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3062 But simultaneously result of request was absolutely normal ldap result,
3063 so I produce this patch..."
3065 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3066 REFERENCE, so I have modified the code to exclude the patch when that macro
3069 55. Some experimental protocols are using DNS PTR records for new purposes. The
3070 keys for these records are domain names, not reversed IP addresses. The
3071 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3072 leaves it alone. Component reversal etc. now happens only for IP addresses.
3075 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3077 57. Double the size of the debug message buffer (to 2048) so that more of very
3078 long debug lines gets shown.
3080 58. The exicyclog utility now does better if the number of log files to keep
3081 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3083 59. Two changes related to the smtp_active_hostname option:
3085 (1) $smtp_active_hostname is now available as a variable.
3086 (2) The default for smtp_banner uses $smtp_active_hostname instead
3087 of $primary_hostname.
3089 60. The host_aton() function is supposed to be passed a string that is known
3090 to be a valid IP address. However, in the case of IPv6 addresses, it was
3091 not checking this. This is a hostage to fortune. Exim now panics and dies
3092 if the condition is not met. A case was found where this could be provoked
3093 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3094 components; fortuitously, this particular loophole had already been fixed
3095 by change 4.50/55 above.
3097 If there are any other similar loopholes, the new check in host_aton()
3098 itself should stop them being exploited. The report I received stated that
3099 data on the command line could provoke the exploit when Exim was running as
3100 exim, but did not say which command line option was involved. All I could
3101 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3102 running as the user.
3105 61. There was a buffer overflow vulnerability in the SPA authentication code
3106 (which came originally from the Samba project). I have added a test to the
3107 spa_base64_to_bits() function which I hope fixes it.
3110 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3111 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3113 63. The daemon start-up calls getloadavg() while still root for those OS that
3114 need the first call to be done as root, but it missed one case: when
3115 deliver_queue_load_max is set with deliver_drop_privilege. This is
3116 necessary for the benefit of the queue runner, because there is no re-exec
3117 when deliver_drop_privilege is set.
3119 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3120 This has been fixed.
3122 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3123 was in use, was not putting the data itself into the right store pool;
3124 consequently, it could be overwritten for a subsequent message in the same
3125 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3128 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3130 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3131 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3132 the function were treating the return as a boolean value, which happened to
3133 work because 0=false and not-0=true, but is not correct code.
3135 68. The host_aton() function was not handling scoped IPv6 addresses (those
3136 with, for example, "%eth0" on the end) correctly.
3138 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3139 negated items (that is, ~something) in unsigned ints. Some compilers
3140 apparently mutter when there is no cast.
3142 70. If an address verification called from an ACL failed, and did not produce a
3143 user-specific message (i.e. there was only a "system" message), nothing was
3144 put in $acl_verify_message. In this situation, it now puts the system
3147 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3148 session; change 4.31/43 added the unwanted input to the log line - except
3149 that it did not do this in the start of session case. It now does.
3151 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3152 This is wrong and can cause the other end to generate a synchronization
3153 error if it is another Exim or anything else that does the synchronization
3154 check. A QUIT command is no longer sent after a timeout.
3156 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3157 during host lookups.
3159 74. The defer_ok option of callout verification was not working if it was used
3160 when verifying addresses in header lines, that is, for this case:
3162 verify = header_sender/callout=defer_ok
3164 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3165 those file descriptors could be used for SMTP connections. If anything
3166 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3167 could be sent to the SMTP client, causing chaos. The daemon now opens
3168 stdin, stdout, and stderr to /dev/null when it puts itself into the
3171 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3172 log by default. The user can override this with suitable Perl magic.
3174 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3175 the log message when discard triggers, was not working for the DATA ACL or
3176 for the non-SMTP ACL.
3178 78. Error message wording change in sieve.c.
3180 79. If smtp_accept_max_per_host was set, the number of connections could be
3181 restricted to fewer than expected, because the daemon was trying to set up
3182 a new connection before checking whether the processes handling previous
3183 connections had finished. The check for completed processes is now done
3184 earlier. On busy systems, this bug wouldn't be noticed because something
3185 else would have woken the daemon, and it would have reaped the completed
3188 80. If a message was submitted locally by a user whose login name contained one
3189 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3190 It caused a spool format error. I have fixed the spool reading code. A
3191 related problem was that the "from" clause in the Received: line became
3192 illegal because of the space(s). It is now covered by ${quote_local_part.
3194 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3197 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3199 83. Patch from Sieve maintainer to fix unterminated string problem in
3200 "vacation" handling.
3202 84. Some minor changes to the Linux configuration files to help with other
3203 OS variants using glibc.
3205 85. One more patch for Sieve to update vacation handling to latest spec.
3208 ----------------------------------------------------
3209 See the note above about the 4.44 and 4.50 releases.
3210 ----------------------------------------------------
3216 1. Change 4.43/35 introduced a bug that caused file counts to be
3217 incorrectly computed when quota_filecount was set in an appendfile
3220 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3221 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3224 3. Give more explanation in the error message when the command for a transport
3225 filter fails to execute.
3227 4. There are several places where Exim runs a non-Exim command in a
3228 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3229 was being done only for the command run by the queryprogram router. It is
3230 now done for all such subprocesses. The other cases are: ${run, transport
3231 filters, and the commands run by the lmtp and pipe transports.
3233 5. Some older OS have a limit of 256 on the maximum number of file
3234 descriptors. Exim was using setrlimit() to set 1000 as a large value
3235 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3236 systems. I've change it so that if it can't get 1000, it tries for 256.
3238 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3239 was an oversight, and furthermore, ever since the addition of extra
3240 controls (e.g. 4.43/32), the checks on when to allow different forms of
3241 "control" were broken. There should now be diagnostics for all cases when a
3242 control that does not make sense is encountered.
3244 7. $recipients is now available in the predata ACL (oversight).
3246 8. Tidy the search cache before the fork to do a delivery from a message
3247 received from the command line. Otherwise the child will trigger a lookup
3248 failure and thereby defer the delivery if it tries to use (for example) a
3249 cached ldap connection that the parent has called unbind on.
3251 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3252 of $address_data from the recipient verification was clobbered by the
3253 sender verification.
3255 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3256 was its contents. (It was OK if the option was not defined at all.)
3258 11. A "Completed" log line is now written for messages that are removed from
3259 the spool by the -Mrm option.
3261 12. $host_address is now set to the target address during the checking of
3262 ignore_target_hosts.
3264 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3265 being passed; this would have caused $sender_host_name to have been used if
3266 matching the list had actually called for a host name (not very likely,
3267 since this list is usually IP addresses). A host name is now passed as
3270 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3271 code that specifies a non-restarting handler (typically sigaction() in
3272 modern systems) in an attempt to fix a rare and obscure crash bug.
3274 15. Narrowed the window for a race in the daemon that could cause it to ignore
3275 SIGCHLD signals. This is not a major problem, because they are used only to
3276 wake it up if nothing else does.
3278 16. A malformed maildirsize file could cause Exim to calculate negative values
3279 for the mailbox size or file count. Odd effects could occur as a result.
3280 The maildirsize information is now recalculated if the size or filecount
3283 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3284 support for a long time. Removed HAVE_SYS_VFS_H.
3286 18. Updated exipick to current release from John Jetmore.
3288 19. Allow an empty sender to be matched against a lookup in an address list.
3289 Previously the only cases considered were a regular expression, or an
3292 20. Exim went into a mad DNS lookup loop when doing a callout where the
3293 host was specified on the transport, if the DNS lookup yielded more than
3296 21. The RFC2047 encoding function was originally intended for short strings
3297 such as real names; it was not keeping to the 75-character limit for
3298 encoded words that the RFC imposes. It now respects the limit, and
3299 generates multiple encoded words if necessary. To be on the safe side, I
3300 have increased the buffer size for the ${rfc2047: expansion operator from
3303 22. Failure to deliver a bounce message always caused it to be frozen, even if
3304 there was an errors_to setting on the router. The errors_to setting is now
3307 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3308 canonical form (fully expanded) before being placed in
3309 $sender_host_address.
3311 24. Updated eximstats to version 1.33
3313 25. Include certificate and key file names in error message when GnuTLS fails
3314 to set them up, because the GnuTLS error message doesn't include the name
3315 of the failing file when there is a problem reading it.
3317 26. Expand error message when OpenSSL has problems setting up cert/key files.
3320 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3321 (this can affect the format of dates).
3323 28. exim_tidydb, when checking for the continued existence of a message for
3324 which it has found a message-specific retry record, was not finding
3325 messages that were in split spool directories. Consequently, it was
3326 deleting retry records that should have stayed in existence.
3328 29. eximstats updated to version 1.35
3329 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3330 1.35 - bugfix such that pie charts by volume are generated correctly
3332 30. The SPA authentication driver was not abandoning authentication and moving
3333 on to the next authenticator when an expansion was forced to fail,
3334 contradicting the general specification for all authenticators. Instead it
3335 was generating a temporary error. It now behaves as specified.
3337 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3338 (the order specifies the preference for clients). The order is now AES256,
3339 AES128, 3DES, ARCFOUR128.
3341 31. Small patch to Sieve code - explicitly set From: when generating an
3344 32. Exim crashed if a remote delivery caused a very long error message to be
3345 recorded - for instance if somebody sent an entire SpamAssassin report back
3346 as a large number of 550 error lines. This bug was coincidentally fixed by
3347 increasing the size of one of Exim's internal buffers (big_buffer) that
3348 happened as part of the Exiscan merge. However, to be on the safe side, I
3349 have made the code more robust (and fixed the comments that describe what
3352 33. Some experimental protocols are using DNS PTR records for new purposes. The
3353 keys for these records are domain names, not reversed IP addresses. The
3354 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3355 leaves it alone. Component reversal etc. now happens only for IP addresses.
3358 34. The host_aton() function is supposed to be passed a string that is known
3359 to be a valid IP address. However, in the case of IPv6 addresses, it was
3360 not checking this. This is a hostage to fortune. Exim now panics and dies
3361 if the condition is not met. A case was found where this could be provoked
3362 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3363 components; fortuitously, this particular loophole had already been fixed
3364 by change 4.50/55 or 4.44/33 above.
3366 If there are any other similar loopholes, the new check in host_aton()
3367 itself should stop them being exploited. The report I received stated that
3368 data on the command line could provoke the exploit when Exim was running as
3369 exim, but did not say which command line option was involved. All I could
3370 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3371 running as the user.
3374 35. There was a buffer overflow vulnerability in the SPA authentication code
3375 (which came originally from the Samba project). I have added a test to the
3376 spa_base64_to_bits() function which I hope fixes it.
3379 36. The daemon start-up calls getloadavg() while still root for those OS that
3380 need the first call to be done as root, but it missed one case: when
3381 deliver_queue_load_max is set with deliver_drop_privilege. This is
3382 necessary for the benefit of the queue runner, because there is no re-exec
3383 when deliver_drop_privilege is set.
3385 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3386 was in use, was not putting the data itself into the right store pool;
3387 consequently, it could be overwritten for a subsequent message in the same
3388 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3391 38. Sometimes the final signoff response after QUIT could fail to get
3392 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3393 < 0 before doing a fflush(). This bug looks as though it goes back to the
3394 introduction of TLS in release 3.20, but "sometimes" must have been rare
3395 because the tests only now provoked it.
3401 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3402 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3403 Now it returns an integer. A number of calls were still expecting a T/F
3404 return. Fortuitously, in all cases, the tests worked in OK situations,
3405 which is the norm. However, things would have gone wrong on any write
3406 failures on the smtp file descriptor. This function is used when sending
3407 messages over SMTP and also when doing verify callouts.
3409 2. When Exim is called to do synchronous delivery of a locally submitted
3410 message (the -odf or -odi options), it no longer closes stderr before doing
3413 3. Implemented the mua_wrapper option.
3415 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3417 5. Implemented the functions header_remove(), header_testname(),
3418 header_add_at_position(), and receive_remove_recipient(), and exported them
3421 6. If an ACL "warn" statement specified the addition of headers, Exim already
3422 inserted X-ACL-Warn: at the start if there was no header name. However, it
3423 was not making this test for the second and subsequent header lines if
3424 there were newlines in the string. This meant that an invalid header could
3425 be inserted if Exim was badly configured.
3427 7. Allow an ACL "warn" statement to add header lines at the start or after all
3428 the Received: headers, as well as at the end.
3430 8. Added the rcpt_4xx retry error code.
3432 9. Added postmaster_mailfrom=xxx to callout verification option.
3434 10. Added mailfrom=xxxx to the callout verification option, for verify=
3437 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3438 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3440 12. Inserted some casts to stop certain compilers warning when using pointer
3441 differences as field lengths or precisions in printf-type calls (mostly
3442 affecting debugging statements).
3444 13. Added optional readline() support for -be (dynamically loaded).
3446 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3447 same clock tick as a message's arrival, so that its received time was the
3448 same as the "first fail" time on the retry record, and that message
3449 remained on the queue past the ultimate address timeout, every queue runner
3450 would try a delivery (because it was past the ultimate address timeout) but
3451 after another failure, the ultimate address timeout, which should have then
3452 bounced the address, did not kick in. This was a "< instead of <=" error;
3453 in most cases the first failure would have been in the next clock tick
3454 after the received time, and all would be well.
3456 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3457 being recognized when the domain list was tested by the match_domain
3458 condition in an expansion string.
3460 16. Added the ${str2b64: operator.
3462 17. Exim was always calling setrlimit() to set a large limit for the number of
3463 processes, without checking whether the existing limit was already
3464 adequate. (It did check for the limit on file descriptors.) Furthermore,
3465 errors from getrlimit() and setrlimit() were being ignored. Now they are
3466 logged to the main and panic logs, but Exim does carry on, to try to do its
3467 job under whatever limits there are.
3469 18. Imported PCRE 5.0.
3471 19. Trivial typo in log message " temporarily refused connection" (the leading
3474 20. If the log selector return_path_on_delivery was set and an address was
3475 redirected to /dev/null, the delivery process crashed because it assumed
3476 that a return path would always be set for a "successful" delivery. In this
3477 case, the whole delivery is bypassed as an optimization, and therefore no
3480 21. Internal re-arrangement: the function for sending a challenge and reading
3481 a response while authentication was assuming a zero-terminated challenge
3482 string. It's now changed to take a pointer and a length, to allow for
3483 binary data in such strings.
3485 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3487 23. Exim was not respecting finduser_retries when seeking the login of the
3488 uid under which it was called; it was always trying 10 times. (The default
3489 setting of finduser_retries is zero.) Also, it was sleeping after the final
3490 failure, which is pointless.
3492 24. Implemented tls_on_connect_ports.
3494 25. Implemented acl_smtp_predata.
3496 26. If the domain in control=submission is set empty, Exim assumes that the
3497 authenticated id is a complete email address when it generates From: or
3498 Sender: header lines.
3500 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3501 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3502 chown and chgrp in /bin and hostname in /usr/bin.
3504 28. Exim was keeping the "process log" file open after each use, just as it
3505 does for the main log. This opens the possibility of it remaining open for
3506 long periods when the USR1 signal hits a daemon. Occasional processlog
3507 errors were reported, that could have been caused by this. Anyway, it seems
3508 much more sensible not to leave this file open at all, so that is what now
3511 29. The long-running daemon process does not normally write to the log once it
3512 has entered its main loop, and it closes the log before doing so. This is
3513 so that log files can straightforwardly be renamed and moved. However,
3514 there are a couple of unusual error situations where the daemon does write
3515 log entries, and I had neglected to close the log afterwards.
3517 30. The text of an SMTP error response that was received during a remote
3518 delivery was being truncated at 512 bytes. This is too short for some of
3519 the long messages that one sometimes sees. I've increased the limit to
3522 31. It is now possible to make retry rules that apply only when a message has a
3523 specific sender, in particular, an empty sender.
3525 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3526 it possible to be selective about when SMTP synchronization is enforced.
3528 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3530 32. Implemented hosts_connection_nolog.
3532 33. Added an ACL for QUIT.
3534 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3537 35. Added mailbox_size and mailbox_filecount to appendfile.
3539 36. Added control = no_multiline_responses to ACLs.
3541 37. There was a bug in the logic of the code that waits for the clock to tick
3542 in the case where the clock went backwards by a substantial amount such
3543 that the microsecond fraction of "now" was more than the microsecond
3544 fraction of "then" (but the whole seconds number was less).
3546 38. Added support for the libradius Radius client library this is found on
3547 FreeBSD (previously only the radiusclient library was supported).
3553 1. When certain lookups returned multiple values in the form name=value, the
3554 quoting of the values was not always being done properly. Specifically:
3555 (a) If the value started with a double quote, but contained no whitespace,
3557 (b) If the value contained whitespace other than a space character (i.e.
3558 tabs or newlines or carriage returns) it was not quoted.
3559 This fix has been applied to the mysql and pgsql lookups by writing a
3560 separate quoting function and calling it from the lookup code. The fix
3561 should probably also be applied to nisplus, ibase and oracle lookups, but
3562 since I cannot test any of those, I have not disturbed their existing code.
3564 2. A hit in the callout cache for a specific address caused a log line with no
3565 reason for rejecting RCPT. Now it says "Previous (cached) callout
3566 verification failure".
3568 3. There was an off-by-one bug in the queryprogram router. An over-long
3569 return line was truncated at 256 instead of 255 characters, thereby
3570 overflowing its buffer with the terminating zero. As well as fixing this, I
3571 have increased the buffer size to 1024 (and made a note to document this).
3573 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3574 when Exim is waiting for an SMTP response from a remote server, Exim
3575 restarts its select() call on the socket, thereby resetting its timeout.
3576 This is not a problem when such interrupts are rare. Somebody set up a cron
3577 job to run exiwhat every 2 minutes, which is less than the normal select()
3578 timeout (5 or 10 minutes). This meant that the select() timeout never
3579 kicked in because it was always reset. I have fixed this by comparing the
3580 time when an interrupt arrives with the time at the start of the first call
3581 to select(). If more time than the timeout has elapsed, the interrupt is
3582 treated as a timeout.
3584 5. Some internal re-factoring in preparation for the addition of Sieve
3585 extensions (by MH). In particular, the "personal" test is moved to a
3586 separate function, and given an option for scanning Cc: and Bcc: (which is
3587 not set for Exim filters).
3589 6. When Exim created an email address using the login of the caller as the
3590 local part (e.g. when creating a From: or Sender: header line), it was not
3591 quoting the local part when it contained special characters such as @.
3593 7. Installed new OpenBSD configuration files.
3595 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3596 try to make them clearer.
3598 9. Callout options, other than the timeout value, were being ignored when
3599 verifying sender addresses in header lines. For example, when using
3601 verify = header_sender/callout=no_cache
3603 the cache was (incorrectly) being used.
3605 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3606 only the Cygwin environment.
3608 11. When return_path_on_delivery was set as a log selector, if different remote
3609 addresses in the same message used different return paths and parallel
3610 remote delivery occurred, the wrong values would sometimes be logged.
3611 (Whenever a remote delivery process finished, the return path value from
3612 the most recently started remote delivery process was logged.)
3614 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3615 lines when AUTH and/or TLS are in use. This is the "received protocol"
3616 field. Exim used to use "asmtp" for authenticated SMTP, without any
3617 indication (in the protocol name) for TLS use. Now it follows the RFC and
3618 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3619 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3620 names appear in log lines as well as in Received: header lines.
3622 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3623 extensions, and comparison tests, and to fix some bugs.
3625 14. Changes to the "personal" filter test:
3627 (1) The test was buggy in that it was just doing the equivalent of
3628 "contains" tests on header lines. For example, if a user's address was
3629 anne@some.where, the "personal" test would incorrectly be true for
3631 To: susanne@some.where
3633 This test is now done by extracting each address from the header in turn,
3634 and checking the entire address. Other tests that are part of "personal"
3635 are now done using regular expressions (for example, to check local parts
3636 of addresses in From: header lines).
3638 (2) The list of non-personal local parts in From: addresses has been
3639 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3640 taken from the Sieve specification recommendations.
3642 (3) If the message contains any header line starting with "List-" it is
3643 treated as non-personal.
3645 (4) The test for "circular" in the Subject: header line has been removed
3646 because it now seems ill-conceived.
3648 15. Minor typos in src/EDITME comments corrected.
3650 16. Installed latest exipick from John Jetmore.
3652 17. If headers_add on a router specified a text string that was too long for
3653 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3654 of string_sprintf() is now avoided.
3656 18. $message_body_size was not set (it was always zero) when running the DATA
3657 ACL and the local_scan() function.
3659 19. For the "mail" command in an Exim filter, no default was being set for
3660 the once_repeat time, causing a random time value to be used if "once" was
3661 specified. (If the value happened to be <= 0, no repeat happened.) The
3662 default is now 0s, meaning "never repeat". The "vacation" command was OK
3663 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3664 (I found it when inspecting the code).
3666 20. There is now an overall timeout for performing a callout verification. It
3667 defaults to 4 times the callout timeout, which applies to individual SMTP
3668 commands during the callout. The overall timeout applies when there is more
3669 than one host that can be tried. The timeout is checked before trying the
3670 next host. This prevents very long delays if there are a large number of
3671 hosts and all are timing out (e.g. when the network connections are timing
3672 out). The value of the overall timeout can be changed by specifying an
3673 additional sub-option for "callout", called "maxwait". For example:
3675 verify = sender/callout=5s,maxwait=20s
3677 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3678 to the end before writing, but this should make it even safer).
3680 22. Exim was forgetting that it had advertised PIPELINING for the second and
3681 subsequent messages on an SMTP connection. It was also not resetting its
3682 memory on STARTTLS and an internal HELO.
3684 23. When Exim logs an SMTP synchronization error within a session, it now
3685 records whether PIPELINING has been advertised or not.
3687 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3688 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3689 rather than long int.
3691 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3693 26. Added the never_mail option to autoreply.
3699 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3700 crash if the getsockname() call failed; this can happen if a connection is
3701 closed very soon after it is established. The problem was simply in the
3702 order in which certain operations were done, causing Exim to try to write
3703 to the SMTP stream before it had set up the file descriptor. The bug has
3704 been fixed by making things happen in the correct order.
3710 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3711 before the connection was closed, thus losing the rejection response.
3713 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3714 some early Solaris releases, but causes trouble in current releases where
3715 socklen_t is defined.
3717 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3720 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3723 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3724 that releasing the top of it at the end releases what was used for sub-
3725 expansions (unless the block got too big). However, discard this block if
3726 the first thing is a variable or header, so that we can use its block when
3727 it is dynamic (useful for very large $message_headers, for example).
3729 6. Lookups now cache *every* query, not just the most recent. A new, separate
3730 store pool is used for this. It can be recovered when all lookup caches are
3731 flushed. Lookups now release memory at the end of their result strings.
3732 This has involved some general refactoring of the lookup sources.
3734 7. Some code has been added to the store_xxx() functions to reduce the amount
3735 of flapping under certain conditions.
3737 8. log_incoming_interface used to affect only the <= reception log lines. Now
3738 it causes the local interface and port to be added to several more SMTP log
3739 lines, for example "SMTP connection from", and rejection lines.
3741 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3743 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3745 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3746 could be overwritten at the end of the current message (or the start of a
3747 new message if it was set in a HELO ACL). The value is now preserved for
3748 the duration of the SMTP connection.
3750 12. If a transport had a headers_rewrite setting, and a matching header line
3751 contained an unqualified address, that address was qualified, even if it
3752 did not match any rewriting rules. The underlying bug was that the values
3753 of the flags that permit the existence of unqualified sender and recipient
3754 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3755 non-local messages, and by -bnq for local messages) were not being
3756 preserved with the message after it was received.
3758 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3759 "next input=" as part of the text comprising the host identity instead of
3760 the correct text. The code was using the same buffer for two different
3761 strings. However, depending on which order the printing function evaluated
3762 its arguments, the bug did not always show up. Under Linux, for example, my
3763 test suite worked just fine.
3765 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3766 doesn't work with some older versions of Perl. It has been changed to "my",
3767 which in any case is probably the better facility to use.
3769 15. A really picky compiler found some instances of statements for creating
3770 error messages that either had too many or two few arguments for the format
3773 16. The size of the buffer for calls to the DNS resolver has been increased
3774 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3775 for addresses that have a lot of PTR records. This alleviates a problem; it
3776 does not fully solve it.
3778 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3779 buffer now truncates the list and logs the incident, which is the same
3780 action as happens when Exim is looking up a host name and its aliases.
3781 Previously in this situation something unpredictable would happen;
3782 sometimes it was "internal error: store_reset failed".
3784 18. If a server dropped the connection unexpectedly when an Exim client was
3785 using GnuTLS and trying to read a response, the client delivery process
3786 crashed while trying to generate an error log message.
3788 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3789 string, for example:
3791 warn message = H1: something\nH2: something
3793 the text was added as a single header line from Exim's point of view
3794 though it ended up OK in the delivered message. However, searching for the
3795 second and subsequent header lines using $h_h2: did not work. This has been
3796 fixed. Similarly, if a system filter added multiple headers in this way,
3797 the routers could not see them.
3799 20. Expanded the error message when iplsearch is called with an invalid key to
3800 suggest using net-iplsearch in a host list.
3802 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3803 ACLs are no longer actually imposed (and a message to that effect is
3806 22. If a "gecos" field in a passwd entry contained escaped characters, in
3807 particular, if it contained a \" sequence, Exim got it wrong when building
3808 a From: or a Sender: header from that name. A second bug also caused
3809 incorrect handling when an unquoted " was present following a character
3810 that needed quoting.
3812 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3813 was not being matched caselessly.
3815 24. Arranged for all hyphens in the exim.8 source to be escaped with
3818 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3819 properly. Recipient callout cache records were still being keyed to include
3820 the sender, even when use_sender was set false. This led to far more
3821 callouts that were necessary. The sender is no longer included in the key
3822 when use_sender is false.
3824 26. Added "control = submission" modifier to ACLs.
3826 27. Added the ${base62d: operator to decode base 62 numbers.
3828 28. dnsdb lookups can now access SRV records.
3830 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3831 the configuration file.
3833 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3834 (-v) mode. This makes the output for a verbose queue run more intelligible.
3836 31. Added a use_postmaster feature to recipient callouts.
3838 32. Added the $body_zerocount variable, containing the number of binary zero
3839 bytes in the message body.
3841 33. The time of last modification of the "new" subdirectory is now used as the
3842 "mailbox time last read" when there is a quota error for a maildir
3845 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3847 35. Added +ignore_unknown as a special item in host lists.
3849 36. Code for decoding IPv6 addresses in host lists is now included, even if
3850 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3851 address was recognized as an IP address, but was then not correctly decoded
3852 into binary, causing unexpected and incorrect effects when compared with
3859 1. Very minor rewording of debugging text in manualroute to say "list of
3860 hosts" instead of "hostlist".
3862 2. If verify=header_syntax was set, and a header line with an unqualified
3863 address (no domain) and a large number of spaces between the end of the
3864 name and the colon was received, the reception process suffered a buffer
3865 overflow, and (when I tested it) crashed. This was caused by some obsolete
3866 code that should have been removed. The fix is to remove it!
3868 3. When running in the test harness, delay a bit after writing a bounce
3869 message to get a bit more predictability in the log output.
3871 4. Added a call to search_tidyup() just before forking a reception process. In
3872 theory, someone could use a lookup in the expansion of smtp_accept_max_
3873 per_host which, without the tidyup, could leave open a database connection.
3875 5. Added the variables $recipient_data and $sender_data which get set from a
3876 lookup success in an ACL "recipients" or "senders" condition, or a router
3877 "senders" option, similar to $domain_data and $local_part_data.
3879 6. Moved the writing of debug_print from before to after the "senders" test
3882 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3883 problems for message scanning, either using a data ACL, or using
3884 local_scan() because the Received: header was not generated till after they
3885 were called (in order to set the time as the time of reception completion).
3886 I have revised the way this works. The header is now generated after the
3887 body is received, but before the ACL or local_scan() are called. After they
3888 are run, the timestamp in the header is updated.
3894 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3895 before starting a queue runner without re-exec. This happened only when
3896 deliver_drop_privilege was set or when the Exim user was set to root. The
3897 effect of the bug was that timeouts during subsequent deliveries caused
3898 crashes instead of being properly handled. The handler is now left at its
3899 default (and expected) setting.
3901 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3902 message, again when deliver_drop_privilege is set or Exim is run as root.
3903 The bug described in (1) was not present in this case, but the tidying up
3904 of the other signals was missing. I have made the two cases consistent.
3906 3. The ignore_target_hosts setting on a manualroute router was being ignored
3907 for hosts that were looked up using the /MX notation.
3909 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3912 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3913 operated on the sender address. After changing the $sender_address to <>
3914 for the sender address verify, Exim was re-instated it as the original
3915 (before rewriting) address, but remembering that it had rewritten it, so it
3916 wasn't rewriting it again. This bug also had the effect of breaking the
3917 sender address verification caching when the sender address was rewritten.
3919 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3920 This has been changed so that if the ip literal address matches
3921 ignore_target_hosts, the router declines.
3923 7. Added expansion conditions match_domain, match_address, and match_local_
3924 part (NOT match_host).
3926 8. The placeholder for the Received: header didn't have a length field set.
3928 9. Added code to Exim itself and to exim_lock to test for a specific race
3929 condition that could lead to file corruption when using MBX delivery. The
3930 issue is with the lockfile that is created in /tmp. If this file is removed
3931 after a process has opened it but before that process has acquired a lock,
3932 there is the potential for a second process to recreate the file and also
3933 acquire a lock. This could lead to two Exim processes writing to the file
3934 at the same time. The added code performs the same test as UW imapd; it
3935 checks after acquiring the lock that its file descriptor still refers to
3936 the same named file.
3938 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3939 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3942 11. Added the smtp_active_hostname option. If used, this will typically be made
3943 to depend on the incoming interface address. Because $interface_address is
3944 not set up until the daemon has forked a reception process, error responses
3945 that can happen earlier (such as "too many connections") no longer contain
3948 12. If an expansion in a condition on a "warn" statement fails because a lookup
3949 defers, the "warn" statement is abandoned, and the next ACL statement is
3950 processed. Previously this caused the whole ACL to be aborted.
3952 13. Added the iplsearch lookup type.
3954 14. Added ident_timeout as a log selector.
3956 15. Added tls_certificate_verified as a log selector.
3958 16. Added a global option tls_require_ciphers (compare the smtp transport
3959 option of the same name). This controls incoming TLS connections.
3961 17. I finally figured out how to make tls_require_ciphers do a similar thing
3962 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3963 before starting the TLS session.
3965 18. Tabs are now shown as \t in -bP output.
3967 19. If the log selector return_path_on_delivery was set, Exim crashed when
3968 bouncing a message because it had too many Received: header lines.
3970 20. If two routers both had headers_remove settings, and the first one included
3971 a superfluous trailing colon, the final name in the first list and the
3972 first name in the second list were incorrectly joined into one item (with a
3973 colon in the middle).
3979 1. Added -C and -D options to the exinext utility, mainly to make it easier
3980 to include in the automated testing, but these could be helpful when
3981 multiple configurations are in use.
3983 2. The exinext utility was not formatting the output nicely when there was
3984 an alternate port involved in the retry record key, nor when there was a
3985 message id as well (for retries that were specific to a specific message
3986 and a specific host). It was also confused by IPv6 addresses, because of
3987 the additional colons they contain. I have fixed the IPv4 problem, and
3988 patched it up to do a reasonable job for IPv6.
3990 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3991 delivery, the log line now contains "pipelined" if PIPELINING was used.
3993 4. An SMTP transport process used to panic and die if the bind() call to set
3994 an explicit outgoing interface failed. This has been changed; it is now
3995 treated in the same way as a connect() failure.
3997 5. A reference to $sender_host_name in the part of a conditional expansion
3998 that was being skipped was still causing a DNS lookup. This no longer
4001 6. The def: expansion condition was not recognizing references to header lines
4002 that used bh_ and bheader_.
4004 7. Added the _cache feature to named lists.
4006 8. The code for checking quota_filecount in the appendfile transport was
4007 allowing one more file than it should have been.
4009 9. For compatibility with Sendmail, the command line option
4017 and sets the incoming protocol and host name (for trusted callers). The
4018 host name and its colon can be omitted when only the protocol is to be set.
4019 Note the Exim already has two private options, -pd and -ps, that refer to
4020 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4021 "s", but I don't think that's a major issue.
4023 10. A number of refactoring changes to the code, none of which should affect
4026 (a) The number of logging options was getting close to filling up the
4027 32-bit word that was used as a bit map. I have split them into two classes:
4028 those that are passed in the argument to log_write(), and those that are
4029 only ever tested independently outside of that function. These are now in
4030 separate 32-bit words, so there is plenty of room for expansion again.
4031 There is no change in the user interface or the logging behaviour.
4033 (b) When building, for example, log lines, the code previously used a
4034 macro that called string_cat() twice, in order to add two strings. This is
4035 not really sufficiently general. Furthermore, there was one instance where
4036 it was actually wrong because one of the argument was used twice, and in
4037 one call a function was used. (As it happened, calling the function twice
4038 did not affect the overall behaviour.) The macro has been replaced by a
4039 function that can join an arbitrary number of extra strings onto a growing
4042 (c) The code for expansion conditions now uses a table and a binary chop
4043 instead of a serial search (which was left over from when there were very
4044 few conditions). Also, it now recognizes conditions like "pam" even when
4045 the relevant support is not compiled in: a suitably worded error message is
4046 given if an attempt is made to use such a condition.
4048 11. Added ${time_interval:xxxxx}.
4050 12. A bug was causing one of the ddress fields not to be passed back correctly
4051 from remote delivery subprocesses. The field in question was not being
4052 subsequently used, so this caused to problems in practice.
4054 13. Added new log selectors queue_time and deliver_time.
4056 14. Might have fixed a bug in maildirsizefile handling that threw up
4057 "unexpected character" debug warnings, and recalculated the data
4058 unnecessarily. In any case, I expanded the warning message to give more
4061 15. Added the message "Restricted characters in address" to the statements in
4062 the default ACL that block characters like @ and % in local parts.
4064 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4065 Three changes have been made:
4067 (a) There was a serious bug; a negative response to MAIL caused the whole
4068 recipient domain to be cached as invalid, thereby blocking all messages
4069 to all local parts at the same domain, from all senders. This bug has
4070 been fixed. The domain is no longer cached after a negative response to
4071 MAIL if the sender used is not empty.
4073 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4076 (c) A new callout option, "use_sender" has been added for people who want
4077 the modified behaviour.
4083 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4086 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4087 indeed breaks things for older releases.
4089 3. Added additional logging to the case where there is a problem reading data
4090 from a filter that is running in a subprocess using a pipe, in order to
4091 try to track down a specific problem.
4093 4. Testing facility fudge: when running in the test harness and attempting
4094 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4095 getting "No route to host". Convert this to a timeout.
4097 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4100 6. Some OS don't have socklen_t but use size_t instead. This affects the
4101 fifth argument of getsockopt() amongst other things. This is now
4102 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4103 can be set for individual OS. I have set it for SunOS5, OSF1, and
4104 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4105 some earlier ones do not.
4107 7. Change 4.30/15 was not doing the test caselessly.
4109 8. The standard form for an IPv6 address literal was being rejected by address
4110 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4111 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4112 this, as well as the form without the "IPv6" on the front (but only when
4113 address literals are enabled, of course).
4115 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4117 10. Exim crashed if a message with an empty sender address specified by -f
4118 encountered a router with an errors_to setting. This could be provoked only
4119 by a command such as
4123 where an empty string was supplied; "<>" did not hit this bug.
4125 11. Installed PCRE release 4.5.
4127 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4128 remained set. It is now erased.
4130 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4131 times from message ids (which are base 36 rather than the normal 62).
4133 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4134 were being counted as actual protocol errors, and logged if the log
4135 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4136 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4137 and DATA following a set of rejected RCPTs do not count as protocol errors.
4138 In other words, Exim assumes they were pipelined, though this may not
4139 actually be the case. Of course, in all cases the client gets an
4140 appropriate error code.
4142 15. If a lookup fails in an ACL condition, a message about the failure may
4143 be available; it is used if testing the ACL cannot continue, because most
4144 such messages specify what the cause of the deferral is. However, some
4145 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4146 that caused an old message to be retained and used if a later statement
4147 caused a defer, replacing the real cause of the deferral.
4149 16. If an IP address had so many PTR records that the DNS lookup buffer
4150 was not large enough to hold them, Exim could crash while trying to process
4151 the truncated data. It now detects and logs this case.
4153 17. Further to 4.21/58, another change has been made: if (and only if) the
4154 first line of a message (the first header line) ends with CRLF, a bare LF
4155 in a subsequent header line has a space inserted after it, so as not to
4156 terminate the header.
4158 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4159 unnecessarily, used atoi() instead of strtol(), and didn't check the
4160 termination when getting file sizes from file names by regex.
4162 19. Completely re-implemented the support for maildirsize files, in the light
4163 of a number of problems with the previous contributed implementation
4164 (4.30/29). In particular:
4166 . If the quota is zero, the maildirsize file is maintained, but no quota is
4169 . If the maildir directory does not exist, it is created before any attempt
4170 to write a maildirsize file.
4172 . The quota value in the file is just a cache; if the quota is changed in
4173 the transport, the new value overrides.
4175 . A regular expression is available for excluding directories from the
4178 20. The autoreply transport checks the characters in options that define the
4179 message's headers; it allows continued headers, but it was checking with
4180 isspace() after an embedded newline instead of explicitly looking for a
4183 21. If all the "regular" hosts to which an address was routed had passed their
4184 expiry times, and had not reached their retry times, the address was
4185 bounced, even if fallback hosts were defined. Now Exim should go on to try
4188 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4189 equivalent code in the SMTP transport. Some hosts send humungous responses
4190 to HELO/EHLO, more than 1024 it seems.
4192 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4193 gives compiler warnings in some environments. I've now done it "properly",
4196 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4197 (because of problems with the built-in one) was declared to return uschar *
4198 instead of char *, causing compiler failure.
4200 25. Fixed a file descriptor leak when processing alias/forward files.
4202 26. Fixed a minor format string issue in dbfn.c.
4204 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4206 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4207 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4210 29. When a sender address is verified, it is cached, to save repeating the test
4211 when there is more than one recipient in a message. However, when the
4212 verification involves a callout, it is possible for different callout
4213 options to be set for different recipients. It is too complicated to keep
4214 track of this in the cache, so now Exim always runs a verification when a
4215 callout is required, relying on the callout cache for the optimization.
4216 The overhead is duplication of the address routing, but this should not be
4219 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4220 to be verified with callout=postmaster, and the main callout worked but the
4221 postmaster check failed, the verification correctly failed. However, if a
4222 subsequent RCPT command asked for sender verification *without* the
4223 postmaster check, incorrect caching caused this verification also to fail,
4226 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4227 it was not caching the DNS options (qualify_single, search_parents) that
4228 were used when the lookup failed. A subsequent lookup with different
4229 options therefore always gave the same answer, though there were cases
4230 where it should not have. (Example: a "domains = !$mx_any" option on a
4231 dnslookup router: the "domains" option is always processed without any
4232 widening, but the router might have qualify_single set.) Now Exim uses the
4233 cached value only when the same options are set.
4235 32. Added John Jetmore's "exipick" utility to the distribution.
4237 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4238 than a timeout (e.g. a certificate is required, and is not provided), an
4239 Exim server now closes the connection immediately. Previously it waited for
4240 the client to close - but if the client is SSL, it seems that they each
4241 wait for each other, leading to a delay before one of them times out.
4243 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4244 maintained 0.8.x compatibility because I don't think many are using it, and
4245 it is clearly obsolete.
4247 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4250 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4251 client certificate was expired. A simple patch fixes this, though I don't
4252 understand the full logic of why the verify callback is called multiple
4255 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4256 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4257 which causes problems with some clients (such as the Certicom SSL Plus
4258 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4259 disables the coutermeasure allowing Eudora to connect."
4261 38. Exim was not checking that a write() to a log file succeeded. This could
4262 lead to Bad Things if a log got too big, in particular if it hit a file
4263 size limit. Exim now panics and dies if it cannot write to a log file, just
4264 as it does if it cannot open a log file.
4266 39. Modified OS/Makefile-Linux so that it now contains
4268 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4270 The two -D definitions ensure that Exim is compiled with large file
4271 support, which makes it possible to handle log files that are bigger than
4274 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4275 instance) a domain was checked against a named list that involved a lookup,
4276 causing $domain_data to be set, then another domain was checked against the
4277 same list, then the first domain was re-checked, the value of $domain_data
4278 after the final check could be wrong. In particular, if the second check
4279 failed, it could be set empty. This bug probably also applied to
4282 41. The strip_trailing_dot option was not being applied to the address given
4283 with the -f command-line option.
4285 42. The code for reading a message's header from the spool was incrementing
4286 $received_count, but never initializing it. This meant that the value was
4287 incorrect (doubled) while delivering a message in the same process in which
4288 it was received. In the most common configuration of Exim, this never
4289 happens - a fresh exec is done - but it can happen when
4290 deliver_drop_privilege is set.
4292 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4293 it now includes up to 150 characters of the unexpected data in the log
4296 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4297 and building data strings. The size of both of these buffers was 10 000
4298 bytes - far larger than anybody would *ever* want, thought I. Needless to
4299 say, somebody hit the limit. I have increased the maximum line length to
4300 20 000 and the maximum data length of concatenated lines to 100 000. I have
4301 also fixed two bugs, because there was no checking on these buffers. Tsk,
4302 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4303 buffer is too small.
4305 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4306 lsearch lookups. Now it does.
4308 46. When parsing a route_list item in a manualroute router, a fixed-length
4309 buffer was used for the list of hosts. I made this 1024 bytes long,
4310 thinking that nobody would ever have a list of hosts that long. Wrong.
4311 Somebody had a whole pile of complicated expansion conditions, and the
4312 string was silently truncated, leading to an expansion error. It turns out
4313 that it is easier to change to an unlimited length (owing to other changes
4314 that have happened since this code was originally written) than to build
4315 structure for giving a limitation error. The length of the item that
4316 expands into the list of hosts is now unlimited.
4318 47. The lsearch lookup could not handle data where the length of text line was
4319 more than 4095 characters. Such lines were truncated, leading to shortened
4320 data being returned. It should now handle lines of any length.
4322 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4323 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4326 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4327 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4328 no longer happens in comment lines. A list of replacements is now placed
4329 at the head of all of the source files, except those whose only change is
4330 to replace PERL_COMMAND in the very first #! line.
4332 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4333 messages on the queue, with a bottom-up merge sort, using code contributed
4334 by Michael Haardt. This should make operations like -bp somewhat faster on
4335 large queues. It won't affect queue runners, except when queue_run_in_order
4338 51. Installed eximstats 1.31 in the distribution.
4340 52. Added support for SRV lookups to the dnslookup router.
4342 53. If an ACL referred to $message_body or $message_body_end, the value was not
4343 reset for any messages that followed in the same SMTP session.
4345 54. The store-handling optimization for building very long strings was not
4346 differentiating between the different store pools. I don't think this
4347 actually made any difference in practice, but I've tidied it.
4349 55. While running the routers to verify a sender address, $sender_address
4350 was still set to the sender address. This is wrong, because when routing to
4351 send a bounce to the sender, it would be empty. Therefore, I have changed
4352 it so that, while verifying a sender address, $sender_address is set to <>.
4353 (There is no change to what happens when verifying a recipient address.)
4355 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4356 target A or AAAA records (if not already returned) without resetting the
4357 qualify_single or search_parents options of the DNS resolver. These are
4358 inappropriate in this case because the targets of MX and SRV records must
4359 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4360 target that, when qualified, matched something in the local domain. These
4361 two options are now turned off when doing these lookups.
4363 57. It seems that at least some releases of Reiserfs (which does not have the
4364 concept of a fixed number of inodes) returns zero and not -1 for the
4365 number of available inodes. This interacted badly with check_spool_inodes,
4366 which assumed that -1 was the "no such thing" setting. What I have done is
4367 to check that the total number of inodes is greater than zero before doing
4368 the test of how many are available.
4370 58. When a "warn" ACL statement has a log_message modifier, the message is
4371 remembered, and not repeated. This is to avoid a lot of repetition when a
4372 message has many recipients that cause the same warning to be written.
4373 Howewer, Exim was preserving the list of already written lines for an
4374 entire SMTP session, which doesn't seem right. The memory is now reset if a
4375 new message is started.
4377 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4378 debugging output unless log_rewrite was also set.
4380 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4381 of (int)(handle) when we know that handle contains (void *)(-1).
4383 61. The Exim daemon panic-logs an error return when it closes the incoming
4384 connection. However "connection reset by peer" seems to be common, and
4385 isn't really an error worthy of noting specially, so that particular error
4388 62. When Exim is trying to find all the local interfaces, it used to panic and
4389 die if the ioctl to get the interface flags failed. However, it seems that
4390 on at least one OS (Solaris 9) it is possible to have an interface that is
4391 included in the list of interfaces, but for which you get a failure error
4392 for this call. This happens when the interface is not "plumbed" into a
4393 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4394 failure of the "get flags" call assumes that the interface is down.
4396 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4397 makes life easier for people who are doing arithmetic on fields extracted
4398 from dates, where you often get leading zeros that should not be
4399 interpreted as octal.
4401 64. Added qualify_domain to the redirect router, to override the global
4404 65. If a pathologically long header line contained very many addresses (the
4405 report of this problem mentioned 10 000) and each of them was rewritten,
4406 Exim could use up a very large amount of memory. (It kept on making new
4407 copies of the header line as it rewrote, and never released the old ones.)
4408 At the expense of a bit more processing, the header rewriting function has
4409 been changed so that it no longer eats memory in this way.
4411 66. The generation of the Received: header has been moved from the time that a
4412 message starts to be received, to the time that it finishes. The timestamp
4413 in the Received: header should now be very close to that of the <= log
4414 line. There are two side-effects of this change:
4416 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4417 logged header lines no longer include the local Received: line, because
4418 it has not yet been created. The same applies to a copy of the message
4419 that is returned to a non-SMTP sender when a message is rejected.
4421 (b) When a filter file is tested using -bf, no additional Received: header
4422 is added to the test message. After some thought, I decided that this
4425 This change does not affect the value of $received_for. It is still set
4426 after address rewriting, but before local_scan() is called.
4428 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4430 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4431 gave an unhelpful panic error message, and a defer error. I have managed to
4432 change this behaviour so that it now rejects any supplied certificate,
4433 which seems right, as the list of acceptable certificates is empty.
4435 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4436 gave an unhelpful defer error. I have not managed to make this reject any
4437 supplied certificates, but the error message it gives is "no certificate
4438 supplied", which is not helpful.
4440 70. exigrep's output now also includes lines that are not associated with any
4441 message, but which match the given pattern. Implemented by a patch from
4442 Martin Sluka, which also tidied up the Perl a bit.
4444 71. Recipient callout verification, like sender verification, was using <> in
4445 the MAIL FROM command. This isn't really the right thing, since the actual
4446 sender may affect whether the remote host accepts the recipient or not. I
4447 have changed it to use the actual sender in the callout; this means that
4448 the cache record is now keyed on a recipient/sender pair, not just the
4449 recipient address. There doesn't seem to be a real danger of callout loops,
4450 since a callout by the remote host to check the sender would use <>.
4451 [SEE ABOVE: changed after hitting problems.]
4453 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4454 temporary errors. However, in the case of such a code being given after
4455 the end of a data transmission (i.e. after ".") Exim was failing to write
4456 a retry record for the message. (Yes, there was some broken host that was
4457 actually sending 8xx at this point.)
4459 73. An unknown lookup type in a host list could cause Exim to panic-die when
4460 the list was checked. (An example that provoked this was putting <; in the
4461 middle of a list instead of at the start.) If this happened during a DATA
4462 ACL check, a -D file could be left lying around. This kind of configuration
4463 error no longer causes Exim to die; instead it causes a defer errror. The
4464 incident is still logged to the main and panic logs.
4466 74. Buglet left over from Exim 3 conversion. The message "too many messages
4467 in one connection" was written to the rejectlog but not the mainlog, except
4468 when address rewriting (yes!) was being logged.
4470 75. Added write_rejectlog option.
4472 76. When a system filter was run not as root (that is, when system_filter_user
4473 was set), the values of the $n variables were not being returned to the
4474 main process; thus, they were not subsequently available in the $sn
4477 77. Added +return_path_on_delivery log selector.
4479 78. A connection timeout was being treated differently from recipients deferred
4480 when testing hosts_max_try with a message that was older than the host's
4481 retry timeout. (The host should not be counted, thus allowing all hosts to
4482 be tried at least once before bouncing.) This may have been the cause of an
4483 occasionally reported bug whereby a message would remain on the queue
4484 longer than the retry timeout, but would be bounced if a delivery was
4485 forced. I say "may" because I never totally pinned down the problem;
4486 setting up timeout/retry tests is difficult. See also the next item.
4488 79. The ultimate address timeout was not being applied to errors that involved
4489 a combination of host plus message (for example, a timeout on a MAIL
4490 command). When an address resolved to a number of possible hosts, and they
4491 were not all tried for each delivery (e.g. because of hosts_max_try), a
4492 message could remain on the queue longer than the retry timeout.
4494 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4497 81. Fixed an obscure SMTP outgoing bug which required at least the following
4498 conditions: (a) there was another message waiting for the same server;
4499 (b) the server returned 5xx to all RCPT commands in the first message so
4500 that the message was not completed; (c) the server dropped the connection
4501 or gave a negative response to the RSET that Exim sends to abort the
4502 transaction. The observed case was a dropped connection after DATA that had
4503 been sent in pipelining mode. That is, the server had advertised PIPELINING
4504 but was not implementing it correctly. The effect of the bug was incorrect
4505 behaviour, such as trying another host, and this could lead to a crash.
4511 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4512 and daemon.c were passed as pointers to ints; they should have been
4513 pointers to socklen_t variables (which are typically unsigned ints).
4515 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4518 3. Fixed a really odd bug that affected only the testing scheme; patching a
4519 certain fixed string in the binary changed the value of another string that
4520 happened to be identical to the end of the original first string.
4522 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4523 name", it returns that address as the IP address. On some operating
4524 systems (e.g. Solaris), it also passes back the IP address string as the
4525 "host name". However, on others (e.g. Linux), it passes back an empty
4526 string. Exim wasn't checking for this, and was changing the host name to an
4527 empty string, assuming it had been canonicized.
4529 5. Although rare, it is permitted to have more than one PTR record for a given
4530 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4531 all the names associated with an address, because they do in Solaris.
4532 However, it seems that they do not in Linux for data that comes from the
4533 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4534 I found this out when I moved to a new Linux workstation and tried to run
4535 the Exim test suite.
4537 To get round this problem I have changed the code so that it now does its
4538 own call to the DNS to look up PTR records when searching for a host name.
4539 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4540 addresses that are only in /etc/hosts are still found.
4542 This behaviour is, however, controlled by an option called host_lookup_
4543 order, which defaults to "bydns:byaddr". If people want to use the other
4544 order, or indeed, just use one or the other means of lookup, they can
4545 specify it in this variable.
4547 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4548 some operating systems, this comes back from gethostbyaddr() as an empty
4549 string, and this is what Exim used to test for. However, it seems that in
4550 other systems, "." is yielded. Exim now tests for this case too.
4552 7. The values of check_spool_space and check_log_space are now held internally
4553 as a number of kilobytes instead of an absolute number of bytes. If a
4554 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4555 kilobyte. This means that much larger values can be stored.
4557 8. Exim monitor: an attempt to get the action menu when not actually pointing
4558 at a message produces an empty menu entitled "No message selected". This
4559 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4560 no entries in it ("Shell widget menu has zero width and/or height"). So I
4561 have added a single, blank menu entry in this case.
4563 9. Added ${quote_local_part.
4565 10. MIME decoding is now applied to the contents of Subject: header lines when
4568 11. Now that a reference to $sender_host_address automatically causes a reverse
4569 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4570 host lookup before query-style lookups in lists that might use this
4571 variable. This has therefore been abolished, and the "net-" prefix is no
4572 longer necessary for query-style lookups.
4574 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4575 have been a typo for LFLAGS, so it has been changed.
4577 13. The install script calls Exim with "-C /dev/null" in order to find the
4578 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4579 to be output. Howeve, since Exim outputs its version number before the
4580 error, it didn't break the script. It just looked ugly. I fixed this by
4581 always allowing "-C /dev/null" if the caller is root.
4583 14. Ignore overlarge ACL variable number when reading spool file - insurance
4584 against a later release with more variables having written the file.
4586 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4587 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4588 this, as well as the form without the "IPv6" on the front.
4590 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4591 OS/Makefile-Darwin file.
4593 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4594 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4596 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4597 variable. However, some OS use ints for inodes. I've added cast to long int
4598 to get rid of the compiler warning.
4600 19. I had forgotten to lock out "/../" in configuration file names when
4601 ALT_CONFIG_PREFIX was set.
4603 20. Routers used for verification do not need to specify transports. However,
4604 if such a router generated a host list, and callout was configured, Exim
4605 crashed, because it could not find a port number from the (non-existent)
4606 transport. It now assumes port 25 in this circumstance.
4608 21. Added the -t option to exigrep.
4610 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4611 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4612 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4613 them is set, LOOKUP_LSEARCH is forced.
4615 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4617 24. Added sender and host information to the "rejected by local_scan()" log
4618 line; previously there was no indication of these.
4620 25. Added .include_if_exists.
4622 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4623 Linux. It turns out that not all file systems support this. Apparently some
4624 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4625 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4626 not supported on the file descriptor, is now ignored when Exim is trying to
4627 sync a directory. This applies only to Linux.
4629 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4631 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4632 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4635 29. Added support for maildirsize files from supplied patch (modified a bit).
4637 30. The use of :fail: followed by an empty string could lead Exim to respond to
4638 sender verification failures with (e.g.):
4640 550 Verification failed for <xxx>
4641 550 Sender verify failed
4643 where the first response line was missing the '-' that indicates it is not
4644 the final line of the response.
4646 31. The loop for finding the name of the user that called Exim had a hardwired
4647 limit of 10; it now uses the value of finduser_retries, which is used for
4648 all other user lookups.
4650 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4653 33. Exim was neglecting to zero errno before one call of strtol() when
4654 expanding a string and expecting an integer value. On some systems this
4655 resulted in spurious "integer overflow" errors. Also, it was casting the
4656 result into an int without checking.
4658 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4659 did not work. The code looks as if it has *never* worked, though it appears
4660 to have been documented since at least releast 1.62. I have made it work.
4662 35. The "timeout_DNS" error in retry rules, also documented since at least
4663 1.62, also never worked. As it isn't clear exactly what this means, and
4664 clearly it isn't a major issue, I have abolished the feature by treating it
4665 as "timeout", and writing a warning to the main and panic logs.
4667 36. The display of retry rules for -brt wasn't always showing the error code
4670 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4671 timeout_connect_A, timeout_connect_MX.
4673 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4674 to the empty sender.
4676 39. The daemon was not analysing the content of -oX till after it had closed
4677 stderr and disconnected from the controlling terminal. This meant that any
4678 syntax errors were only noted on the panic log, and the return code from
4679 the command was 0. By re-arranging the code a little, I've made the
4680 decoding happen first, so such errors now appear on stderr, and the return
4681 code is 1. However, the actual setting up of the sockets still happens in
4682 the disconnected process, so errors there are still only recorded on the
4685 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4686 connections (as happens on some IP stacks) was logged at start up time as
4687 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4688 it from "IPv6 and IPv4", which means that two separate sockets are being
4691 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4692 says whether AF_INET or AF_INET6 was passed as an argument.
4694 42. Exiwhat output was messed up when time zones were included in log
4697 43. Exiwhat now gives more information about the daemon's listening ports,
4698 and whether -tls-on-connect was used.
4700 44. The "port" option of the smtp transport is now expanded.
4702 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4703 silently ignored. Now an error message is written to the main and panic
4706 46. There's a new ACL modifier called "logwrite" which writes to a log file
4707 as soon as it is encountered.
4709 47. Added $local_user_uid and $local_user_gid at routing time.
4711 48. Exim crashed when trying to verify a sender address that was being
4714 49. Exim was recognizing only a space character after ".include". It now also
4715 recognizes a tab character.
4717 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4718 extracting the relevant information from the specification. The man page no
4719 longer contains scrambled data for the -d option, and I've added a section
4720 at the front about calling Exim under different names.
4722 51. Added "extra_headers" argument to the "mail" command in filter files.
4724 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4727 53. Installed eximstats 1.29.
4729 54. Added transport_filter_timeout as a generic transport option.
4731 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4732 Cc: header lines. This was required by RFC 822, but it not required by RFC
4735 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4736 incoming messages that did not have them. Now it does so only if the
4737 message originates locally, that is, if there is no associated remote host
4738 address. When Resent- header lines are present, this applies to the Resent-
4739 lines rather than the non-Resent- lines.
4741 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4742 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4744 58. Messages for configuration errors now include the name of the main
4745 configuration file - useful now that there may be more than one file in a
4746 list (.included file names were always shown).
4748 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4749 for those rare installations that do not start the daemon as root or run it
4750 setuid root. I've cut out the call to initgroups() if the daemon is not
4753 60. The Exim user and group can now be bound into the binary as text strings
4754 that are looked up at the start of Exim's processing.
4756 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4758 62. Added $mailstore_basename variable.
4760 63. Installed patch to sieve.c from Michael Haardt.
4762 64. When Exim failed to open the panic log after failing to open the main log,
4763 the original message it was trying to log was written to stderr and debug
4764 output, but if they were not available (the usual case in production), it
4765 was lost. Now it is written to syslog before the two lines that record the
4766 failures to open the logs.
4768 65. Users' Exim filters run in subprocesses under the user's uid. It is
4769 possible for a "deliver" command or an alias in a "personal" command to
4770 provoke an address rewrite. If logging of address rewriting is configured,
4771 this fails because the process is not running as root or exim. There may be
4772 a better way of dealing with this, but for the moment (because 4.30 needs
4773 to be released), I have disabled address rewrite logging when running a
4774 filter in a non-root, non-exim process.
4780 1. The buildconfig auxiliary program wasn't quoting the value set for
4781 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4782 not defined. This bug was masked in 4.22 by the effect that was fixed in
4785 2. Some messages that were rejected after a message id was allocated were
4786 shown as "incomplete" by exigrep. It no longer does this for messages that
4787 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4789 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4790 have allow_domain_literals set, the ID did not get logged in the <= line.
4791 Domain literals are now always recognized in Message-ID: header lines.
4793 4. The first argument for a ${extract expansion item is the key name or field
4794 number. Leading and trailing spaces in this item were not being ignored,
4795 causing some misleading effects.
4797 5. When deliver_drop_privilege was set, single queue runner processes started
4798 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4799 same command in the process it spins off) were not dropping privilege.
4801 6. When the daemon running as "exim" started a queue runner, it always
4802 re-executed Exim in the spun-off process. This is a waste of effort when
4803 deliver_drop_privilege is set. The new process now just calls the
4804 queue-runner function directly.
4810 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4813 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4814 ignored. Though the use of -oP was forcing the writing of a pid file, it
4815 was always written to the default place.
4817 3. If the message "no IP address found for host xxxx" is generated during
4818 incoming verification, it is now followed by identification of the incoming
4819 connection (so you can more easily find what provoked it).
4821 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4823 5. Added some features to "harden" Exim a bit more against certain attacks:
4825 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4826 be put in Local/Makefile. This is like the never_users runtime option,
4827 but it cannot be overridden. The default setting is "root".
4829 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4830 prefix string with which any file named in a -C command line option
4833 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4834 is retained for -C and -D only if the caller of Exim is root. Without
4835 it, the exim user may also use -C and -D and retain privilege.
4837 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4838 command line option is disabled.
4840 6. Macro names set by the -D option must start with an upper case letter, just
4841 like macro names defined in the configuration file.
4843 7. Added "dereference=" facility to LDAP.
4845 8. Two instances of the typo "uknown" in the source files are fixed.
4847 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4848 the Configure-Makefile script screwed up while processing it.
4850 10. Incorporated PCRE 4.4.
4852 11. The SMTP synchronization check was not operating right at the start of an
4853 SMTP session. For example, it could not catch a HELO sent before the client
4854 waited for the greeting. There is now a check for outstanding input at the
4855 point when the greeting is written. Because of the duplex, asynchronous
4856 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4857 way, but not yet received, when the check is performed.
4859 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4860 on TCP/IP sockets, because this apparently causes some broken clients to
4863 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4864 unchanged) from the Cygwin maintainer.
4866 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4867 was supported instead of testing for mbx. Effectively a typo.
4869 15. The spa authenticator server code was not checking that the input it
4870 received was valid base64.
4872 16. The debug output line for the "set" modifier in ACLs was not showing the
4873 name of the variable that was being set.
4875 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4877 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4878 lookup on its own. Something else was needed to trigger the lookup. For
4879 example, a match in host_lookup or the need for a host name in a host list.
4880 Now, if $sender_host_name is referenced and the host name has not yet been
4881 looked up, a lookup is performed. If the lookup fails, the variable remains
4882 empty, and $host_lookup_failed is set to "1".
4884 19. Added "eqi" as a case-independent comparison operator.
4886 20. The saslauthd authentication condition could segfault if neither service
4887 nor realm was specified.
4889 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4890 error message that was logged was misleading, and incoming SMTP
4891 connections were dropped. The message is now more accurate, and temporary
4892 errors are given to SMTP connections.
4894 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4895 (except RSET) with a 421 error, until QUIT is received. However, it was
4896 failing to send a response to QUIT.
4898 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4899 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4900 ACL is now run after the helo_try_verify_hosts code.
4902 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4903 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4904 case-independent, so other case variants are also recognized). Apparently
4905 some systems use these upper case variants.
4907 25. If more than two messages were waiting for the same host, and a transport
4908 filter was specified for the transport, Exim sent two messages over the
4909 same TCP/IP connection, and then failed with "socket operation on non-
4910 socket" when it tried to send the third.
4912 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4914 27. The extern definition of crypt16() in expand.c was not being excluded when
4915 the OS had its own crypt16() function.
4917 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4918 as a preferred synonym for return_size_limit, both as an option and as an
4921 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4923 30. Changed the default configuration ACL to relax the local part checking rule
4924 for addresses that are not in any local domains. For these addresses,
4925 slashes and pipe symbols are allowed within local parts, but the sequence
4926 /../ is explicitly forbidden.
4928 31. SPA server authentication was not clearing the challenge buffer before
4931 32. log_message in a "warn" ACL statement was writing to the reject log as
4932 well as to the main log, which contradicts the documentation and doesn't
4933 seem right (because no rejection is happening). So I have stopped it.
4935 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4936 However, I am unable to do any testing of this.
4938 34. Fixed an infelicity in the appendfile transport. When checking directories
4939 for a mailbox, to see if any needed to be created, it was accidentally
4940 using path names with one or more superfluous leading slashes; tracing
4941 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4943 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4944 contents are added to the log line that is written for every discarded
4945 recipient. (Previously a log_message setting was ignored.)
4947 36. The ${quote: operator now quotes the string if it is empty.
4949 37. The install script runs exim in order to find its version number. If for
4950 some reason other than non-existence or emptiness, which it checks, it
4951 could not run './exim', it was installing it with an empty version number,
4952 i.e. as "exim-". This error state is now caught, and the installation is
4955 38. An argument was missing from the function that creates an error message
4956 when Exim fails to connect to the socket for saslauthd authentication.
4957 This could cause Exim to crash, or give a corrupted message.
4959 39. Added isip, isip4, and isip6 to ${if conditions.
4961 40. The ACL variables $acl_xx are now saved with the message, and can be
4962 accessed later in routers, transports, and filters.
4964 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4965 strings in the file are not string-expanded.
4967 42. If a MAIL command specified a SIZE value that was too large to fit into an
4968 int variable, the check against message_size_limit failed. Such values are
4969 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4970 day this will have to be increased, but I don't think I want to be around
4971 when emails are that large.
4978 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4979 iconv() is not standard in FreeBSD.
4981 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4982 IPv6 enabled. The observed symptom was a segmentation fault on return from
4983 the function os_common_find_running_interfaces() in src/os.c.
4985 3. In the check_special_case() function in daemon.c I had used "errno" as an
4986 argument name, which causes warnings on some systems. This was basically a
4987 typo, since it was named "eno" in the comments!
4989 4. The code that waits for the clock to tick (at a resolution of some fraction
4990 of a second) so as to ensure message-id uniqueness was always waiting for
4991 at least one whole tick, when it could have waited for less. [This is
4992 almost certainly not relevant at current processor speeds, where it is
4993 unlikely to ever wait at all. But we try to future-proof.]
4995 5. The function that sleeps for a time interval that includes fractions of a
4996 second contained a race. It did not block SIGALRM between setting the
4997 timer, and suspending (a couple of lines later). If the interval was short
4998 and the sigsuspend() was delayed until after it had expired, the suspension
4999 never ended. On busy systems this could lead to processes getting stuck for
5002 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5003 process, before it forks any delivery processes. The open lookup caching
5004 mechanism meant that the open file or database connection was passed into
5005 the delivery process. The problem was that delivery processes always tidy
5006 up cached lookup data. This could cause a problem for the next delivery
5007 process started by the queue runner, because the external queue runner
5008 process does not know about the closure. So the next delivery process
5009 still has data in the lookup cache. In the case of a file lookup, there was
5010 no problem because closing a file descriptor in a subprocess doesn't affect
5011 the parent. However, if the lookup was caching a connection to a database,
5012 the connection was closed, and the second delivery process was likely to
5013 see errors such as "PGSQL: query failed: server closed the connection
5014 unexpectedly". The problem has been fixed by closing all cached lookups
5015 in a queue runner before running a delivery process.
5017 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5018 seem to have the "const" qualifier which it has on other OS. I've
5021 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5022 *of the same type* (client or server) with the same public name that an
5023 error should be diagnosed.
5025 9. When Exim looked up a host name for an IP address, but failed to find the
5026 original IP address when looking up the host name (a safety check), it
5027 output the message "<ip address> does not match any IP for NULL", which was
5028 confusing, to say the least. The bug was that the host name should have
5029 appeared instead of "NULL".
5031 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5032 user that is built into the binary, and the -C or -D options is used, root
5033 privilege is dropped before the configuration file is read. In addition,
5034 logging is switched to stderr instead of the normal log files. If the
5035 configuration then re-defines the Exim user, the unprivileged environment
5036 is probably not what is expected, so Exim logs a panic warning message (but
5039 However, if deliver_drop_privilege is set, the unprivileged state may well
5040 be exactly what is intended, so the warning has been cut out in that case,
5041 and Exim is allowed to try to write to its normal log files.
5047 1. smtp_return_error_details was not giving details for temporary sender
5048 or receiver verification errors.
5050 2. Diagnose a configuration error if two authenticators have the same public
5053 3. Exim used not to create the message log file for a message until the first
5054 delivery attempt. This could be confusing when incoming messages were held
5055 for policy or load reasons. The message log file is now created at the time
5056 the message is received, and an initial "Received" line is written to it.
5058 4. The automatically generated man page for command line options had a minor
5059 bug that caused no ill effects; however, a more serious problem was that
5060 the procedure for building the man page automatically didn't always
5061 operate. Consequently, release 4.20 contains an out-of-date version. This
5062 shouldn't happen again.
5064 5. When building Exim with embedded Perl support, the script that builds the
5065 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5066 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5068 6. The freeze_tell option was not being used for messages that were frozen on
5069 arrival, either by an ACL or by local_scan().
5071 7. Added the smtp_incomplete_transaction log selector.
5073 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5074 was accepting AUTH without a new EHLO.
5076 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5077 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5080 10. Logging of TCP/IP connections (when configured) now happens in the main
5081 daemon process instead of the child process, so that the TCP/IP connection
5082 count is more accurate (but it can never be perfect).
5084 11. The use of "drop" in a nested ACL was not being handled correctly in the
5085 outer ACL. Now, if condition failure induced by the nested "drop" causes
5086 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5087 or "require"), the connection is dropped.
5089 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5090 that yield "discard" can now be used with an "accept" or a "discard" verb,
5091 but an error is generated for any others (because I can't see a useful way
5092 to define what should happen).
5094 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5095 are now processed in the same way as lines in the Exim configuration file.
5096 In particular, continuation lines are supported.
5098 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5100 15. Added -ti meaning -t -i.
5102 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5103 domains, and warn by logging if others are found.
5105 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5106 returned from GIFCONF when Exim is trying to find the list of interfaces on
5107 a host. The code in os.c has been modified to copy each ifreq to an aligned
5108 structure in all cases.
5110 Also, in some cases, the returned ifreq's were being copied to a 'struct
5111 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5112 means the last couple of bytes of an IPv6 address could be chopped if the
5113 ifreq contained only a normal sockaddr (14 bytes storage).
5115 18. Named domain lists were not supported in the hosts_treat_as_local option.
5116 An entry such as +xxxx was not recognized, and was treated as a literal
5119 19. Ensure that header lines added by a DATA ACL are included in the reject log
5120 if the ACL subsequently rejects the message.
5122 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5123 MD5 (which is deprecated).
5125 21. When testing a filter file using -bf, Exim was writing a message when it
5126 took the sender from a "From " line in the message, but it was not doing so
5127 when it took $return_path from a Return-Path: header line. It now does.
5129 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5130 with a valid header line field name (a series of printing characters
5131 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5133 23. Changed "disc" in the source to "disk" to conform to the documentation and
5134 the book and for uniformity.
5136 24. Ignore Sendmail's -Ooption=value command line item.
5138 25. When execve() failed while trying to run a command in a pipe transport,
5139 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5140 could be confused with a return value of 69 from the command itself. This
5141 has been changed to 127, the value the shell returns if it is asked to run
5142 a non-existent command. The wording for the related log line suggests a
5143 non-existent command as the problem.
5145 26. If received_header_text expands to an empty string, do not add a Received:
5146 header line to the message. (Well, it adds a token one on the spool, but
5147 marks it "old" so that it doesn't get used or transmitted.)
5149 27. Installed eximstats 1.28 (addition of -nt option).
5151 28. There was no check for failure on the call to getsockname() in the daemon
5152 code. This can fail if there is a shortage of resources on the system, with
5153 ENOMEM, for example. A temporary error is now given on failure.
5155 29. Contrary to the C standard, it seems that in some environments, the
5156 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5157 program. Exim now does this explicitly; it affects the formatting of
5158 timestamps using strftime().
5160 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5161 complaints. I've now initialized all the variables, to avoid this.
5163 32. Header lines added by a system filter were not being "seen" during
5164 transport-time rewrites.
5166 33. The info_callback() function passed to OpenSSL is set up with type void
5167 (*)(SSL *, int, int), as described somewhere. However, when calling the
5168 function (actually a macro) that sets it up, the type void(*)() is
5169 expected. I've put in a cast to prevent warnings from picky compilers.
5171 34. If a DNS black list lookup found a CNAME record, but there were no A
5172 records associated with the domain it pointed at, Exim crashed.
5174 35. If a DNS black list lookup returned more than one A record, Exim ignored
5175 all but the first. It now scans all returned addresses if a particular IP
5176 value is being sought. In this situation, the contents of the
5177 $dnslist_value variable are a list of all the addresses, separated by a
5180 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5181 to accept a host name and all its aliases if the forward lookup for any of
5182 them yielded the IP address of the incoming connection. Now it accepts only
5183 those names whose forward lookup yields the correct IP address. Any other
5184 names are discarded. This closes a loophole whereby a rogue DNS
5185 administrator could create reverse DNS records to break through a
5186 wildcarded host restriction in an ACL.
5188 37. If a user filter or a system filter that ran in a subprocess used any of
5189 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5190 the wrong values were passed to the pipe command ($thisaddress had the
5191 value of $0, $0 had the value of $1, etc). This bug was introduced by
5192 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5194 38. Improved the line breaking for long SMTP error messages from ACLs.
5195 Previously, if there was no break point between 40 and 75 characters, Exim
5196 left the rest of the message alone. Two changes have been made: (a) I've
5197 reduced the minimum length to 35 characters; (b) if it can't find a break
5198 point between 35 and 75 characters, it looks ahead and uses the first one
5199 that it finds. This may give the occasional overlong line, but at least the
5200 remaining text gets split now.
5202 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5203 file descriptors might be low, and that setting 1000 would always raise it.
5204 It turns out that in some environments, the limit is already over 1000 and
5205 that lowering it causes trouble. So now Exim takes care not to decrease it.
5207 40. When delivering a message, the value of $return_path is set to $sender_
5208 address at the start of routing (routers may change the value). By an
5209 oversight, this default was not being set up when an address was tested by
5210 -bt or -bv, which affected the outcome if any router or filter referred to
5213 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5214 log only when "message" or "log_message" are set. However, if one of the
5215 conditions was an address verification, or a call to a nested ACL, the
5216 messages generated by the underlying test were being passed through. This
5217 no longer happens. The underlying message is available in $acl_verify_
5218 message for both "message" and "log_message" expansions, so it can be
5219 passed through if needed.
5221 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5222 new expansion $bh_ to give the encoded byte string without charset
5223 translation. Translation happens only if iconv() is available; HAVE_ICONV
5224 indicates this at build time. HEADERS_CHARSET gives the charset to
5225 translate to; headers_charset can change it in the configuration, and
5226 "headers charset" can change it in an individual filter file.
5228 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5229 that creates RFC 2047 encoded "words" labels them as that charset instead
5230 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5231 expansion operator; (ii) when Exim creates a From: line for a local
5232 message; (iii) when a header line is rewritten to include a "phrase" part.
5234 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5235 buggy, causing it to skip the first lines of messages whose message ID
5236 ended in 'D'. This would not have bitten before Exim release 4.14, because
5237 message IDs were unlikely to end in 'D' before then. The effect was to have
5238 incorrect size information for certain domains.
5240 45. #include "config.h" was missing at the start of the crypt16.c module. This
5241 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5244 46. If there was a timeout during a "random" callout check, Exim treated it as
5245 a failure of the random address, and carried on sending RSET and the real
5246 address. If the delay was just some slowness somewhere, the response to the
5247 original RCPT would be taken as a response to RSET and so on, causing
5248 mayhem of various kinds.
5250 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5251 when I implemented it. It didn't allow for the fact that some option values
5252 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5253 the right test for positive values.
5255 48. Domain names in DNS records are case-independent. Exim always looks them up
5256 in lower case. Some resolvers return domain names in exactly the case they
5257 appear in the zone file, that is, they may contain uppercase letters. Not
5258 all resolvers do this - some return always lower case. Exim was treating a
5259 change of case by a resolver as a change of domain, similar to a widening
5260 of a domain abbreviation. This triggered its re-routing code and so it was
5261 trying to route what was effectively the same domain again. This normally
5262 caused routing to fail (because the router wouldn't handle the domain
5263 twice). Now Exim checks for this case specially, and just changes the
5264 casing of the domain that it ultimately uses when it transmits the message
5267 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5270 50. If a filter generated a file delivery with a non-absolute name (possible if
5271 no home directory exists for the router), the forbid_file option was not
5274 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5275 the existing equality matching.
5277 52. Exim was using ints instead of ino_t variables in some places where it was
5278 dealing with inode numbers.
5280 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5281 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5282 called TMPDIR, and if it finds it is different, it changes its value.
5284 54. The smtp_printf() function is now made available to local_scan() so
5285 additional output lines can be written before returning. There is also an
5286 smtp_fflush() function to enable the detection of a dropped connection.
5287 The variables smtp_input and smtp_batched_input are exported to
5290 55. Changed the default runtime configuration: the message "Unknown user"
5291 has been removed from the ACL, and instead placed on the localuser router,
5292 using the cannot_route_message feature. This means that any verification
5293 failures that generate their own messages won't get overridden. Similarly,
5294 the "Unrouteable address" message that was in the ACL for unverifiable
5295 relay addresses has also been removed.
5297 56. Added hosts_avoid_esmtp to the smtp transport.
5299 57. The exicyclog script was not checking for the esoteric option
5300 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5301 will work only if exicyclog is run under the appropriate euid.
5303 58. Following a discussion on the list, the rules by which Exim recognises line
5304 endings on incoming messages have been changed. The -dropcr and drop_cr
5305 options are now no-ops, retained only for backwards compatibility. The
5306 following line terminators are recognized: LF CRLF CR. However, special
5307 processing applies to CR:
5309 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5310 nor a local message in the state where . is a terminator.
5312 (ii) If a bare CR is encountered in a header line, an extra space is added
5313 after the line terminator so as not to end the header. The reasoning
5314 behind this is that bare CRs in header lines are most likely either
5315 to be mistakes, or people trying to play silly games.
5317 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5318 was being incorrectly given as 18 bytes larger than it should have been.
5319 This is a VOB (very old bug).
5321 60. This may never have affected anything current, but just in case it has:
5322 When the local host is found other than at the start of a list of hosts,
5323 the local host, those with the same MX, and any that follow, are discarded.
5324 When the list in question was part of a longer list of hosts, the following
5325 hosts (not currently being processed) were also being discarded. This no
5326 longer happens. I'm not sure if this situation could ever has previously
5329 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5332 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5333 header. This is something that is recommended in a new Internet Draft, and
5334 is something that is documented as being done by Sendmail. There are two
5335 possible values. For messages generated by the autoreply transport, Exim
5338 Auto-Submitted: auto-replied
5340 whereas for all other generated messages (e.g. bounces) it adds
5342 Auto-Submitted: auto-generated
5344 63. The "personal" condition in filters now includes a test for the
5345 Auto-Submitted: header. If it contains the string "auto-" the message it
5346 not considered personal.
5348 64. Added rcpt_include_affixes as a generic transport option.
5350 65. Added queue_only_override (default true).
5352 66. Added the syslog_duplication option.
5354 67. If what should have been the first header line of a message consisted of
5355 a space followed by a colon, Exim was mis-interpreting it as a header line.
5356 It isn't of course - it is syntactically invalid and should therefore be
5357 treated as the start of the message body. The misbehaviour could have
5358 caused a number of strange effects, including loss of data in subsequent
5359 header lines, and spool format errors.
5361 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5362 client host had authenticated. This control can now be exercised by an ACL
5363 for more flexibility.
5365 69. By default, callouts do not happen when testing with -bh. There is now a
5366 variant, -bhc, which does actually run the callout code, including
5367 consulting and updating the callout cache.
5369 70. Added support for saslauthd authentication, courtesy of Alexander
5372 71. If statvfs() failed on the spool or log directories while checking their
5373 size for availability, Exim confusingly gave the error "space shortage".
5374 Furthermore, in debugging mode it crashed with a floating point exception.
5375 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5376 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5377 serious problem, Exim now writes to the main and panic logs when this
5378 happens, with details of the failure. It then refuses to accept the
5379 incoming message, giving the message "spool directory problem" or "log
5380 directory problem" with a 421 code for SMTP messages.
5382 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5383 0, 1, and 2 exist, because some OS complain for execs without them (see
5384 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5385 descriptors. However, the code omitted to check that the open succeeded,
5386 causing mysterious errors if for some reason the permissions on /dev/null
5387 got screwed. Now Exim writes a message to the main and panic logs, and
5388 bombs out if it can't open /dev/null.
5390 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5391 interact so that it is all more flexible. It is supposed to remain
5392 backwards compatible. Also added extra_local_interfaces.
5394 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5395 to bomb out with an assertion failure - to the client this appears as a
5396 connection drop. This problem occurs in the part of the code that was taken
5397 from the Samba project. Fortunately, the assertion is in a very simple
5398 function, so I have fixed this by reproducing the function inline in the
5399 one place where it is called, and arranging for authentication to fail
5400 instead of killing the process with assert().
5402 75. The SPA client code was not working when the server requested OEM rather
5403 than Unicode encoding.
5405 76. Added code to make require_files with a specific uid setting more usable in
5406 the case where statting the file as root fails - usually a non-root-mounted
5407 NFS file system. When this happens and the failure is EACCES, Exim now
5408 forks a subprocess and does the per-uid checking as the relevant uid.
5410 77. Added process_log_path.
5412 78. If log_file_path was not explicitly set, a setting of check_log_space or
5413 check_log_inodes was ignored.
5415 79. If a space check for the spool or log partitions fails, the incident is now
5416 logged. Of course, in the latter case the data may get lost...
5418 80. Added the %p formatting code to string_format() so that it can be used to
5419 print addresses in debug_print(). Adjusted all the address printing in the
5420 debugging in store.c to use %p rather than %d.
5422 81. There was a concern that a line of code in smtp_in.c could overflow a
5423 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5424 initially expressed, the concern was not well-founded, because trailing
5425 spaces are removed early. However, if the trailing spaces were followed by
5426 a NULL, they did not get removed, so the overflow was possible. Two fixes
5429 (a) I re-wrote the offending code in a cleaner fashion.
5430 (b) If an incoming SMTP command contains a NULL character, it is rejected
5433 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5434 runs initgroups(), so that if the Exim user is in any additional groups,
5435 they will be used during message reception.
5441 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob