1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 NM/01 Workround for PCRE version dependancy in version reporting
12 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
13 This fixes portability to compilers other than gcc, notably
14 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
16 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
17 makefiles for portability to HP-UX and POSIX correctness.
19 PP/01 Permit LOOKUP_foo enabling on the make command-line.
20 Also via indented variable definition in the Makefile.
21 (Debugging by Oliver Heesakkers).
23 PP/02 Restore caching of spamd results with expanded spamd_address.
24 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
26 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
27 Improves build reliability. Fix from: Frank Elsner
29 NM/02 Fix wide character breakage in the rfc2047 coding
30 Fixes bug 1064. Patch from Andrey N. Oktyabrski
32 NM/03 Allow underscore in dnslist lookups
33 Fixes bug 1026. Patch from Graeme Fowler
35 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
36 Code patches from Adam Ciarcinski of NetBSD.
38 NM/04 Fixed exiqgrep to cope with mailq missing size issue
41 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
42 is logged, to avoid truncation. Patch from John Horne.
44 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
45 Patch from Jakob Hirsch.
47 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
48 SQL string expansion failure details.
49 Patch from Andrey Oktyabrski.
55 TF/01 Failure to get a lock on a hints database can have serious
56 consequences so log it to the panic log.
58 TF/02 Log LMTP confirmation messages in the same way as SMTP,
59 controlled using the smtp_confirmation log selector.
61 TF/03 Include the error message when we fail to unlink a spool file.
63 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
64 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
65 for maintaining out-of-tree patches for some time.
67 PP/01 Bugzilla 139: Documentation and portability issues.
68 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
69 Handle per-OS dynamic-module compilation flags.
71 PP/02 Let /dev/null have normal permissions.
72 The 4.73 fixes were a little too stringent and complained about the
73 permissions on /dev/null. Exempt it from some checks.
74 Reported by Andreas M. Kirchwitz.
76 PP/03 Report version information for many libraries, including
77 Exim version information for dynamically loaded libraries. Created
78 version.h, now support a version extension string for distributors
79 who patch heavily. Dynamic module ABI change.
81 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
82 privilege escalation vulnerability whereby the Exim run-time user
83 can cause root to append content of the attacker's choosing to
86 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
89 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
90 If dropping privileges for untrusted macros, we disabled normal logging
91 on the basis that it would fail; for the Exim run-time user, this is not
92 the case, and it resulted in successful deliveries going unlogged.
93 Fixed. Reported by Andreas Metzler.
99 PP/01 Date: & Message-Id: revert to normally being appended to a message,
100 only prepend for the Resent-* case. Fixes regression introduced in
101 Exim 4.70 by NM/22 for Bugzilla 607.
103 PP/02 Include check_rfc2047_length in configure.default because we're seeing
104 increasing numbers of administrators be bitten by this.
106 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
108 PP/03 Bugzilla 994: added openssl_options main configuration option.
110 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
112 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
114 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
116 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
117 without a peer certificate, leading to a segfault because of an
118 assumption that peers always have certificates. Be a little more
119 paranoid. Problem reported by Martin Tscholak.
121 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
122 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
123 NB: ClamAV planning to remove STREAM in "middle of 2010".
124 CL also introduces -bmalware, various -d+acl logging additions and
125 more caution in buffer sizes.
127 PP/09 Implemented reverse_ip expansion operator.
129 PP/10 Bugzilla 937: provide a "debug" ACL control.
131 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
133 PP/12 Bugzilla 973: Implement --version.
135 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
137 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
139 PP/15 Bugzilla 816: support multiple condition rules on Routers.
141 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
142 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
143 ignore trailing whitespace.
145 JJ/02 prevent non-panic DKIM error from being sent to paniclog
147 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
150 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
151 Notification from Dr Andrew Aitchison.
153 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
154 ExtendedDetectionInfo response format.
155 Notification from John Horne.
157 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
160 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
161 XSL and documented dependency on system catalogs, with examples of how
164 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
167 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
168 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
169 configuration file which is writeable by the Exim user or group.
171 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
172 of configuration files to cover files specified with the -C option if
173 they are going to be used with root privileges, not just the default
176 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
177 option (effectively making it always true).
179 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
180 files to be used while preserving root privileges.
182 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
183 that rogue child processes cannot use them.
185 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
186 run-time user, instead of root.
188 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
189 Exim run-time user without dropping privileges.
191 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
192 result string, instead of calling string_vformat() twice with the same
195 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
196 for other users. Others should always drop root privileges if they use
197 -C on the command line, even for a whitelisted configure file.
199 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
201 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
207 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
208 $data_path, and $header_path variables; fixed documentation bugs and
211 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
212 exipick to access non-standard spools, including the "frozen" queue
215 NM/01 Bugzilla 965: Support mysql stored procedures.
216 Patch from Alain Williams
218 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
220 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
221 Patch from Andreas Metzler
223 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
224 Patch from Kirill Miazine
226 NM/05 Bugzilla 671: Added umask to procmail example.
228 JJ/03 installed exipick 20100323.0, fixing doc bug
230 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
231 directory. Notification and patch from Dan Rosenberg.
233 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
235 TK/02 Improve log output when DKIM signing operation fails.
237 MH/01 Treat the transport option dkim_domain as a colon separated
238 list, not as a single string, and sign the message with each element,
239 omitting multiple occurences of the same signer.
241 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
242 Bugzilla 985, 986. Patch by Simon Arlott
244 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
245 Patch by Simon Arlott
247 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
248 MBX locking. Notification from Dan Rosenberg.
254 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
256 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
258 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
260 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
262 NM/04 Bugzilla 915: Flag broken perl installation during build.
268 TK/01 Added patch by Johannes Berg that expands the main option
269 "spamd_address" if it starts with a dollar sign.
271 TK/02 Write list of recipients to X-Envelope-Sender header when building
272 the mbox-format spool file for content scanning (suggested by Jakob
275 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
276 (http://www.dcc-servers.net/) support via dccifd. Activated by
277 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
279 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
280 by Mark Daniel Reidel <mr@df.eu>.
282 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
283 When building exim an external PCRE library is now needed -
284 PCRE is a system library on the majority of modern systems.
285 See entry on PCRE_LIBS in EDITME file.
287 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
288 conversation. Added nologin parameter to request.
289 Patch contributed by Kirill Miazine.
291 TF/01 Do not log submission mode rewrites if they do not change the address.
293 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
295 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
296 log files in place. Contributed by Roberto Lima.
298 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
300 TF/03 Bugzilla 615: When checking the local_parts router precondition
301 after a local_part_suffix or local_part_prefix option, Exim now
302 does not use the address's named list lookup cache, since this
303 contains cached lookups for the whole local part.
305 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
306 Robert Millan. Documentation is in experimental-spec.txt.
308 TF/04 Bugzilla 668: Fix parallel build (make -j).
310 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
312 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
313 Patch provided by Jan Srzednicki.
315 TF/05 Leading white space used to be stripped from $spam_report which
316 wrecked the formatting. Now it is preserved.
318 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
319 that they are available at delivery time.
321 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
323 TF/08 TLS error reporting now respects the incoming_interface and
324 incoming_port log selectors.
326 TF/09 Produce a more useful error message if an SMTP transport's hosts
327 setting expands to an empty string.
329 NM/06 Bugzilla 744: EXPN did not work under TLS.
330 Patch provided by Phil Pennock.
332 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
333 Patch provided by Richard Godbee.
335 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
336 acl_smtp_notquit, added index entry.
338 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
339 Patch provided by Eugene Bujak.
341 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
342 accept(). Patch provided by Maxim Dounin.
344 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
345 Patch provided by Phil Pennock.
347 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
349 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
350 Patch provided by Brad "anomie" Jorsch.
352 NM/14 Bugzilla 622: Added timeout setting to transport filter.
353 Patch provided by Dean Brooks.
355 TK/05 Add native DKIM support (does not depend on external libraries).
357 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
358 Patch provided by Graeme Fowler.
360 NM/16 Bugzilla 851: Documentation example syntax fix.
362 NM/17 Changed NOTICE file to remove references to embedded PCRE.
364 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
367 NM/19 Bugzilla 745: TLS version reporting.
368 Patch provided by Phil Pennock.
370 NM/20 Bugzilla 167: bool: condition support.
371 Patch provided by Phil Pennock.
373 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
374 clients. Patch provided by Phil Pennock.
376 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
377 Patch provided by Brad "anomie" Jorsch.
379 NM/23 Bugzilla 687: Fix misparses in eximstats.
380 Patch provided by Heiko Schlittermann.
382 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
383 Patch provided by Heiko Schlittermann.
385 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
386 plus update to original patch.
388 NM/26 Bugzilla 799: Documentation correction for ratelimit.
390 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
391 Patch provided by David Brownlee.
393 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
395 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
397 NM/30 Bugzilla 888: TLS documentation bugfixes.
399 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
401 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
402 Unlike the original bugzilla I have changed all shell scripts in src tree.
404 NM/33 Bugzilla 898: Transport filter timeout fix.
405 Patch by Todd Rinaldo.
407 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
408 Patch by Serge Demonchaux.
410 NM/35 Bugzilla 39: Base64 decode bug fixes.
411 Patch by Jakob Hirsch.
413 NM/36 Bugzilla 909: Correct connect() call in dcc code.
415 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
417 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
419 NM/39 Bugzilla 911: Fixed MakeLinks build script.
425 TK/01 Add preliminary DKIM support. Currently requires a forked version of
426 ALT-N's libdkim that I have put here:
427 http://duncanthrax.net/exim-experimental/
429 Note to Michael Haardt: I had to rename some vars in sieve.c. They
430 were called 'true' and it seems that C99 defines that as a reserved
431 keyword to be used with 'bool' variable types. That means you could
432 not include C99-style headers which use bools without triggering
433 build errors in sieve.c.
435 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
436 as mailq or other aliases. Changed the --help handling significantly
437 to do whats expected. exim_usage() emits usage/help information.
439 SC/01 Added the -bylocaldomain option to eximstats.
441 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
443 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
445 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
451 PH/01 Another patch from the Sieve maintainer.
453 PH/02 When an IPv6 address is converted to a string for single-key lookup
454 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
455 dots are used instead of colons so that keys in lsearch files need not
456 contain colons. This was done some time before quoting was made available
457 in lsearch files. However, iplsearch files do require colons in IPv6 keys
458 (notated using the quote facility) so as to distinguish them from IPv4
459 keys. This meant that lookups for IP addresses in host lists did not work
460 for iplsearch lookups.
462 This has been fixed by arranging for IPv6 addresses to be expressed with
463 colons if the lookup type is iplsearch. This is not incompatible, because
464 previously such lookups could never work.
466 The situation is now rather anomolous, since one *can* have colons in
467 ordinary lsearch keys. However, making the change in all cases is
468 incompatible and would probably break a number of configurations.
470 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
473 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
474 conversion specification without a maximum field width, thereby enabling
475 a rogue spamd server to cause a buffer overflow. While nobody in their
476 right mind would setup Exim to query an untrusted spamd server, an
477 attacker that gains access to a server running spamd could potentially
478 exploit this vulnerability to run arbitrary code as the Exim user.
480 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
481 $primary_hostname instead of what libspf2 thinks the hosts name is.
483 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
484 a directory entry by the name of the lookup key. Previously, if a
485 symlink pointed to a non-existing file or a file in a directory that
486 Exim lacked permissions to read, a lookup for a key matching that
487 symlink would fail. Now it is enough that a matching directory entry
488 exists, symlink or not. (Bugzilla 503.)
490 PH/03 The body_linecount and body_zerocount variables are now exported in the
493 PH/04 Added the $dnslist_matched variable.
495 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
496 This means they are set thereafter only if the connection becomes
499 PH/06 Added the client_condition to authenticators so that some can be skipped
500 by clients under certain conditions.
502 PH/07 The error message for a badly-placed control=no_multiline_responses left
503 "_responses" off the end of the name.
505 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
507 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
508 (without spaces) instead of just copying the configuration text.
510 PH/10 Added the /noupdate option to the ratelimit ACL condition.
512 PH/11 Added $max_received_linelength.
514 PH/12 Added +ignore_defer and +include_defer to host lists.
516 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
517 way in which PCRE > 7.0 is built.
519 PH/14 Implemented queue_only_load_latch.
521 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
522 MAIL command. The effect was to mangle the value on 64-bit systems.
524 PH/16 Another patch from the Sieve maintainer.
526 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
528 PH/18 If a system quota error occurred while trying to create the file for
529 a maildir delivery, the message "Mailbox is full" was not appended to the
530 bounce if the delivery eventually timed out. Change 4.67/27 below applied
531 only to a quota excession during the actual writing of the file.
533 PH/19 It seems that peer DN values may contain newlines (and other non-printing
534 characters?) which causes problems in log lines. The DN values are now
535 passed through string_printing() before being added to log lines.
537 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
538 and InterBase are left for another time.)
540 PH/21 Added message_body_newlines option.
542 PH/22 Guard against possible overflow in moan_check_errorcopy().
544 PH/23 POSIX allows open() to be a macro; guard against that.
546 PH/24 If the recipient of an error message contained an @ in the local part
547 (suitably quoted, of course), incorrect values were put in $domain and
548 $local_part during the evaluation of errors_copy.
554 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
555 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
558 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
559 issue a MAIL command.
561 PH/02 In an ACL statement such as
563 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
565 if a client was not listed at all, or was listed with a value other than
566 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
567 the condition was not true (as it should be), so access was not denied.
568 The bug was that the ! inversion was incorrectly passed on to the second
569 item. This has been fixed.
571 PH/03 Added additional dnslists conditions == and =& which are different from
572 = and & when the dns lookup returns more than one IP address.
574 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
575 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
577 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
578 FSYNC, which compiles an option called disable_fsync that allows for
579 bypassing fsync(). The documentation is heavily laced with warnings.
581 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
583 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
584 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
585 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
586 including adding "make clean"; (3) Added -fPIC when compiling the test
587 dynamically loaded module, to get rid of a warning.
589 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
590 message fails, move_frozen_messages = true and ignore_bounce_errors_after
591 = 0s. The bug is otherwise harmless.
593 PH/07 There was a bug in the dovecot authenticator such that the value of
594 $auth1 could be overwritten, and so not correctly preserved, after a
595 successful authentication. This usually meant that the value preserved by
596 the server_setid option was incorrect.
598 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
600 PH/09 Installed PCRE release 7.0.
602 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
603 run for batched SMTP input. It is now run at the start of every message
604 in the batch. While fixing this I discovered that the process information
605 (output by running exiwhat) was not always getting set for -bs and -bS
606 input. This is fixed, and it now also says "batched" for BSMTP.
608 PH/11 Added control=no_pipelining.
610 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
611 patch, slightly modified), and move the expansion of helo_data till after
612 the connection is made in the smtp transport (so it can use these
615 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
617 PH/14 Added log_selector = +pid.
619 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
621 PH/16 Add ${if forany and ${if forall.
623 PH/17 Added dsn_from option to vary the From: line in DSNs.
625 PH/18 Flush SMTP output before performing a callout, unless control =
626 no_callout_flush is set.
628 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
629 was true (the default) a successful delivery failed to delete the retry
630 item, thus causing premature timeout of the address. The bug is now
633 PH/20 Added hosts_avoid_pipelining to the smtp transport.
635 PH/21 Long custom messages for fakedefer and fakereject are now split up
636 into multiline reponses in the same way that messages for "deny" and
637 other ACL rejections are.
639 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
640 with slight modification.
642 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
643 draft, changing the syntax and factoring some duplicate code".
645 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
646 for deliveries of the second and subsequent messages over the same SMTP
649 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
650 ${reduce, with only minor "tidies".
652 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
654 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
655 expansion side effects.
657 PH/27 When a message times out after an over-quota error from an Exim-imposed
658 quota, the bounce message says "mailbox is full". This message was not
659 being given when it was a system quota that was exceeded. It now should
662 MH/03 Made $recipients available in local_scan(). local_scan() already has
663 better access to the recipient list through recipients_list[], but
664 $recipients can be useful in postmaster-provided expansion strings.
666 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
667 in the case of a MAIL command with additional options following the
668 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
669 were accidentally chopped off.
671 PH/29 SMTP synchronization checks are implemented when a command is read -
672 there is a check that no more input is waiting when there shouldn't be
673 any. However, for some commands, a delay in an ACL can mean that it is
674 some time before the response is written. In this time, more input might
675 arrive, invalidly. So now there are extra checks after an ACL has run for
676 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
677 pipelining has not been advertised.
679 PH/30 MH's patch to allow iscntrl() characters to be list separators.
681 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
682 returned in the SMTP response when smtp_return_error_details was false.
685 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
686 instead of the C I/O that was originally supplied, because problems were
689 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
690 Exim which did not show up earlier: it was assuming that a call to
691 SSL_CTX_set_info_callback() might give an error value. In fact, there is
692 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
693 was a macro that became an assignment, so it seemed to work. This has
694 changed to a proper function call with a void return, hence the compile
695 error. Exim's code has been fixed.
697 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
700 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
702 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
704 PH/37 If a message is not accepted after it has had an id assigned (e.g.
705 because it turns out to be too big or there is a timeout) there is no
706 "Completed" line in the log. When some messages of this type were
707 selected by exigrep, they were listed as "not completed". Others were
708 picked up by some special patterns. I have improved the selection
709 criteria to be more general.
711 PH/38 The host_find_failed option in the manualroute router can now be set
712 to "ignore", to completely ignore a host whose IP address cannot be
713 found. If all hosts are ignored, the behaviour is controlled by the new
714 host_all_ignored option.
716 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
717 homing or because of multiple MX records with /mx) generated more than
718 one IP address, and the following item turned out to be the local host,
719 all the secondary addresses of the first item were incorrectly removed
720 from the list, along with the local host and any following hosts (which
721 is what is supposed to happen).
723 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
724 whoever called Exim into the -H file. In the case of the daemon it was
725 behaving confusingly. When first started, it used values for whoever
726 started the daemon, but after a SIGHUP it used the Exim user (because it
727 calls itself on a restart). I have changed the code so that it now always
730 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
731 message are rejected with the same error (e.g. no authentication or bad
732 sender address), and a DATA command is nevertheless sent (as can happen
733 with PIPELINING or a stupid MUA), the error message that was given to the
734 RCPT commands is included in the rejection of the DATA command. This is
735 intended to be helpful for MUAs that show only the final error to their
738 PH/42 Another patch from the Sieve maintainer.
740 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
741 Eximstats - Fixed some broken HTML links and added missing column headers
743 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
744 columns for Rejects, Temp Rejects, Ham, and Spam rows.
746 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
748 PH/43 Yet another patch from the Sieve maintainer.
750 PH/44 I found a way to check for a TCP/IP connection going away before sending
751 the response to the final '.' that terminates a message, but only in the
752 case where the client has not sent further data following the '.'
753 (unfortunately, this is allowed). However, in many cases there won't be
754 any further data because there won't be any more messages to send. A call
755 to select() can be used: if it shows that the input is "ready", there is
756 either input waiting, or the socket has been closed. An attempt to read
757 the next input character can distinguish the two cases. Previously, Exim
758 would have sent an OK response which the client would never have see.
759 This could lead to message repetition. This fix should cure that, at
760 least in a lot of common cases.
762 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
763 advertised in response to EHLO.
769 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
770 fixed by 4.65/MH/01 (is this a record?) are fixed:
772 (i) An empty string was always treated as zero by the numeric comparison
773 operators. This behaviour has been restored.
775 (ii) It is documented that the numeric comparison operators always treat
776 their arguments as decimal numbers. This was broken in that numbers
777 starting with 0 were being interpreted as octal.
779 While fixing these problems I realized that there was another issue that
780 hadn't been noticed. Values of message_size_limit (both the global option
781 and the transport option) were treated as octal if they started with 0.
782 The documentation was vague. These values are now always treated as
783 decimal, and I will make that clear in the documentation.
789 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
790 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
793 MH/01 Don't check that the operands of numeric comparison operators are
794 integers when their expansion is in "skipping" mode (fixes bug
795 introduced by 4.64-PH/07).
797 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
798 child addresses, Exim now panics and dies. Previously, because the count
799 is held in a short int, deliveries were likely to be lost. As such a
800 large number of recipients for a single message is ridiculous
801 (performance will be very, very poor), I have chosen to impose a limit
802 rather than extend the field.
808 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
809 leftover -K file (the existence of which was triggered by #402).
810 While we were at it, introduced process PID as part of the -K
811 filename. This should rule out race conditions when creating
814 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
815 processing considerably. Previous code took too long for large mails,
816 triggering a timeout which in turn triggers #401.
818 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
819 in the DK code in transports.c. sendfile() is not really portable,
820 hence the _LINUX specificness.
822 TF/01 In the add_headers option to the mail command in an Exim filter,
823 there was a bug that Exim would claim a syntax error in any
824 header after the first one which had an odd number of characters
827 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
828 callout verification, Exim cached a "reject" for the entire domain. This
829 is correct for most verifications, but it is not correct for a recipient
830 verification with use_sender or use_postmaster set, because in that case
831 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
832 case of MAIL FROM:<> rejection from other early rejections (e.g.
833 rejection of HELO). When verifying a recipient using a non-null MAIL
834 address, the cache is ignored if it shows MAIL FROM:<> rejection.
835 Whatever the result of the callout, the value of the domain cache is
836 left unchanged (for any other kind of callout, getting as far as trying
837 RCPT means that the domain itself is ok).
839 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
842 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
843 manifest itself as EPIPE rather than ECONNECT. When tidying away a
844 session, the daemon ignores ECONNECT errors and logs others; it now
845 ignores EPIPE as well.
847 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
848 (quoted-printable decoding).
850 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
851 later the small subsequent patch to fix an introduced bug.
853 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
855 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
857 PH/08 An error is now given if message_size_limit is specified negative.
859 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
860 to be given (somewhat) arbitrary names.
862 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
865 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
866 miscellaneous code fixes
868 PH/10 Added the log_reject_target ACL modifier to specify where to log
871 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
872 hostname. This is wrong, because it relates to the incoming message (and
873 probably the interface on which it is arriving) and not to the outgoing
874 callout (which could be using a different interface). This has been
875 changed to use the value of the helo_data option from the smtp transport
876 instead - this is what is used when a message is actually being sent. If
877 there is no remote transport (possible with a router that sets up host
878 addresses), $smtp_active_hostname is used.
880 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
881 tweaks were necessary in order to get it to work (see also 21 below):
882 (a) The code assumed that strncpy() returns a negative number on buffer
883 overflow, which isn't the case. Replaced with Exim's string_format()
885 (b) There were several signed/unsigned issues. I just did the minimum
886 hacking in of casts. There is scope for a larger refactoring.
887 (c) The code used strcasecmp() which is not a standard C function.
888 Replaced with Exim's strcmpic() function.
889 (d) The code set only $1; it now sets $auth1 as well.
890 (e) A simple test gave the error "authentication client didn't specify
891 service in request". It would seem that Dovecot has changed its
892 interface. Fortunately there's a specification; I followed it and
893 changed what the client sends and it appears to be working now.
895 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
898 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
899 address A is aliased to B and C, where B exists and C does not. Without
900 -v the output is "A verified" because verification stops after a
901 successful redirection if more than one address is generated. However,
902 with -v the child addresses are also verified. Exim was outputting "A
903 failed to verify" and then showing the successful verification for C,
904 with its parentage. It now outputs "B failed to verify", showing B's
905 parentage before showing the successful verification of C.
907 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
908 look up a TXT record in a specific list after matching in a combined
911 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
912 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
913 they consult the DNS. I had assumed they would set it the way they
914 wanted; and indeed my experiments on Linux seem to show that in some
915 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
916 To be on the safe side, however, I have now made the interface to
917 host_find_byname() similar to host_find_bydns(), with an argument
918 containing the DNS resolver options. The host_find_byname() function now
919 sets these options at its start, just as host_find_bydns() does. The smtp
920 transport options dns_qualify_single and dns_search_parents are passed to
921 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
922 of host_find_byname() use the default settings of RES_DEFNAMES
923 (qualify_single) but not RES_DNSRCH (search_parents).
925 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
926 spool_read_header() do less string testing, by means of a preliminary
927 switch on the second character of optional "-foo" lines. (This is
928 overdue, caused by the large number of possibilities that now exist.
929 Originally there were few.) While I was there, I also converted the
930 str(n)cmp tests so they don't re-test the leading "-" and the first
931 character, in the hope this might squeeze out yet more improvement.
933 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
934 flag allowing group syntax was set by the header_syntax check but not
935 turned off, possible causing trouble later; (2) The flag was not being
936 set at all for the header_verify test, causing "group"-style headers to
937 be rejected. I have now set it in this case, and also caused header_
938 verify to ignore an empty address taken from a group. While doing this, I
939 came across some other cases where the code for allowing group syntax
940 while scanning a header line wasn't quite right (mostly, not resetting
941 the flag correctly in the right place). These bugs could have caused
942 trouble for malformed header lines. I hope it is now all correct.
944 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
945 with the "reply" argument non-NULL. The code, however (which originally
946 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
947 but it didn't always do it. This confused somebody who was copying the
948 code for some other use. I have removed all the tests.
950 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
951 feature that was used to support insecure browsers during the U.S. crypto
952 embargo. It requires special client support, and Exim is probably the
953 only MTA that supported it -- and would never use it because real RSA is
954 always available. This code has been removed, because it had the bad
955 effect of slowing Exim down by computing (never used) parameters for the
956 RSA_EXPORT functionality.
958 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
959 authenticator to fail if there's a tab character in the incoming data
960 (there should never be unless someone is messing about, as it's supposed
961 to be base64-encoded). Also added, on Timo's advice, the "secured" option
962 if the connection is using TLS or if the remote IP is the same as the
963 local IP, and the "valid-client-cert option" if a client certificate has
966 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
967 authenticators. This can be used for authorization after authentication
968 succeeds. (In the case of plaintext, it servers for both authentication
971 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
972 if any retry times were supplied.
974 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
975 connection, where there is no client IP address to check. In this
976 situation, the verify now always succeeds.
978 PH/25 Applied John Jetmore's -Mset patch.
980 PH/26 Added -bem to be like -Mset, but loading a message from a file.
982 PH/27 In a string expansion for a processed (not raw) header when multiple
983 headers of the same name were present, leading whitespace was being
984 removed from all of them, but trailing whitespace was being removed only
985 from the last one. Now trailing whitespace is removed from each header
986 before concatenation. Completely empty headers in a concatenation (as
989 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
990 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
992 PH/29 [Removed. This was a change that I later backed out, and forgot to
993 correct the ChangeLog entry (that I had efficiently created) before
994 committing the later change.]
996 PH/30 Exim was sometimes attempting to deliver messages that had suffered
997 address errors (4xx response to RCPT) over the same connection as other
998 messages routed to the same hosts. Such deliveries are always "forced",
999 so retry times are not inspected. This resulted in far too many retries
1000 for the affected addresses. The effect occurred only when there were more
1001 hosts than the hosts_max_try setting in the smtp transport when it had
1002 the 4xx errors. Those hosts that it had tried were not added to the list
1003 of hosts for which the message was waiting, so if all were tried, there
1004 was no problem. Two fixes have been applied:
1006 (i) If there are any address or message errors in an SMTP delivery, none
1007 of the hosts (tried or untried) are now added to the list of hosts
1008 for which the message is waiting, so the message should not be a
1009 candidate for sending over the same connection that was used for a
1010 successful delivery of some other message. This seems entirely
1011 reasonable: after all the message is NOT "waiting for some host".
1012 This is so "obvious" that I'm not sure why it wasn't done
1013 previously. Hope I haven't missed anything, but it can't do any
1014 harm, as the worst effect is to miss an optimization.
1016 (ii) If, despite (i), such a delivery is accidentally attempted, the
1017 routing retry time is respected, so at least it doesn't keep
1018 hammering the server.
1020 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1021 in ${readsocket because some servers need this prod.
1023 PH/32 Added some extra debug output when updating a wait-xxx database.
1025 PH/33 The hint "could be header name not terminated by colon", which has been
1026 given for certain expansion errors for a long time, was not being given
1027 for the ${if def:h_colon_omitted{... case.
1029 PH/34 The spec says: "With one important exception, whenever a domain list is
1030 being scanned, $domain contains the subject domain." There was at least
1031 one case where this was not true.
1033 PH/35 The error "getsockname() failed: connection reset by peer" was being
1034 written to the panic log as well as the main log, but it isn't really
1035 panic-worthy as it just means the connection died rather early on. I have
1036 removed the panic log writing for the ECONNRESET error when getsockname()
1039 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1040 runs only) independently of the message's sender address. This meant
1041 that, if the 4xx error was in fact related to the sender, a different
1042 message to the same recipient with a different sender could confuse
1043 things. In particualar, this can happen when sending to a greylisting
1044 server, but other circumstances could also provoke similar problems.
1045 I have changed the default so that the retry time for these errors is now
1046 based a combination of the sender and recipient addresses. This change
1047 can be overridden by setting address_retry_include_sender=false in the
1050 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1051 remote server are returned as part of bounce messages. This was not
1052 happening for LMTP over a pipe (the lmtp transport), but now it is the
1053 same for both kinds of LMTP.
1055 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1056 in header lines that were in fact CNAMEs. This is no longer the case.
1058 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1059 and queue runs started by the daemon processed all messages. This has
1060 been fixed so that -R and -S can now usefully be given with -q<time>.
1062 PH/40 Import PCRE release 6.7 (fixes some bugs).
1064 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1066 PH/42 Give an error if -q is specified more than once.
1068 PH/43 Renamed the variables $interface_address and $interface_port as
1069 $received_ip_address and $received_port, to make it clear that these
1070 values apply to message reception, and not to the outgoing interface when
1071 a message is delivered. (The old names remain recognized, of course.)
1073 PH/44 There was no timeout on the connect() call when using a Unix domain
1074 socket in the ${readsocket expansion. There now is.
1076 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1077 be meaningful with "accept".
1079 SC/01 Eximstats V1.43
1080 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1082 SC/02 Eximstats V1.44
1083 Use a glob alias rather than an array ref in the generated
1084 parser. This improves both readability and performance.
1086 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1087 Collect SpamAssassin and rejection statistics.
1088 Don't display local sender or destination tables unless
1089 there is data to show.
1090 Added average volumes into the top table text output.
1092 SC/04 Eximstats V1.46
1093 Collect data on the number of addresses (recipients)
1094 as well as the number of messages.
1096 SC/05 Eximstats V1.47
1097 Added 'Message too big' to the list of mail rejection
1098 reasons (thanks to Marco Gaiarin).
1100 SC/06 Eximstats V1.48
1101 Mainlog lines which have GMT offsets and are too short to
1102 have a flag are now skipped.
1104 SC/07 Eximstats V1.49 (Alain Williams)
1105 Added the -emptyok flag.
1107 SC/08 Eximstats V1.50
1108 Fixes for obtaining the IP address from reject messages.
1110 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1111 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1112 whitesspace changes from 4.64-PH/27
1114 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1117 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1118 are found, allow negative numbers in numeric criteria)
1120 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1122 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1123 to match changes made in 4.64-PH/43
1125 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1127 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1128 "domain" is actually an IP address, and force a failure. This locks out
1129 those revolvers/nameservers that support "A-for-A" lookups, in
1130 contravention of the specifications.
1132 PH/48 When a host name was looked up from an IP address, and the subsequent
1133 forward lookup of the name timed out, the host name was left in
1134 $sender_host_name, contrary to the specification.
1136 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1137 restricted to single-key lookups, Exim was not diagnosing an error if
1138 * or *@ was used with a query-style lookup.
1140 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1142 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1143 long ago, but noone interested enough thought of it. Let's just say that
1144 the "1.1" means that there are some new functions that weren't there at
1145 some point in the past.
1147 PH/51 Error processing for expansion failure of helo_data from an smtp
1148 transport during callout processing was broken.
1150 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1151 tested/used via the -bh/-bhc/-bs options.
1153 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1154 bug, fixed in subsequent PCRE releases).
1156 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1157 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1159 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1165 SC/01 Use a glob alias rather than an array ref in eximstats generated
1166 parser. This improves both readability and performance.
1168 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1169 Don't display local sender or destination tables in eximstats unless
1170 there is data to show.
1171 Added average volumes into the eximstats top table text output.
1173 SC/03 Collect data on the number of addresses (recipients) as well
1174 as the number of messages in eximstats.
1176 TF/01 Correct an error in the documentation for the redirect router. Exim
1177 does (usually) call initgroups() when daemonizing.
1179 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1180 with consistent privilege compared to when running as a daemon.
1182 TF/03 Note in the spec that $authenticated_id is not set for local
1183 submissions from trusted users.
1185 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1186 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1188 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1189 by adding some example configuration directives to the default
1190 configuration file. A little bit of work is required to uncomment the
1191 directives and define how usernames and passwords are checked, but
1192 there is now a framework to start from.
1194 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1195 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1196 without this. I don't know how relevant this is to other LDAP libraries.
1198 PH/02 Add the verb name to the "unknown ACL verb" error.
1200 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1202 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1204 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1205 directory not expanded when it should be if an expanded home directory
1206 was set for the address (which is overridden by the transport).
1208 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1211 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1212 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1213 because it is too late at that time, and has no effect.
1215 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1216 security issue with \' (bugzilla #107). I could not use the
1217 PQescapeStringConn() function, because it needs a PGconn value as one of
1220 PH/08 When testing addresses using -bt, indicate those final addresses that
1221 are duplicates that would not cause an additional delivery. At least one
1222 person was confused, thinking that -bt output corresponded to deliveries.
1223 (Suppressing duplicates isn't a good idea as you lose the information
1224 about possibly different redirections that led to the duplicates.)
1226 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1227 systems where poll() doesn't work, in particular OS X.
1229 PH/10 Added more information to debugging output for retry time not reached.
1231 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1232 operations in malware.c.
1234 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1237 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1238 syslog_duplication set false, log lines that would normally be written
1239 both the the main log and to the reject log were not written to syslog at
1242 PH/14 In the default configuration, change the use of "message" in ACL warn
1243 statements to "add_header".
1245 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1246 not followed by a command (e.g. "seen endif").
1248 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1249 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1252 PH/17 Added extra conditions to the default value of delay_warning_condition
1256 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1257 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1258 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1261 The Auto-Submitted: and various List- headers are standardised, whereas I
1262 don't think Precedence: ever was.
1264 PH/18 Refactored debugging code in route_finduser() to show more information,
1265 in particular, the error code if getpwnam() issues one.
1267 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1268 This is apparently needed in addition to the PH/07 change above to avoid
1269 any possible encoding problems.
1271 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1272 but not after initializing Perl.
1274 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1275 output them only if debugging. By default they are written stderr,
1276 apparently, which is not desirable.
1278 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1281 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1284 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1286 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1287 authenticated or an ident call has been made. Suppress the default
1288 values for $authenticated_id and $authenticated_sender (but permit -oMai
1289 and -oMas) when testing with -bh.
1291 PH/24 Re-jigged the order of the tests in the default configuration so that the
1292 tests for valid domains and recipients precede the DNS black list and CSA
1293 tests, on the grounds that those ones are more expensive.
1295 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1296 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1297 This bug exists in every version of Exim that I still have, right back to
1300 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1301 However, an attempt to turn on case-sensitivity in a regex key by
1302 including (?-i) didn't work because the subject string was already
1303 lowercased, and the effects were non-intuitive. It turns out that a
1304 one-line patch can be used to allow (?-i) to work as expected.
1310 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1311 other effects) broke the use of negated acl sub-conditions.
1313 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1316 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1317 "Deny" causes Exim to reject the incoming connection with a 554 error.
1318 Unfortunately, if there is a major crisis, such as a disk failure,
1319 tcp-wrappers gives "deny", whereas what one would like would be some
1320 kind of temporary error. A kludge has been added to help with this.
1321 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1322 554 error is used if errno is still zero or contains ENOENT (which occurs
1323 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1326 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1328 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1329 errors. Otherwise a message that provokes a temporary error (when other
1330 messages do not) can cause a whole host to time out.
1332 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1333 addresses were routed directly to files or pipes from a redirect router.
1334 File deliveries just didn't batch; pipe deliveries might have suffered
1337 PH/06 A failure to get a lock for a hints database would erroneously always say
1338 "Failed to get write lock", even when it was really a read lock.
1340 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1341 of 0600. This has been changed to use the value of the lockfile_mode
1342 option (which defaults to 0600).
1344 PH/08 Applied small patch from the Sieve maintainer.
1346 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1347 folder from quota calculations, a direct delivery into this folder messed
1348 up the contents of the maildirsize file. This was because the regex was
1349 used only to exclude .Trash (or whatever) when the size of the mailbox
1350 was calculated. There was no check that a delivery was happening into an
1351 excluded directory. This bug has been fixed by ignoring all quota
1352 processing for deliveries into excluded directories.
1354 PH/10 Added the maildirfolder_create_regex option to appendfile.
1360 PH/01 The code for finding all the local interface addresses on a FreeBSD
1361 system running IPv6 was broken. This may well have applied to all BSD
1362 systems, as well as to others that have similar system calls. The broken
1363 code found IPv4 interfaces correctly, but gave incorrect values for the
1364 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1365 that it would not match correctly against @[] and not recognize the IPv6
1368 PH/02 The ipliteral router was not recognizing addresses of the form user@
1369 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1371 PH/03 Added disable_ipv6.
1373 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1374 decoded form, because it is most often used to construct To: headers
1375 lines in autoreplies, and the decoded form may well be syntactically
1376 invalid. However, $reply_address has leading white space removed, and all
1377 newlines turned into spaces so that the autoreply transport does not
1380 PH/05 If group was specified without a user on a router, and no group or user
1381 was specified on a transport, the group from the router was ignored.
1383 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1384 for visible compile-time settings that can be used to change these
1385 numbers, for those that want even more. Backwards compatibility with old
1386 spool files has been maintained. However, going back to a previous Exim
1387 release will lost any variables that are in spool files.
1389 PH/07 Two small changes when running in the test harness: increase delay when
1390 passing a TCP/IP connection to a new process, in case the original
1391 process has to generate a bounce, and remove special handling of
1392 127.0.0.2 (sic), which is no longer necessary.
1394 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1395 be the same on different OS.
1397 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1400 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1401 whether --show-vars was specified or not
1403 JJ/02 exipick: Added support for new ACL variable spool format introduced
1406 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1407 syntactically invalid From: or Reply-to: line, and a filter used this to
1408 generate an autoreply, and therefore failed to obtain an address for the
1409 autoreply, Exim could try to deliver to a non-existent relative file
1410 name, causing unrelated and misleading errors. What now happens is that
1411 it logs this as a hard delivery error, but does not attempt to create a
1414 PH/11 The exinext utility has a -C option for testing purposes, but although
1415 the given file was scanned by exinext itself; it wasn't being passed on
1416 when Exim was called.
1418 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1419 an end-of-file indication when reading a command response.
1421 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1422 compiled. In many other places in Exim, IPv6 addresses are always
1423 recognized, so I have changed this. It also means that IPv4 domain
1424 literals of the form [IPV4:n.n.n.n] are now always recognized.
1426 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1427 used if the router is not running as root, for example, when verifying at
1428 ACL time, or when using -bh. The debugging output from this situation was
1429 non-existent - all you got was a failure to exec. I have made two
1432 (a) Failures to set uid/gid, the current directory, or a process leader
1433 in a subprocess such as that created by queryprogram now generate
1434 suitable debugging ouput when -d is set.
1436 (b) The queryprogram router detects when it is not running as root,
1437 outputs suitable debugging information if -d is set, and then runs
1438 the subprocess without attempting to change uid/gid.
1440 PH/15 Minor change to Makefile for building test_host (undocumented testing
1443 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1444 additional section of a DNS packet that returns MX or SRV records.
1445 Instead, it always explicitly searches for A/AAAA records. This avoids
1446 major problems that occur when a DNS server includes only records of one
1447 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1448 fixed another bug: if SRV records were looked up and the corresponding
1449 address records were *not* found in the additional section, the port
1450 values from the SRV records were lost.
1452 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1453 using the correct key (the original address) when searching the retry
1454 rules in order to find which one to use for generating the retry hint.
1456 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1457 adding the default one. Similarly, if it contains a Reply-To: header, the
1458 errors_reply_to option, if set, is not used.
1460 PH/19 When calculating a retry time, Exim used to measure the "time since
1461 failure" by looking at the "first failed" field in the retry record. Now
1462 it does not use this if it is later than than the arrival time of the
1463 message. Instead it uses the arrival time. This makes for better
1464 behaviour in cases where some deliveries succeed, thus re-setting the
1465 "first failed" field. An example is a quota failure for a huge message
1466 when small messages continue to be delivered. Without this change, the
1467 "time since failure" will always be short, possible causing more frequent
1468 delivery attempts for the huge message than are intended.
1469 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1471 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1472 $1, $2, $3) because the numerical variables can be reset during some
1473 expansion items (e.g. "match"), thereby losing the authentication data.
1475 PH/21 Make -bV show the size of off_t variables so that the test suite can
1476 decide whether to run tests for quotas > 2G.
1478 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1479 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1480 filecount value is greater than 2G or if a quota value is greater than 2G
1481 on a system where the size of off_t is not greater than 4, a panic error
1484 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1485 never match. The debug and -bh output now contains an explicit error
1486 message indicating a malformed IPv4 address or mask.
1488 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1489 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1490 PH/23 above applies.
1492 PH/25 Do not write to syslog when running in the test harness. The only
1493 occasion when this arises is a failure to open the main or panic logs
1494 (for which there is an explicit test).
1496 PH/26 Added the /no_tell option to "control=freeze".
1498 PH/27 If a host name lookup failed very early in a connection, for example, if
1499 the IP address matched host_lookup and the reverse lookup yielded a name
1500 that did not have a forward lookup, an error message of the form "no IP
1501 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1502 could be logged. Now it outputs the IP address instead of "NULL".
1504 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1505 allows the sender and the authenticated sender to be set when
1506 submitting a message from within Exim. Since child_open_exim() is
1507 documented for local_scan(), the new function should be too.
1509 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1510 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1511 results in an empty string is now treated as unset.
1513 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1515 PH/31 Added sender_verify_fail logging option.
1517 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1518 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1519 tidied the source and removed it altogether.
1521 PH/33 When a queue run was abandoned because the load average was too high, a
1522 log line was always written; now it is written only if the queue_run log
1523 selector is set. In addition, the log line for abandonment now contains
1524 information about the queue run such as the pid. This is always present
1525 in "start" and "stop" lines but was omitted from the "abandon" line.
1527 PH/34 Omit spaces between a header name and the colon in the error message that
1528 is given when verify = headers_syntax fails (if there are lots of them,
1529 the message gets confusing).
1531 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1532 names, as there are now some PTR records that contain slashes. This check
1533 is only to protect against broken name servers that fall over on strange
1534 characters, so the fact that it applies to all lookups doesn't matter.
1536 PH/36 Now that the new test suite is complete, we can remove some of the
1537 special code in Exim that was needed for the old test suite. For example,
1538 sorting DNS records because real resolvers return them in an arbitrary
1539 order. The new test suite's fake resolver always returns records in the
1542 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1543 bounces) except when queue_only is set, to avoid logging races between
1544 the different processes.
1546 PH/38 Panic-die if .include specifies a non-absolute path.
1548 PH/39 A tweak to the "H" retry rule from its user.
1550 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1551 a label. They prevented compilation on older perls.
1553 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1554 a warning to be raised on newish perls.
1556 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1557 on queue. Changes to match documented behaviour of showing count of
1558 messages matching specified criteria.
1560 PH/40 Changed the default ident timeout from 30s to 5s.
1562 PH/41 Added support for the use of login_cap features, on those BSD systems
1563 that have them, for controlling the resources used by pipe deliveries.
1565 PH/42 The content-scanning code uses fopen() to create files in which to put
1566 message data. Previously it was not paying any attention to the mode of
1567 the files. Exim runs with umask(0) because the rest of the code creates
1568 files with open(), and sets the required mode explicitly. Thus, these
1569 files were ending up world-writeable. This was not a big issue, because,
1570 being within the spool directory, they were not world-accessible. I have
1571 created a function called modefopen, which takes an additional mode
1572 argument. It sets umask(777), creates the file, chmods it to the required
1573 mode, then resets the umask. All the relevant calls to fopen() in the
1574 content scanning code have been changed to use this function.
1576 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1577 to 24 hours. This avoids potential overflow problems when processing G
1578 and H retry rules. I suspect nobody ever tinkers with this value.
1580 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1582 PH/45 When the plaintext authenticator is running as a client, the server's
1583 challenges are checked to ensure they are valid base64 strings. By
1584 default, the authentication attempt is cancelled if an invalid string is
1585 received. Setting client_ignore_invalid_base64 true ignores these errors.
1586 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1587 they are received. Thus, the responses can be made to depend on the
1588 challenges. If an invalid string is ignored, an empty string is placed in
1591 PH/46 Messages that are created by the autoreply transport now contains a
1592 References: header, in accordance with RFCs 2822 and 3834.
1594 PH/47 Added authenticated_sender_force to the smtp transport.
1596 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1598 PH/49 Installed latest patch from the Sieve maintainer.
1600 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1601 was also set, the appendfile transport was unnecessarily scanning a
1602 directory of message files (e.g. for maildir delivery) to find the count
1603 of files (along with the size), even though it did not need this
1604 information. It now does the scan only if it needs to find either the
1605 size of the count of files.
1607 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1609 PH/52 Two bugs concerned with error handling when the smtp transport is
1612 (i) Exim was not creating retry information for temporary errors given
1613 for individual recipients after the DATA command when the smtp transport
1614 was used in LMTP mode. This meant that they could be retried too
1615 frequently, and not timed out correctly.
1617 (ii) Exim was setting the flag that allows error details to be returned
1618 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1619 recipients that were returned after the DATA command.
1621 PH/53 This is related to PH/52, but is more general: for any failing address,
1622 when detailed error information was permitted to be returned to the
1623 sender, but the error was temporary, then after the final timeout, only
1624 "retry timeout exceeded" was returned. Now it returns the full error as
1625 well as "retry timeout exceeded".
1627 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1628 do this, and (what is worse) MTAs that accept it.
1630 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1631 will now be deprecated.
1633 PH/56 New os.c-cygwin from the Cygwin maintainer.
1635 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1636 formats (previously only available in exim formats via -bpr, -bpru,
1637 and -bpra. Now also available in native and exiqgrep formats)
1639 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1640 with very large, slow to parse queues
1642 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1644 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1646 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1647 responses to authentication challenges, though it was showing the
1648 challenges; (ii) I've removed the CR characters from the debug output for
1651 PH/58 Allow for the insertion of a newline as well as a space when a string
1652 is turned into more than one encoded-word during RFC 2047 encoding. The
1653 Sieve code now uses this.
1655 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1656 data_4xx, lost_connection, tls_required.
1658 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1659 message was being sent as an SMTP response.
1661 PH/61 Add -l and -k options to exicyclog.
1663 PH/62 When verifying, if an address was redirected to one new address, so that
1664 verification continued, and the new address failed or deferred after
1665 having set something in $address_data, the value of $address_data was not
1666 passed back to the ACL. This was different to the case when no
1667 redirection occurred. The value is now passed back in both cases.
1669 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1670 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1671 use login_cap.h, so on its own it isn't the distinguishing feature. The
1672 new name refers directly to the setclassresources() function.
1674 PH/65 Added configuration files for NetBSD3.
1676 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1678 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1679 is preferred over IPv4.
1681 PH/68 The bounce_return_message and bounce_return_body options were not being
1682 honoured for bounces generated during the reception of non-SMTP messages.
1683 In particular, this applied to messages rejected by the ACL. This bug has
1684 been fixed. However, if bounce_return_message is true and bounce_return_
1685 body is false, the headers that are returned for a non-SMTP message
1686 include only those that have been read before the error was detected.
1687 (In the case of an ACL rejection, they have all been read.)
1689 PH/69 The HTML version of the specification is now built in a directory called
1690 spec_html instead of spec.html, because the latter looks like a path with
1691 a MIME-type, and this confuses some software.
1693 PH/70 Catch two compiler warnings in sieve.c.
1695 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1696 function verify_get_ident() calls ip_connect() to connect a socket, but
1697 if the "connect()" function timed out, ip_connect() used to close the
1698 socket. However, verify_get_ident() also closes the socket later, and in
1699 between Exim writes to the log, which may get opened at this point. When
1700 the socket was closed in ip_connect(), the log could get the same file
1701 descriptor number as the socket. This naturally causes chaos. The fix is
1702 not to close the socket in ip_connect(); the socket should be closed by
1703 the function that creates it. There was only one place in the code where
1704 this was missing, in the iplookup router, which I don't think anybody now
1705 uses, but I've fixed it anyway.
1707 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1708 well as to direct DNS lookups. Otherwise the handling of names in host
1709 lists is inconsistent and therefore confusing.
1715 PH/01 Two changes to the default runtime configuration:
1717 (1) Move the checks for relay_from_hosts and authenticated clients from
1718 after to before the (commented out) DNS black list checks.
1720 (2) Add control=submission to the relay_from_hosts and authenticated
1721 clients checks, on the grounds that messages accepted by these
1722 statements are most likely to be submissions.
1724 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1726 (1) Generate an error if the third argument for the ${prvs expansion is
1729 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1732 (3) Reset the variables that are obtained from the first argument of
1733 ${prvscheck and used in the second argument before leaving the code,
1734 because their memory is reclaimed, so using them afterwards may do
1737 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1738 one (it's much easier than Tom thought :-).
1740 (5) Because of (4), we can now allow for the use of $prvscheck_result
1741 inside the third argument.
1743 PH/03 For some reason, the default setting of PATH when running a command from
1744 a pipe transport was just "/usr/bin". I have changed it to
1747 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1748 anything to be listed in the output from -bV.
1750 PH/05 When a filter generated an autoreply, the entire To: header line was
1751 quoted in the delivery log line, like this:
1753 => >A.N.Other <ano@some.domain> <original@ddress> ...
1755 This has been changed so that it extracts the operative address. There
1756 may be more than one such address. If so, they are comma-separated, like
1759 => >ano@some.domain,ona@other.domain <original@ddress> ...
1761 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1762 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1763 not being looked up in the rDNS to get a host name, Exim was showing the
1764 IP address twice in Received: lines, even though the IP addresses were
1765 identical. For example:
1767 Received: from [1.2.3.4] (helo=[1.2.3.4])
1769 However, if the real host name was known, it was omitting the HELO data
1770 if it matched the actual IP address. This has been tidied up so that it
1771 doesn't show the same IP address twice.
1773 PH/07 When both +timestamp and +memory debugging was on, the value given by
1774 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1775 called by the debug printing, thereby overwriting the timestamp buffer.
1776 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1778 PH/08 When the original message was included in an autoreply transport, it
1779 always said "this is a copy of the message, including all the headers",
1780 even if body_only or headers_only was set. It now gives an appropriate
1783 PH/09 Applied a patch from the Sieve maintainer which:
1785 o fixes some comments
1786 o adds the (disabled) notify extension core
1787 o adds some debug output for the result of if/elsif tests
1788 o points to the current vacation draft in the documentation
1789 and documents the missing references header update
1793 o fixes a bug in processing the envelope test (when testing
1794 multiple envelope elements, the last element determinted the
1797 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1798 Electronic Mail") by including:
1800 Auto-submitted: auto-generated
1802 in the messages that it generates (bounce messages and others, such as
1803 warnings). In the case of bounce messages for non-SMTP mesages, there was
1804 also a typo: it was using "Auto_submitted" (underscore instead of
1805 hyphen). Since every message generated by Exim is necessarily in response
1806 to another message, thes have all been changed to:
1808 Auto-Submitted: auto-replied
1810 in accordance with these statements in the RFC:
1812 The auto-replied keyword:
1814 - SHOULD be used on messages sent in direct response to another
1815 message by an automatic process,
1817 - MUST NOT be used on manually-generated messages,
1819 - MAY be used on Delivery Status Notifications (DSNs) and Message
1820 Disposition Notifications (MDNs),
1822 - MUST NOT be used on messages generated by automatic or periodic
1823 processes, except for messages which are automatic responses to
1826 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1827 to the default Received: header definition.
1829 PH/12 Added log selector acl_warn_skipped (default on).
1831 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1832 variables because (a) they are in the wrong storage pool and (b) even if
1833 they were copied, it wouldn't work properly because of the caching.
1835 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1836 checking when decoding. Apparently there are clients that generate
1837 overlong encoded strings. Why am I not surprised?
1839 PH/15 If the first argument of "${if match_address" was not empty, but did not
1840 contain an "@" character, Exim crashed. Now it writes a panic log message
1841 and treats the condition as false.
1843 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1845 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1846 extension "envelope-auth". The code is finished and in agreement with
1847 other implementations, but there is no documentation so far and in fact,
1848 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1849 not changing the active code.
1851 Print executed "if" and "elsif" statements when debugging is used. This
1852 helps a great deal to understand what a filter does.
1854 Document more things not specified clearly in RFC3028. I had all this
1855 sorted out, when out of a sudden new issues came to my mind. Oops."
1857 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1860 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1861 canonical form (as documented). However, after a host name lookup from
1862 the IP address, check_host() was doing a simple string comparison with
1863 addresses acquired from the DNS when checking that the found name did
1864 have the original IP as one of its addresses. Since any found IPv6
1865 addresses are likely to be in abbreviated form, the comparison could
1866 fail. Luckily, there already exists a function for doing the comparison
1867 by converting both addresses to binary, so now that is used instead of
1868 the text comparison.
1870 PH/20 There was another similar case to PH/19, when a complete host name was
1871 given in a host list; looking up its IP address could give an abbreviated
1872 form, whereas the current host's name might or might not be abbreviated.
1873 The same fix has been applied.
1879 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1880 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1883 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1884 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1886 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1888 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1889 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1890 Also, for "Auto-Submitted", treat anything other than "no" as
1891 non-personal, in accordance with RFC 3834. (Previously it treated
1892 anything starting "auto-" as non-personal.)
1894 TF/01 The control=submission/name=... option had a problem with syntax
1895 errors if the name included a slash character. The /name= option
1896 now slurps the rest of the string, so it can include any characters
1897 but it must come last in the list of options (after /sender_retain
1900 PH/05 Some modifications to the interface to the fake nameserver for the new
1908 TK/01 Added the "success_on_redirect" address verification option. See
1909 NewStuff for rationale and an example.
1911 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1913 PH/02 Patch to exigrep to allow it to work on syslog lines.
1915 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1916 fread() to skip over the body file's header line, because in Cygwin the
1917 header line is locked and is inaccessible.
1919 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1920 co-exist for some time) to make it clear that it is the Exim ID that is
1921 referenced, not the Message-ID: header line.
1923 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1924 string_format() function, because snprintf() does not exist on all
1927 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1928 ${stat: expansion item.
1930 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1931 protocol synchronization error", to keep the pedants happy.
1933 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1934 well as for IRIX systems, when gcc is being used. See the host.c source
1937 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1939 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1942 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1943 transport and to the smtp transport in LMTP mode.
1945 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1947 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1948 run a filter in a subprocess. This could lead to confusion in subsequent
1949 lookups in the parent process. There should also be a search_tidyup() at
1950 the end of the subprocess.
1952 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1953 only if the host matched helo_try_verify_hosts, which caused the
1954 verification to occur when the EHLO/HELO command was issued. The ACL just
1955 tested the remembered result. Now, if a previous verification attempt has
1956 not happened, "verify = helo" does it there and then.
1958 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1960 TK/03 Fix log output including CR from clamd.
1962 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1963 exist provoked a memory error which could cause a segfault.
1965 PH/15 Installed PCRE 6.2
1967 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1969 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1970 of the problem. Specifically, suggested +O2 rather than +O1 for the
1973 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1975 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1976 record, the port was not correctly logged when the outgoing_port log
1977 selector was set (it logged the transort's default port).
1979 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1980 fallback_hosts, and "hosts" in the smtp transport.
1982 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1983 host errors such as "Connection refused".
1985 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1986 authentication with radiusclient 0.4.9:
1988 - Error returned from rc_read_config was caught wrongly
1989 - Username/password not passed on to radius server due to wrong length.
1991 The presumption is that some radiusclient API changes for 4.51/PH/17
1992 were not taken care of correctly. The code is still untested by me (my
1993 Linux distribution still has 0.3.2 of radiusclient), but it was
1994 contributed by a Radius user.
1996 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1997 expanding the "port" option of the smtp transport.
1999 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2000 while reading a MIME header. Thanks to Tom Hughes for a patch.
2002 PH/24 Include config.h inside local_scan.h so that configuration settings are
2005 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2006 that in an ACL for RCPT (for example), you can examine exactly what was
2009 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2010 commands, but it was not correctly comparing the address with the actual
2011 client host address. Thus, it would show the EHLO address in Received:
2012 header lines when this was not necessary.
2014 PH/27 Added the % operator to ${eval:}.
2016 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2017 it should be ignoring failures (because with -C, for example, it has lost
2018 privilege). It wasn't ignoring creation failures other than "already
2021 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2024 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2025 because an input line was too long, either on its own, or by virtue of
2026 too many continuations, the temporary file was not being removed, and the
2027 return code was incorrect.
2029 PH/31 Missing "BOOL" in function definition in filtertest.c.
2031 PH/32 Applied Sieve patches from the maintainer.
2033 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2035 PH/33 Added "verify = not_blind".
2037 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2038 Local/Makefile (with some defaults set). These are used in built scripts
2039 such as exicyclog, but they have never been used in the exim_install
2040 script (though there are many overriding facilities there). I have
2041 arranged that the exim_install script now takes note of these two
2044 PH/35 Installed configuration files for Dragonfly.
2046 PH/36 When a locally submitted message by a trusted user did not contain a
2047 From: header, and the sender address was obtained from -f or from an SMTP
2048 MAIL command, and the trusted user did not use -F to supply a sender
2049 name, $originator_name was incorrectly used when constructing a From:
2050 header. Furthermore, $originator_name was used for submission mode
2051 messages from external hosts without From: headers in a similar way,
2052 which is clearly wrong.
2054 PH/37 Added control=suppress_local_fixups.
2056 PH/38 When log_selector = +received_sender was set, and the addition of the
2057 sender made the log line's construction buffer exactly full, or one byte
2058 less than full, an overflow happened when the terminating "\n" was
2061 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2062 when the result of a list match is failure because a DNS lookup failed.
2064 PH/40 RM_COMMAND is now used in the building process.
2066 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2067 the "build-* directories that it finds.
2069 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2070 address in a domain literal was a prefix of an interface address.
2072 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2073 when verifying a sender address, unless rewrite_headers is false.
2075 PH/44 Wrote a long comment about why errors_to addresses are verified as
2076 recipients, not senders.
2078 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2079 the ratelimit ACL was added.
2081 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2083 PH/46 Added extra information about PostgreSQL errors to the error string.
2085 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2086 avoiding the need to install special zones in a real server. This is
2087 backwards compatible; if it can't find the fake resolver, it drops back.
2088 Thus, both old and new test suites can be run.
2090 TF/02 Added util/ratelimit.pl
2092 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2093 clock is set back in time.
2095 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2096 Candler <B.Candler@pobox.com>.
2098 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2099 OK for addresses that are the result of redirections.
2101 PH/48 A number of further additions for the benefit of the new test suite,
2102 including a fake gethostbyname() that interfaces to the fake DNS resolver
2105 TF/06 The fix for widen_domains has also been applied to qualify_single and
2106 search_parents which are the other dnslookup options that can cause
2109 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2112 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2114 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2115 service" if it encountered an MX record with an empty target hostname.
2116 The message is now "an MX or SRV record indicated no SMTP service".
2118 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2119 if the DNS of the sending site is misconfigured. This is quite a
2120 common situation. This change restores the behaviour of treating a
2121 helo verification defer as a failure.
2123 PH/51 If self=fail was set on a router, the bounce message did not include the
2124 actual error message.
2130 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2132 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2133 command itself ended in error, the underlying message about the transport
2134 filter timeout was being overwritten with the pipe command error. Now the
2135 underlying error message should be appended to the second error message.
2137 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2139 PH/02 Reduce the amount of output that "make" produces by default. Full output
2140 can still be requested.
2142 PH/03 The warning log line about a condition test deferring for a "warn" verb
2143 was being output only once per connection, rather than after each
2144 occurrence (because it was using the same function as for successful
2145 "warn" verbs). This seems wrong, so I have changed it.
2147 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2148 it should not have, which might have caused a crash in the right
2149 circumstances, but probably never did.
2151 PH/04 Installed a modified version of Tony Finch's patch to make submission
2152 mode fix the return path as well as the Sender: header line, and to
2153 add a /name= option so that you can make the user's friendly name appear
2156 TF/03 Added the control = fakedefer ACL modifier.
2158 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2159 Mark Lowes for thorough testing.
2161 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2163 TK/03 Merged latest SRS patch from Miles Wilton.
2165 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2166 with the definition in sysexits.h (which is #included earlier).
2167 Fortunately, Exim does not actually use EX_OK. The code used to try to
2168 preserve the sysexits.h value, by assumimg that macro definitions were
2169 scanned for macro replacements. I have been disabused of this notion,
2170 so now the code just undefines EX_OK before #including unistd.h.
2172 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2173 in the smtp transport. When a block could not be written in a single
2174 write() function, the timeout was being re-applied to each part-write.
2175 This seems wrong - if the receiver was accepting one byte at a time it
2176 would take for ever. The timeout is now adjusted when this happens. It
2177 doesn't have to be particularly precise.
2179 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2180 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2182 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2183 without a domain if the check to <postmaster@domain> fails.
2185 SC/01 Eximstats: added -xls and the ability to specify output files
2186 (patch written by Frank Heydlauf).
2188 SC/02 Eximstats: use FileHandles for outputing results.
2190 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2192 SC/04 Eximstats: fixed display of large numbers with -nvr option
2194 SC/05 Eximstats: fixed merging of reports with empty tables.
2196 SC/06 Eximstats: added the -include_original_destination flag
2198 SC/07 Eximstats: removed tabs and trailing whitespace.
2200 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2202 TK/06 MBOX spool code: Add real "From " MBOX separator line
2203 so the .eml file is really in mbox format (even though
2204 most programs do not really care). Patch from Alex Miller.
2206 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2207 The latter is generated from $received_to and is only set if the
2208 message has one envelope recipient. SA can use these headers,
2209 obviously out-of-the-box. Patch from Alex Miller.
2211 PH/08 The ${def test on a variable was returning false if the variable's
2212 value was "0", contrary to what the specification has always said!
2213 The result should be true unless the variable is empty.
2215 PH/09 The syntax error of a character other than { following "${if
2216 def:variable_name" (after optional whitespace) was not being diagnosed.
2217 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2218 accidental colon was present, for example, could give incorrect results.
2220 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2221 result is used (not including appendfile, where other changes are about
2224 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2225 This involved changing a lot of size variables from int to off_t. It
2226 should work with maildirs and everything.
2228 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2229 spamd dying while we are connected to it.
2231 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2234 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2235 function for BDB 4.3.
2237 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2239 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2240 the sources for PCRE have been re-arranged and more files are now
2243 PH/15 The code I had for printing potentially long long variables in PH/11
2244 above was not the best (it lost precision). The length of off_t variables
2245 is now inspected at build time, and an appropriate printing format (%ld
2246 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2247 to be "long long int" or "long int". This is needed for the internal
2248 formatting function string_vformat().
2250 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2251 the configuration file to be ":syslog", then the script "guesses" where
2252 the logs files are, rather than using the compiled in default. In our
2253 case the guess is not the same as the compiled default, so the script
2254 suddenly stopped working when I started to use syslog. The patch checks
2255 to see if log_file_path is "". If so, it attempts to read it from exim
2256 with no configuration file to get the compiled in version, before it
2257 falls back to the previous guessing code."
2259 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2260 implementing BATV in an Exim configuration. See NewStuff for the gory
2263 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2264 Makefile that are specific to HP-UX.
2266 PH/18 If the "use_postmaster" option was set for a recipient callout together
2267 with the "random" option, the postmaster address was used as the MAIL
2268 FROM address for the random test, but not for the subsequent recipient
2269 test. It is now used for both.
2271 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2272 patch removes a few documentation additions to RFC 3028, because the
2273 latest draft now contains them. It adds the new en;ascii-case comparator
2274 and a new error check for 8bit text in MIME parts. Comparator and
2275 require names are now matched exactly. I enabled the subaddress
2276 extension, but it is not well tested yet (read: it works for me)."
2278 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2279 rework some of the code of TK/09 above to avoid the hardwired use of
2280 "%lld" and "long long". Replaced the call to snprintf() with a call to
2283 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2284 records point to non-existent hosts", "retry timeout exceeded", and
2285 "retry time not reached for any host after a long failure period".
2287 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2288 experimental DomainKeys support:
2290 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2291 (2) On an error such as an illegally used "control", the wrong name for
2292 the control was given.
2294 These problems did NOT occur unless DomainKeys support was compiled.
2296 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2298 PH/24 Added ${if match_ip condition.
2300 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2301 databases so that it will be absolutely obvious if a crash occurs in the
2302 DB library. This is a regular occurrence (often caused by mis-matched
2305 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2306 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2307 picked up on a user's system that detects such things. There doesn't seem
2308 to be a gcc warning option for this - only an attribute that has to be
2309 put on the function's prototype. It seems that in Fedora Core 4 they have
2310 set this on a number of new functions. No doubt there will be more in due
2313 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2314 specify a transport. However, if an address that was verified by such a
2315 router was the subject of a callout, Exim crashed because it tried to
2316 read the rcpt_include_affixes from the non-existent transport. Now it
2317 just assumes that the setting of that option is false. This bug was
2318 introduced by 4.51/PH/31.
2320 PH/28 Changed -d+all to exclude +memory, because that information is very
2321 rarely of interest, but it makes the output a lot bigger. People tend to
2322 do -d+all out of habit.
2324 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2325 code in os-type was giving problems when libc.so lives in lib64, like on
2328 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2329 aren't the modern standard, and it seems that some systems' include files
2330 don't always have them. Exim was already checking for some of the newer
2331 ones like T_AAAA, and defining it itself. I've added checks for all the
2332 record types that Exim uses.
2334 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2335 not automatically generating a new one, as it is supposed to. This
2336 prevented TLS from working. If the file did exist, but contained invalid
2337 data, a new version was generated, as expected. It was only the case of a
2338 non-existent file that was broken.
2340 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2341 with a change in libdomainkeys > 0.64.
2343 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2344 from DNS. If the selector record carries the flag, it now has
2345 precedence over the domain-wide flag.
2347 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2349 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2350 the use of a port name that isn't defined in /etc/services) occurred, the
2351 message was deferred as in a normal delivery, and thus remained on the
2352 spool, instead of being failed because of the mua_wrapper setting. This
2353 is now fixed, and I tidied up some of the mua_wrapper messages at the
2356 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2357 the messages in a hash of arrays rather than using individual hashes.
2358 This is a bit cleaner and results in dramatic memory savings, albeit
2359 at a slight CPU cost.
2361 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2362 as requested by Marc Sherman.
2364 SC/10 Eximstats: added histograms for user specified patterns as requested
2367 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2369 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2370 fopen() in the content-scanning modules that did not already have it.
2376 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2377 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2379 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2381 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2383 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2384 to test Sieve filters that use "vacation".
2386 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2387 that changes the way the GnuTLS parameters are stored in the cache file.
2388 The new format can be generated externally. For backward compatibility,
2389 if the data in the cache doesn't make sense, Exim assumes it has read an
2390 old-format file, and it generates new data and writes a new file. This
2391 means that you can't go back to an older release without removing the
2394 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2395 work if there are any delivery delays because "one_time" forces the
2396 parent to be marked "delivered", so its unseen clone is never tried
2397 again. For this reason, Exim now forbids the simultaneous setting of
2400 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2401 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2402 entry if you want to know the details. The fix, however, neglected to
2403 consider the case when local delivery batching is involved. The test for
2404 "previously delivered" was not happening when checking to see if an
2405 address could be batched with a previous (undelivered) one; under
2406 certain circumstances this could lead to multiple deliveries to the same
2409 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2410 in its include files, and this causes problems building Exim.
2412 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2413 header_syntax) but Exim was just ignoring anything given after a slash.
2414 In particular, this caused confusion with an attempt to use "verify =
2415 reverse_host_lookup/defer_ok". An error is now given when options are
2416 supplied for verify items that do not have them. (Maybe reverse_host_
2417 lookup should have a defer_ok option, but that's a different point.)
2419 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2420 defined by RFC 821) to 2048, because there were problems with some AUTH
2421 commands, and RFC 1869 says the size should be increased for extended
2422 SMTP commands that take arguments.
2424 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2427 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2428 "unknown" error; now it says that the functionality isn't in the binary.
2430 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2431 an address' error message when a string expansion fails (syntax or
2432 whatever). Otherwise the password may appear in the log. Following change
2433 PH/42 below, there is no longer a chance of it appearing in a bounce
2436 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2438 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2439 of its addresses was ever tried. (Bugzilla bug #2.)
2441 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2442 the result incorrectly in the debug output. (It correctly added a newline
2443 to what was transported.)
2445 TF/01 Added $received_time.
2447 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2448 commented out examples of how to interface to a virus scanner and to
2449 SpamAssassin. Also added commented examples of av_scanner and
2450 spamd_address settings.
2452 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2453 and controls are allowed in which ACLs. There were a couple of minor
2454 errors. Some of the entries in the conditions table (which is a table of
2455 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2456 negation of where the condition IS allowed.
2458 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2460 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2461 header file does not have a version number, so I've had to invent a new
2462 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2463 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2464 radiusclient), but it was contributed by a Radius user.
2466 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2467 files or directories, for OpenSSL.
2469 PH/19 When an Exim process that is running as root has to create an Exim log
2470 file, it does so in a subprocess that runs as exim:exim so as to get the
2471 ownership right at creation (otherwise, other Exim processes might see
2472 the file with the wrong ownership). There was no test for failure of this
2473 fork() call, which would lead to the process getting stuck as it waited
2474 for a non-existent subprocess. Forks do occasionally fail when resources
2475 run out. I reviewed all the other calls to fork(); they all seem to check
2478 PH/20 When checking for unexpected SMTP input at connect time (before writing
2479 the banner), Exim was not dealing correctly with a non-positive return
2480 from the read() function. If the client had disconnected by this time,
2481 the result was a log entry for a synchronization error with an empty
2482 string after "input=" when read() returned zero. If read() returned -1
2483 (an event I could not check), uninitialized data bytes were printed.
2484 There were reports of junk text (parts of files, etc) appearing after
2487 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2489 PH/22 Added support for macro redefinition, and (re)definition in between
2490 driver and ACL definitions.
2492 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2493 forgetting to use the resulting value; it was using the unexpanded value.
2495 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2496 hadn't been configured. The fix is from Juergen Kreileder, who
2497 understands it better than I do:
2499 "Here's what I see happening with three configured cyrus_sasl
2500 authenticators configured (plain, login, cram-md5):
2502 On startup auth_cyrus_sasl_init() gets called for each of these.
2503 This means three calls to sasl_listmech() without a specified mech_list.
2504 => SASL tests which mechs of all available mechs actually work
2505 => three warnings about OTP not working
2506 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2508 With the patch, sasl_listmech() also gets called three times. But now
2509 SASL's mech_list option is set to the server_mech specified in the the
2510 authenticator. Or in other words, the answer from sasl_listmech()
2511 gets limited to just the mech you're testing for (which is different
2513 => the return list contains just 'plain' or 'login', 'cram-md5' or
2514 nothing depending on the value of ob->server_mech.
2516 I've just tested the patch: Authentication still works fine,
2517 unavailable mechs specified in the exim configuration are still
2518 caught, and the auth.log warnings about OTP are gone."
2520 PH/25 When debugging is enabled, the contents of the command line are added
2521 to the debugging output, even when log_selector=+arguments is not
2524 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2525 answer is "GNU", and only if the return is "GNU/something" is the answer
2528 PH/27 $acl_verify_message is now set immediately after the failure of a
2529 verification in an ACL, and so is available in subsequent modifiers. In
2530 particular, the message can be preserved by coding like this:
2532 warn !verify = sender
2533 set acl_m0 = $acl_verify_message
2535 Previously, $acl_verify_message was set only while expanding "message"
2536 and "log_message" when a very denied access.
2538 PH/28 Modified OS/os.c-Linux with
2540 -#ifndef OS_LOAD_AVERAGE
2541 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2543 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2544 nomenclature these days.)
2546 PH/29 Installed patch from the Sieve maintainer that adds the options
2547 sieve_useraddress and sieve_subaddress to the redirect router.
2549 PH/30 In these circumstances:
2550 . Two addresses routed to the same list of hosts;
2551 . First host does not offer TLS;
2552 . First host accepts first address;
2553 . First host gives temporary error to second address;
2554 . Second host offers TLS and a TLS session is established;
2555 . Second host accepts second address.
2556 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2557 and peerdn, if requested) that were in fact used only for the second
2560 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2561 attention to any local part prefix or suffix that was matched by the
2562 router that accepted the address. It now behaves in the same way as it
2563 does for delivery: the affixes are removed from the local part unless
2564 rcpt_include_affixes is set on the transport.
2566 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2567 timeout during the DATA phase of an incoming message.
2569 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2570 applied a patch sanctioned by the Sieve maintainer.
2572 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2573 the uid or gid is negative. A case of a negative gid caused this to be
2574 noticed. The fix allows for either to be negative.
2576 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2577 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2580 PH/36 The change PH/12 above was broken. Fixed it.
2582 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2583 the grounds that routing the same address twice would always produce the
2584 same answer. This might have been true once, but it is certainly no
2585 longer true now. Routing a child address may depend on the previous
2586 routing that produced that child. Some complicated redirection strategies
2587 went wrong when messages had multiple recipients, and made Exim's
2588 behaviour dependent on the order in which the addresses were given.
2590 I have moved the duplicate checking until after the routing is complete.
2591 Exim scans the addresses that are assigned to local and remote
2592 transports, and removes any duplicates. This means that more work will be
2593 done, as duplicates will always all be routed, but duplicates are
2594 presumably rare, so I don't expect this is of any significance.
2596 For deliveries to pipes, files, and autoreplies, the duplicate checking
2597 still happens during the routing process, since they are not going to be
2600 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2601 It corrects a timeout issue with spamd. This is Ian's comment: "The
2602 background is that sometimes spamd either never reads data from a
2603 connection it has accepted, or it never writes response data. The exiscan
2604 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2605 blindly assumes that writes won't block so it may never time out."
2607 PH/39 Allow G after quota size as well as K and M.
2609 PH/40 The value set for $authenticated_id in an authenticator may not contain
2610 binary zeroes or newlines because the value is written to log lines and
2611 to spool files. There was no check on this. Now the value is run through
2612 the string_printing() function so that such characters are converted to
2613 printable escape sequences.
2615 PH/41 $message_linecount is a new variable that contains the total number of
2616 lines in the message. Compare $body_linecount, which is the count for the
2619 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2620 bounce and delay warning messages, except in certain special cases, which
2623 (a) An SMTP error message from a remote host;
2624 (b) A message specified in a :fail: redirection;
2625 (c) A message specified in a "fail" command in a system filter;
2626 (d) A message specified in a FAIL return from the queryprogram router;
2627 (e) A message specified by the cannot_route_message router option.
2629 In these cases only, Exim does include the error details in bounce and
2630 warning messages. There are also a few cases where bland messages such
2631 as "unrouteable address" or "local delivery error" are given.
2633 PH/43 $value is now also set for the "else" part of a ${run expansion.
2635 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2636 being worked on, but at least Exim now implements the latest version to
2639 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2640 process to complete was treated as a delivery failure, a timeout while
2641 writing the message to the pipe was logged, but erroneously treated as a
2642 successful delivery. Such timeouts include transport filter timeouts. For
2643 consistency with the overall process timeout, these timeouts are now
2644 treated as errors, giving rise to delivery failures by default. However,
2645 there is now a new Boolean option for the pipe transport called
2646 timeout_defer, which, if set TRUE, converts the failures into defers for
2647 both kinds of timeout. A transport filter timeout is now identified in
2650 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2651 systems where "make" and "gmake" are different, calling "gmake" at top
2652 level broke things. I've arranged for the value of $(MAKE) to be passed
2653 from the Makefile to this script so that it can call the same version of
2657 A note about Exim versions 4.44 and 4.50
2658 ----------------------------------------
2660 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2661 changes of various kinds. As a consequence, a big documentation update was
2662 needed. This delayed the release for rather longer than seemed good, especially
2663 in the light of a couple of (minor) security issues. Therefore, the changes
2664 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2665 release. So 4.44 and 4.50 are in effect two different branches that both start
2668 I have left the 4.50 change log unchanged; it contains all the changes since
2669 4.43. The change log for 4.44 is below; many of its items are identical to
2670 those for 4.50. This seems to be the most sensible way to preserve the
2671 historical information.
2677 1. Minor wording change to the doc/README.SIEVE file.
2679 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2680 computation of the current number of files was incorrect.
2682 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2683 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2686 4. Give more explanation in the error message when the command for a transport
2687 filter fails to execute.
2689 5. There are several places where Exim runs a non-Exim command in a
2690 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2691 was being done only for the command run by the queryprogram router. It is
2692 now done for all such subprocesses. The other cases are: ${run, transport
2693 filters, and the commands run by the lmtp and pipe transports.
2695 6. Added CONFIGURE_GROUP build-time option.
2697 7. Some older OS have a limit of 256 on the maximum number of file
2698 descriptors. Exim was using setrlimit() to set 1000 as a large value
2699 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2700 systems. I've change it so that if it can't get 1000, it tries for 256.
2702 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2703 was an oversight, and furthermore, ever since the addition of extra
2704 controls (e.g. 4.43/32), the checks on when to allow different forms of
2705 "control" were broken. There should now be diagnostics for all cases when a
2706 control that does not make sense is encountered.
2708 9. Added the /retain_sender option to "control=submission".
2710 10. $recipients is now available in the predata ACL (oversight).
2712 11. Tidy the search cache before the fork to do a delivery from a message
2713 received from the command line. Otherwise the child will trigger a lookup
2714 failure and thereby defer the delivery if it tries to use (for example) a
2715 cached ldap connection that the parent has called unbind on.
2717 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2718 of $address_data from the recipient verification was clobbered by the
2719 sender verification.
2721 13. The value of address_data from a sender verification is now available in
2722 $sender_address_data in subsequent conditions in the ACL statement.
2724 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2726 15. Added a new option "connect=<time>" to callout options, to set a different
2729 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2730 was its contents. (It was OK if the option was not defined at all.)
2732 17. A "Completed" log line is now written for messages that are removed from
2733 the spool by the -Mrm option.
2735 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2736 information about exactly what failed.
2738 19. Added -dd to debug only the daemon process.
2740 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2741 handles timeouts, both on the server side and network timeouts. Renamed the
2742 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2744 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2745 It is now set to "smtps".
2747 22. $host_address is now set to the target address during the checking of
2748 ignore_target_hosts.
2750 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2751 being passed; this would have caused $sender_host_name to have been used if
2752 matching the list had actually called for a host name (not very likely,
2753 since this list is usually IP addresses). A host name is now passed as
2756 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2757 code that specifies a non-restarting handler (typically sigaction() in
2758 modern systems) in an attempt to fix a rare and obscure crash bug.
2760 25. Narrowed the window for a race in the daemon that could cause it to ignore
2761 SIGCHLD signals. This is not a major problem, because they are used only to
2762 wake it up if nothing else does.
2764 26. A malformed maildirsize file could cause Exim to calculate negative values
2765 for the mailbox size or file count. Odd effects could occur as a result.
2766 The maildirsize information is now recalculated if the size or filecount
2769 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2770 support for a long time. Removed HAVE_SYS_VFS_H.
2772 28. Installed the latest version of exipick from John Jetmore.
2774 29. In an address list, if the pattern was not a regular expression, an empty
2775 subject address (from a bounce message) matched only if the pattern was an
2776 empty string. Non-empty patterns were not even tested. This was the wrong
2777 because it is perfectly reasonable to use an empty address as part of a
2778 database query. An empty address is now tested by patterns that are
2779 lookups. However, all the other forms of pattern expect the subject to
2780 contain a local part and a domain, and therefore, for them, an empty
2781 address still always fails if the pattern is not itself empty.
2783 30. Exim went into a mad DNS loop when attempting to do a callout where the
2784 host was specified on an smtp transport, and looking it up yielded more
2785 than one IP address.
2787 31. Re-factored the code for checking spool and log partition space into a
2788 function that finds that data and another that does the check. The former
2789 is then used to implement four new variables: $spool_space, $log_space,
2790 $spool_inodes, and $log_inodes.
2792 32. The RFC2047 encoding function was originally intended for short strings
2793 such as real names; it was not keeping to the 75-character limit for
2794 encoded words that the RFC imposes. It now respects the limit, and
2795 generates multiple encoded words if necessary. To be on the safe side, I
2796 have increased the buffer size for the ${rfc2047: expansion operator from
2799 33. It is now permitted to omit both strings after an "if" condition; if the
2800 condition is true, the result is "true". As before, when the second string
2801 is omitted, a false condition yields an empty string. This makes it less
2802 cumbersome to write custom ACL and router conditions.
2804 34. Failure to deliver a bounce message always caused it to be frozen, even if
2805 there was an errors_to setting on the router. The errors_to setting is now
2808 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2809 canonical form (fully expanded) before being placed in
2810 $sender_host_address.
2812 36. The table in the code that translates DNS record types into text (T_A to
2813 "A" for instance) was missing entries for NS and CNAME. It is just possible
2814 that this could have caused confusion if both these types were looked up
2815 for the same domain, because the text type is used as part of Exim's
2816 per-process caching. But the chance of anyone hitting this buglet seems
2819 37. The dnsdb lookup has been extended in a number of ways.
2821 (1) There is a new type, "zns", which walks up the domain tree until it
2822 finds some nameserver records. It should be used with care.
2824 (2) There is a new type, "mxh", which is like "mx" except that it returns
2825 just the host names, not the priorities.
2827 (3) It is now possible to give a list of domains (or IP addresses) to be
2828 looked up. The behaviour when one of the lookups defers can be
2829 controlled by a keyword.
2831 (4) It is now possible to specify the separator character for use when
2832 multiple records are returned.
2834 38. The dnslists ACL condition has been extended: it is now possible to supply
2835 a list of IP addresses and/or domains to be looked up in a particular DNS
2838 39. Added log_selector=+queue_time_overall.
2840 40. When running the queue in the test harness, wait just a tad after forking a
2841 delivery process, to get repeatability of debugging output.
2843 41. Include certificate and key file names in error message when GnuTLS fails
2844 to set them up, because the GnuTLS error message doesn't include the name
2845 of the failing file when there is a problem reading it.
2847 42. Allow both -bf and -bF in the same test run.
2849 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2851 44. The "Exiscan patch" is now merged into the mainline Exim source.
2853 45. Sometimes the final signoff response after QUIT could fail to get
2854 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2855 < 0 before doing a fflush(). This bug looks as though it goes back to the
2856 introduction of TLS in release 3.20, but "sometimes" must have been rare
2857 because the tests only now provoked it.
2859 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2860 (this can affect the format of dates).
2862 47. exim_tidydb, when checking for the continued existence of a message for
2863 which it has found a message-specific retry record, was not finding
2864 messages that were in split spool directories. Consequently, it was
2865 deleting retry records that should have stayed in existence.
2867 48. Steve fixed some bugs in eximstats.
2869 49. The SPA authentication driver was not abandoning authentication and moving
2870 on to the next authenticator when an expansion was forced to fail,
2871 contradicting the general specification for all authenticators. Instead it
2872 was generating a temporary error. It now behaves as specified.
2874 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2875 (the order specifies the preference for clients). The order is now AES256,
2876 AES128, 3DES, ARCFOUR128.
2878 51. Small patch to Sieve code - explicitly set From: when generating an
2881 52. Exim crashed if a remote delivery caused a very long error message to be
2882 recorded - for instance if somebody sent an entire SpamAssassin report back
2883 as a large number of 550 error lines. This bug was coincidentally fixed by
2884 increasing the size of one of Exim's internal buffers (big_buffer) that
2885 happened as part of the Exiscan merge. However, to be on the safe side, I
2886 have made the code more robust (and fixed the comments that describe what
2889 53. Now that there can be additional text after "Completed" in log lines (if
2890 the queue_time_overall log selector is set), a one-byte patch to exigrep
2891 was needed to allow it to recognize "Completed" as not the last thing in
2894 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2895 patch that reportedly fixes this has been added. I am not expert enough to
2896 create a test for it. This is what the patch creator wrote:
2898 "I found a little strange behaviour of ldap code when working with
2899 Windows 2003 AD Domain, where users was placed in more than one
2900 Organization Units. When I tried to give exim partial DN, the exit code
2901 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2902 But simultaneously result of request was absolutely normal ldap result,
2903 so I produce this patch..."
2905 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2906 REFERENCE, so I have modified the code to exclude the patch when that macro
2909 55. Some experimental protocols are using DNS PTR records for new purposes. The
2910 keys for these records are domain names, not reversed IP addresses. The
2911 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2912 leaves it alone. Component reversal etc. now happens only for IP addresses.
2915 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2917 57. Double the size of the debug message buffer (to 2048) so that more of very
2918 long debug lines gets shown.
2920 58. The exicyclog utility now does better if the number of log files to keep
2921 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2923 59. Two changes related to the smtp_active_hostname option:
2925 (1) $smtp_active_hostname is now available as a variable.
2926 (2) The default for smtp_banner uses $smtp_active_hostname instead
2927 of $primary_hostname.
2929 60. The host_aton() function is supposed to be passed a string that is known
2930 to be a valid IP address. However, in the case of IPv6 addresses, it was
2931 not checking this. This is a hostage to fortune. Exim now panics and dies
2932 if the condition is not met. A case was found where this could be provoked
2933 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2934 components; fortuitously, this particular loophole had already been fixed
2935 by change 4.50/55 above.
2937 If there are any other similar loopholes, the new check in host_aton()
2938 itself should stop them being exploited. The report I received stated that
2939 data on the command line could provoke the exploit when Exim was running as
2940 exim, but did not say which command line option was involved. All I could
2941 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2942 running as the user.
2945 61. There was a buffer overflow vulnerability in the SPA authentication code
2946 (which came originally from the Samba project). I have added a test to the
2947 spa_base64_to_bits() function which I hope fixes it.
2950 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2951 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2953 63. The daemon start-up calls getloadavg() while still root for those OS that
2954 need the first call to be done as root, but it missed one case: when
2955 deliver_queue_load_max is set with deliver_drop_privilege. This is
2956 necessary for the benefit of the queue runner, because there is no re-exec
2957 when deliver_drop_privilege is set.
2959 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2960 This has been fixed.
2962 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2963 was in use, was not putting the data itself into the right store pool;
2964 consequently, it could be overwritten for a subsequent message in the same
2965 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2968 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2970 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2971 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2972 the function were treating the return as a boolean value, which happened to
2973 work because 0=false and not-0=true, but is not correct code.
2975 68. The host_aton() function was not handling scoped IPv6 addresses (those
2976 with, for example, "%eth0" on the end) correctly.
2978 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2979 negated items (that is, ~something) in unsigned ints. Some compilers
2980 apparently mutter when there is no cast.
2982 70. If an address verification called from an ACL failed, and did not produce a
2983 user-specific message (i.e. there was only a "system" message), nothing was
2984 put in $acl_verify_message. In this situation, it now puts the system
2987 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2988 session; change 4.31/43 added the unwanted input to the log line - except
2989 that it did not do this in the start of session case. It now does.
2991 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2992 This is wrong and can cause the other end to generate a synchronization
2993 error if it is another Exim or anything else that does the synchronization
2994 check. A QUIT command is no longer sent after a timeout.
2996 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2997 during host lookups.
2999 74. The defer_ok option of callout verification was not working if it was used
3000 when verifying addresses in header lines, that is, for this case:
3002 verify = header_sender/callout=defer_ok
3004 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3005 those file descriptors could be used for SMTP connections. If anything
3006 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3007 could be sent to the SMTP client, causing chaos. The daemon now opens
3008 stdin, stdout, and stderr to /dev/null when it puts itself into the
3011 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3012 log by default. The user can override this with suitable Perl magic.
3014 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3015 the log message when discard triggers, was not working for the DATA ACL or
3016 for the non-SMTP ACL.
3018 78. Error message wording change in sieve.c.
3020 79. If smtp_accept_max_per_host was set, the number of connections could be
3021 restricted to fewer than expected, because the daemon was trying to set up
3022 a new connection before checking whether the processes handling previous
3023 connections had finished. The check for completed processes is now done
3024 earlier. On busy systems, this bug wouldn't be noticed because something
3025 else would have woken the daemon, and it would have reaped the completed
3028 80. If a message was submitted locally by a user whose login name contained one
3029 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3030 It caused a spool format error. I have fixed the spool reading code. A
3031 related problem was that the "from" clause in the Received: line became
3032 illegal because of the space(s). It is now covered by ${quote_local_part.
3034 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3037 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3039 83. Patch from Sieve maintainer to fix unterminated string problem in
3040 "vacation" handling.
3042 84. Some minor changes to the Linux configuration files to help with other
3043 OS variants using glibc.
3045 85. One more patch for Sieve to update vacation handling to latest spec.
3048 ----------------------------------------------------
3049 See the note above about the 4.44 and 4.50 releases.
3050 ----------------------------------------------------
3056 1. Change 4.43/35 introduced a bug that caused file counts to be
3057 incorrectly computed when quota_filecount was set in an appendfile
3060 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3061 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3064 3. Give more explanation in the error message when the command for a transport
3065 filter fails to execute.
3067 4. There are several places where Exim runs a non-Exim command in a
3068 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3069 was being done only for the command run by the queryprogram router. It is
3070 now done for all such subprocesses. The other cases are: ${run, transport
3071 filters, and the commands run by the lmtp and pipe transports.
3073 5. Some older OS have a limit of 256 on the maximum number of file
3074 descriptors. Exim was using setrlimit() to set 1000 as a large value
3075 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3076 systems. I've change it so that if it can't get 1000, it tries for 256.
3078 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3079 was an oversight, and furthermore, ever since the addition of extra
3080 controls (e.g. 4.43/32), the checks on when to allow different forms of
3081 "control" were broken. There should now be diagnostics for all cases when a
3082 control that does not make sense is encountered.
3084 7. $recipients is now available in the predata ACL (oversight).
3086 8. Tidy the search cache before the fork to do a delivery from a message
3087 received from the command line. Otherwise the child will trigger a lookup
3088 failure and thereby defer the delivery if it tries to use (for example) a
3089 cached ldap connection that the parent has called unbind on.
3091 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3092 of $address_data from the recipient verification was clobbered by the
3093 sender verification.
3095 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3096 was its contents. (It was OK if the option was not defined at all.)
3098 11. A "Completed" log line is now written for messages that are removed from
3099 the spool by the -Mrm option.
3101 12. $host_address is now set to the target address during the checking of
3102 ignore_target_hosts.
3104 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3105 being passed; this would have caused $sender_host_name to have been used if
3106 matching the list had actually called for a host name (not very likely,
3107 since this list is usually IP addresses). A host name is now passed as
3110 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3111 code that specifies a non-restarting handler (typically sigaction() in
3112 modern systems) in an attempt to fix a rare and obscure crash bug.
3114 15. Narrowed the window for a race in the daemon that could cause it to ignore
3115 SIGCHLD signals. This is not a major problem, because they are used only to
3116 wake it up if nothing else does.
3118 16. A malformed maildirsize file could cause Exim to calculate negative values
3119 for the mailbox size or file count. Odd effects could occur as a result.
3120 The maildirsize information is now recalculated if the size or filecount
3123 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3124 support for a long time. Removed HAVE_SYS_VFS_H.
3126 18. Updated exipick to current release from John Jetmore.
3128 19. Allow an empty sender to be matched against a lookup in an address list.
3129 Previously the only cases considered were a regular expression, or an
3132 20. Exim went into a mad DNS lookup loop when doing a callout where the
3133 host was specified on the transport, if the DNS lookup yielded more than
3136 21. The RFC2047 encoding function was originally intended for short strings
3137 such as real names; it was not keeping to the 75-character limit for
3138 encoded words that the RFC imposes. It now respects the limit, and
3139 generates multiple encoded words if necessary. To be on the safe side, I
3140 have increased the buffer size for the ${rfc2047: expansion operator from
3143 22. Failure to deliver a bounce message always caused it to be frozen, even if
3144 there was an errors_to setting on the router. The errors_to setting is now
3147 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3148 canonical form (fully expanded) before being placed in
3149 $sender_host_address.
3151 24. Updated eximstats to version 1.33
3153 25. Include certificate and key file names in error message when GnuTLS fails
3154 to set them up, because the GnuTLS error message doesn't include the name
3155 of the failing file when there is a problem reading it.
3157 26. Expand error message when OpenSSL has problems setting up cert/key files.
3160 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3161 (this can affect the format of dates).
3163 28. exim_tidydb, when checking for the continued existence of a message for
3164 which it has found a message-specific retry record, was not finding
3165 messages that were in split spool directories. Consequently, it was
3166 deleting retry records that should have stayed in existence.
3168 29. eximstats updated to version 1.35
3169 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3170 1.35 - bugfix such that pie charts by volume are generated correctly
3172 30. The SPA authentication driver was not abandoning authentication and moving
3173 on to the next authenticator when an expansion was forced to fail,
3174 contradicting the general specification for all authenticators. Instead it
3175 was generating a temporary error. It now behaves as specified.
3177 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3178 (the order specifies the preference for clients). The order is now AES256,
3179 AES128, 3DES, ARCFOUR128.
3181 31. Small patch to Sieve code - explicitly set From: when generating an
3184 32. Exim crashed if a remote delivery caused a very long error message to be
3185 recorded - for instance if somebody sent an entire SpamAssassin report back
3186 as a large number of 550 error lines. This bug was coincidentally fixed by
3187 increasing the size of one of Exim's internal buffers (big_buffer) that
3188 happened as part of the Exiscan merge. However, to be on the safe side, I
3189 have made the code more robust (and fixed the comments that describe what
3192 33. Some experimental protocols are using DNS PTR records for new purposes. The
3193 keys for these records are domain names, not reversed IP addresses. The
3194 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3195 leaves it alone. Component reversal etc. now happens only for IP addresses.
3198 34. The host_aton() function is supposed to be passed a string that is known
3199 to be a valid IP address. However, in the case of IPv6 addresses, it was
3200 not checking this. This is a hostage to fortune. Exim now panics and dies
3201 if the condition is not met. A case was found where this could be provoked
3202 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3203 components; fortuitously, this particular loophole had already been fixed
3204 by change 4.50/55 or 4.44/33 above.
3206 If there are any other similar loopholes, the new check in host_aton()
3207 itself should stop them being exploited. The report I received stated that
3208 data on the command line could provoke the exploit when Exim was running as
3209 exim, but did not say which command line option was involved. All I could
3210 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3211 running as the user.
3214 35. There was a buffer overflow vulnerability in the SPA authentication code
3215 (which came originally from the Samba project). I have added a test to the
3216 spa_base64_to_bits() function which I hope fixes it.
3219 36. The daemon start-up calls getloadavg() while still root for those OS that
3220 need the first call to be done as root, but it missed one case: when
3221 deliver_queue_load_max is set with deliver_drop_privilege. This is
3222 necessary for the benefit of the queue runner, because there is no re-exec
3223 when deliver_drop_privilege is set.
3225 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3226 was in use, was not putting the data itself into the right store pool;
3227 consequently, it could be overwritten for a subsequent message in the same
3228 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3231 38. Sometimes the final signoff response after QUIT could fail to get
3232 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3233 < 0 before doing a fflush(). This bug looks as though it goes back to the
3234 introduction of TLS in release 3.20, but "sometimes" must have been rare
3235 because the tests only now provoked it.
3241 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3242 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3243 Now it returns an integer. A number of calls were still expecting a T/F
3244 return. Fortuitously, in all cases, the tests worked in OK situations,
3245 which is the norm. However, things would have gone wrong on any write
3246 failures on the smtp file descriptor. This function is used when sending
3247 messages over SMTP and also when doing verify callouts.
3249 2. When Exim is called to do synchronous delivery of a locally submitted
3250 message (the -odf or -odi options), it no longer closes stderr before doing
3253 3. Implemented the mua_wrapper option.
3255 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3257 5. Implemented the functions header_remove(), header_testname(),
3258 header_add_at_position(), and receive_remove_recipient(), and exported them
3261 6. If an ACL "warn" statement specified the addition of headers, Exim already
3262 inserted X-ACL-Warn: at the start if there was no header name. However, it
3263 was not making this test for the second and subsequent header lines if
3264 there were newlines in the string. This meant that an invalid header could
3265 be inserted if Exim was badly configured.
3267 7. Allow an ACL "warn" statement to add header lines at the start or after all
3268 the Received: headers, as well as at the end.
3270 8. Added the rcpt_4xx retry error code.
3272 9. Added postmaster_mailfrom=xxx to callout verification option.
3274 10. Added mailfrom=xxxx to the callout verification option, for verify=
3277 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3278 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3280 12. Inserted some casts to stop certain compilers warning when using pointer
3281 differences as field lengths or precisions in printf-type calls (mostly
3282 affecting debugging statements).
3284 13. Added optional readline() support for -be (dynamically loaded).
3286 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3287 same clock tick as a message's arrival, so that its received time was the
3288 same as the "first fail" time on the retry record, and that message
3289 remained on the queue past the ultimate address timeout, every queue runner
3290 would try a delivery (because it was past the ultimate address timeout) but
3291 after another failure, the ultimate address timeout, which should have then
3292 bounced the address, did not kick in. This was a "< instead of <=" error;
3293 in most cases the first failure would have been in the next clock tick
3294 after the received time, and all would be well.
3296 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3297 being recognized when the domain list was tested by the match_domain
3298 condition in an expansion string.
3300 16. Added the ${str2b64: operator.
3302 17. Exim was always calling setrlimit() to set a large limit for the number of
3303 processes, without checking whether the existing limit was already
3304 adequate. (It did check for the limit on file descriptors.) Furthermore,
3305 errors from getrlimit() and setrlimit() were being ignored. Now they are
3306 logged to the main and panic logs, but Exim does carry on, to try to do its
3307 job under whatever limits there are.
3309 18. Imported PCRE 5.0.
3311 19. Trivial typo in log message " temporarily refused connection" (the leading
3314 20. If the log selector return_path_on_delivery was set and an address was
3315 redirected to /dev/null, the delivery process crashed because it assumed
3316 that a return path would always be set for a "successful" delivery. In this
3317 case, the whole delivery is bypassed as an optimization, and therefore no
3320 21. Internal re-arrangement: the function for sending a challenge and reading
3321 a response while authentication was assuming a zero-terminated challenge
3322 string. It's now changed to take a pointer and a length, to allow for
3323 binary data in such strings.
3325 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3327 23. Exim was not respecting finduser_retries when seeking the login of the
3328 uid under which it was called; it was always trying 10 times. (The default
3329 setting of finduser_retries is zero.) Also, it was sleeping after the final
3330 failure, which is pointless.
3332 24. Implemented tls_on_connect_ports.
3334 25. Implemented acl_smtp_predata.
3336 26. If the domain in control=submission is set empty, Exim assumes that the
3337 authenticated id is a complete email address when it generates From: or
3338 Sender: header lines.
3340 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3341 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3342 chown and chgrp in /bin and hostname in /usr/bin.
3344 28. Exim was keeping the "process log" file open after each use, just as it
3345 does for the main log. This opens the possibility of it remaining open for
3346 long periods when the USR1 signal hits a daemon. Occasional processlog
3347 errors were reported, that could have been caused by this. Anyway, it seems
3348 much more sensible not to leave this file open at all, so that is what now
3351 29. The long-running daemon process does not normally write to the log once it
3352 has entered its main loop, and it closes the log before doing so. This is
3353 so that log files can straightforwardly be renamed and moved. However,
3354 there are a couple of unusual error situations where the daemon does write
3355 log entries, and I had neglected to close the log afterwards.
3357 30. The text of an SMTP error response that was received during a remote
3358 delivery was being truncated at 512 bytes. This is too short for some of
3359 the long messages that one sometimes sees. I've increased the limit to
3362 31. It is now possible to make retry rules that apply only when a message has a
3363 specific sender, in particular, an empty sender.
3365 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3366 it possible to be selective about when SMTP synchronization is enforced.
3368 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3370 32. Implemented hosts_connection_nolog.
3372 33. Added an ACL for QUIT.
3374 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3377 35. Added mailbox_size and mailbox_filecount to appendfile.
3379 36. Added control = no_multiline_responses to ACLs.
3381 37. There was a bug in the logic of the code that waits for the clock to tick
3382 in the case where the clock went backwards by a substantial amount such
3383 that the microsecond fraction of "now" was more than the microsecond
3384 fraction of "then" (but the whole seconds number was less).
3386 38. Added support for the libradius Radius client library this is found on
3387 FreeBSD (previously only the radiusclient library was supported).
3393 1. When certain lookups returned multiple values in the form name=value, the
3394 quoting of the values was not always being done properly. Specifically:
3395 (a) If the value started with a double quote, but contained no whitespace,
3397 (b) If the value contained whitespace other than a space character (i.e.
3398 tabs or newlines or carriage returns) it was not quoted.
3399 This fix has been applied to the mysql and pgsql lookups by writing a
3400 separate quoting function and calling it from the lookup code. The fix
3401 should probably also be applied to nisplus, ibase and oracle lookups, but
3402 since I cannot test any of those, I have not disturbed their existing code.
3404 2. A hit in the callout cache for a specific address caused a log line with no
3405 reason for rejecting RCPT. Now it says "Previous (cached) callout
3406 verification failure".
3408 3. There was an off-by-one bug in the queryprogram router. An over-long
3409 return line was truncated at 256 instead of 255 characters, thereby
3410 overflowing its buffer with the terminating zero. As well as fixing this, I
3411 have increased the buffer size to 1024 (and made a note to document this).
3413 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3414 when Exim is waiting for an SMTP response from a remote server, Exim
3415 restarts its select() call on the socket, thereby resetting its timeout.
3416 This is not a problem when such interrupts are rare. Somebody set up a cron
3417 job to run exiwhat every 2 minutes, which is less than the normal select()
3418 timeout (5 or 10 minutes). This meant that the select() timeout never
3419 kicked in because it was always reset. I have fixed this by comparing the
3420 time when an interrupt arrives with the time at the start of the first call
3421 to select(). If more time than the timeout has elapsed, the interrupt is
3422 treated as a timeout.
3424 5. Some internal re-factoring in preparation for the addition of Sieve
3425 extensions (by MH). In particular, the "personal" test is moved to a
3426 separate function, and given an option for scanning Cc: and Bcc: (which is
3427 not set for Exim filters).
3429 6. When Exim created an email address using the login of the caller as the
3430 local part (e.g. when creating a From: or Sender: header line), it was not
3431 quoting the local part when it contained special characters such as @.
3433 7. Installed new OpenBSD configuration files.
3435 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3436 try to make them clearer.
3438 9. Callout options, other than the timeout value, were being ignored when
3439 verifying sender addresses in header lines. For example, when using
3441 verify = header_sender/callout=no_cache
3443 the cache was (incorrectly) being used.
3445 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3446 only the Cygwin environment.
3448 11. When return_path_on_delivery was set as a log selector, if different remote
3449 addresses in the same message used different return paths and parallel
3450 remote delivery occurred, the wrong values would sometimes be logged.
3451 (Whenever a remote delivery process finished, the return path value from
3452 the most recently started remote delivery process was logged.)
3454 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3455 lines when AUTH and/or TLS are in use. This is the "received protocol"
3456 field. Exim used to use "asmtp" for authenticated SMTP, without any
3457 indication (in the protocol name) for TLS use. Now it follows the RFC and
3458 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3459 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3460 names appear in log lines as well as in Received: header lines.
3462 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3463 extensions, and comparison tests, and to fix some bugs.
3465 14. Changes to the "personal" filter test:
3467 (1) The test was buggy in that it was just doing the equivalent of
3468 "contains" tests on header lines. For example, if a user's address was
3469 anne@some.where, the "personal" test would incorrectly be true for
3471 To: susanne@some.where
3473 This test is now done by extracting each address from the header in turn,
3474 and checking the entire address. Other tests that are part of "personal"
3475 are now done using regular expressions (for example, to check local parts
3476 of addresses in From: header lines).
3478 (2) The list of non-personal local parts in From: addresses has been
3479 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3480 taken from the Sieve specification recommendations.
3482 (3) If the message contains any header line starting with "List-" it is
3483 treated as non-personal.
3485 (4) The test for "circular" in the Subject: header line has been removed
3486 because it now seems ill-conceived.
3488 15. Minor typos in src/EDITME comments corrected.
3490 16. Installed latest exipick from John Jetmore.
3492 17. If headers_add on a router specified a text string that was too long for
3493 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3494 of string_sprintf() is now avoided.
3496 18. $message_body_size was not set (it was always zero) when running the DATA
3497 ACL and the local_scan() function.
3499 19. For the "mail" command in an Exim filter, no default was being set for
3500 the once_repeat time, causing a random time value to be used if "once" was
3501 specified. (If the value happened to be <= 0, no repeat happened.) The
3502 default is now 0s, meaning "never repeat". The "vacation" command was OK
3503 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3504 (I found it when inspecting the code).
3506 20. There is now an overall timeout for performing a callout verification. It
3507 defaults to 4 times the callout timeout, which applies to individual SMTP
3508 commands during the callout. The overall timeout applies when there is more
3509 than one host that can be tried. The timeout is checked before trying the
3510 next host. This prevents very long delays if there are a large number of
3511 hosts and all are timing out (e.g. when the network connections are timing
3512 out). The value of the overall timeout can be changed by specifying an
3513 additional sub-option for "callout", called "maxwait". For example:
3515 verify = sender/callout=5s,maxwait=20s
3517 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3518 to the end before writing, but this should make it even safer).
3520 22. Exim was forgetting that it had advertised PIPELINING for the second and
3521 subsequent messages on an SMTP connection. It was also not resetting its
3522 memory on STARTTLS and an internal HELO.
3524 23. When Exim logs an SMTP synchronization error within a session, it now
3525 records whether PIPELINING has been advertised or not.
3527 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3528 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3529 rather than long int.
3531 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3533 26. Added the never_mail option to autoreply.
3539 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3540 crash if the getsockname() call failed; this can happen if a connection is
3541 closed very soon after it is established. The problem was simply in the
3542 order in which certain operations were done, causing Exim to try to write
3543 to the SMTP stream before it had set up the file descriptor. The bug has
3544 been fixed by making things happen in the correct order.
3550 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3551 before the connection was closed, thus losing the rejection response.
3553 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3554 some early Solaris releases, but causes trouble in current releases where
3555 socklen_t is defined.
3557 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3560 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3563 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3564 that releasing the top of it at the end releases what was used for sub-
3565 expansions (unless the block got too big). However, discard this block if
3566 the first thing is a variable or header, so that we can use its block when
3567 it is dynamic (useful for very large $message_headers, for example).
3569 6. Lookups now cache *every* query, not just the most recent. A new, separate
3570 store pool is used for this. It can be recovered when all lookup caches are
3571 flushed. Lookups now release memory at the end of their result strings.
3572 This has involved some general refactoring of the lookup sources.
3574 7. Some code has been added to the store_xxx() functions to reduce the amount
3575 of flapping under certain conditions.
3577 8. log_incoming_interface used to affect only the <= reception log lines. Now
3578 it causes the local interface and port to be added to several more SMTP log
3579 lines, for example "SMTP connection from", and rejection lines.
3581 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3583 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3585 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3586 could be overwritten at the end of the current message (or the start of a
3587 new message if it was set in a HELO ACL). The value is now preserved for
3588 the duration of the SMTP connection.
3590 12. If a transport had a headers_rewrite setting, and a matching header line
3591 contained an unqualified address, that address was qualified, even if it
3592 did not match any rewriting rules. The underlying bug was that the values
3593 of the flags that permit the existence of unqualified sender and recipient
3594 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3595 non-local messages, and by -bnq for local messages) were not being
3596 preserved with the message after it was received.
3598 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3599 "next input=" as part of the text comprising the host identity instead of
3600 the correct text. The code was using the same buffer for two different
3601 strings. However, depending on which order the printing function evaluated
3602 its arguments, the bug did not always show up. Under Linux, for example, my
3603 test suite worked just fine.
3605 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3606 doesn't work with some older versions of Perl. It has been changed to "my",
3607 which in any case is probably the better facility to use.
3609 15. A really picky compiler found some instances of statements for creating
3610 error messages that either had too many or two few arguments for the format
3613 16. The size of the buffer for calls to the DNS resolver has been increased
3614 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3615 for addresses that have a lot of PTR records. This alleviates a problem; it
3616 does not fully solve it.
3618 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3619 buffer now truncates the list and logs the incident, which is the same
3620 action as happens when Exim is looking up a host name and its aliases.
3621 Previously in this situation something unpredictable would happen;
3622 sometimes it was "internal error: store_reset failed".
3624 18. If a server dropped the connection unexpectedly when an Exim client was
3625 using GnuTLS and trying to read a response, the client delivery process
3626 crashed while trying to generate an error log message.
3628 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3629 string, for example:
3631 warn message = H1: something\nH2: something
3633 the text was added as a single header line from Exim's point of view
3634 though it ended up OK in the delivered message. However, searching for the
3635 second and subsequent header lines using $h_h2: did not work. This has been
3636 fixed. Similarly, if a system filter added multiple headers in this way,
3637 the routers could not see them.
3639 20. Expanded the error message when iplsearch is called with an invalid key to
3640 suggest using net-iplsearch in a host list.
3642 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3643 ACLs are no longer actually imposed (and a message to that effect is
3646 22. If a "gecos" field in a passwd entry contained escaped characters, in
3647 particular, if it contained a \" sequence, Exim got it wrong when building
3648 a From: or a Sender: header from that name. A second bug also caused
3649 incorrect handling when an unquoted " was present following a character
3650 that needed quoting.
3652 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3653 was not being matched caselessly.
3655 24. Arranged for all hyphens in the exim.8 source to be escaped with
3658 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3659 properly. Recipient callout cache records were still being keyed to include
3660 the sender, even when use_sender was set false. This led to far more
3661 callouts that were necessary. The sender is no longer included in the key
3662 when use_sender is false.
3664 26. Added "control = submission" modifier to ACLs.
3666 27. Added the ${base62d: operator to decode base 62 numbers.
3668 28. dnsdb lookups can now access SRV records.
3670 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3671 the configuration file.
3673 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3674 (-v) mode. This makes the output for a verbose queue run more intelligible.
3676 31. Added a use_postmaster feature to recipient callouts.
3678 32. Added the $body_zerocount variable, containing the number of binary zero
3679 bytes in the message body.
3681 33. The time of last modification of the "new" subdirectory is now used as the
3682 "mailbox time last read" when there is a quota error for a maildir
3685 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3687 35. Added +ignore_unknown as a special item in host lists.
3689 36. Code for decoding IPv6 addresses in host lists is now included, even if
3690 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3691 address was recognized as an IP address, but was then not correctly decoded
3692 into binary, causing unexpected and incorrect effects when compared with
3699 1. Very minor rewording of debugging text in manualroute to say "list of
3700 hosts" instead of "hostlist".
3702 2. If verify=header_syntax was set, and a header line with an unqualified
3703 address (no domain) and a large number of spaces between the end of the
3704 name and the colon was received, the reception process suffered a buffer
3705 overflow, and (when I tested it) crashed. This was caused by some obsolete
3706 code that should have been removed. The fix is to remove it!
3708 3. When running in the test harness, delay a bit after writing a bounce
3709 message to get a bit more predictability in the log output.
3711 4. Added a call to search_tidyup() just before forking a reception process. In
3712 theory, someone could use a lookup in the expansion of smtp_accept_max_
3713 per_host which, without the tidyup, could leave open a database connection.
3715 5. Added the variables $recipient_data and $sender_data which get set from a
3716 lookup success in an ACL "recipients" or "senders" condition, or a router
3717 "senders" option, similar to $domain_data and $local_part_data.
3719 6. Moved the writing of debug_print from before to after the "senders" test
3722 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3723 problems for message scanning, either using a data ACL, or using
3724 local_scan() because the Received: header was not generated till after they
3725 were called (in order to set the time as the time of reception completion).
3726 I have revised the way this works. The header is now generated after the
3727 body is received, but before the ACL or local_scan() are called. After they
3728 are run, the timestamp in the header is updated.
3734 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3735 before starting a queue runner without re-exec. This happened only when
3736 deliver_drop_privilege was set or when the Exim user was set to root. The
3737 effect of the bug was that timeouts during subsequent deliveries caused
3738 crashes instead of being properly handled. The handler is now left at its
3739 default (and expected) setting.
3741 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3742 message, again when deliver_drop_privilege is set or Exim is run as root.
3743 The bug described in (1) was not present in this case, but the tidying up
3744 of the other signals was missing. I have made the two cases consistent.
3746 3. The ignore_target_hosts setting on a manualroute router was being ignored
3747 for hosts that were looked up using the /MX notation.
3749 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3752 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3753 operated on the sender address. After changing the $sender_address to <>
3754 for the sender address verify, Exim was re-instated it as the original
3755 (before rewriting) address, but remembering that it had rewritten it, so it
3756 wasn't rewriting it again. This bug also had the effect of breaking the
3757 sender address verification caching when the sender address was rewritten.
3759 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3760 This has been changed so that if the ip literal address matches
3761 ignore_target_hosts, the router declines.
3763 7. Added expansion conditions match_domain, match_address, and match_local_
3764 part (NOT match_host).
3766 8. The placeholder for the Received: header didn't have a length field set.
3768 9. Added code to Exim itself and to exim_lock to test for a specific race
3769 condition that could lead to file corruption when using MBX delivery. The
3770 issue is with the lockfile that is created in /tmp. If this file is removed
3771 after a process has opened it but before that process has acquired a lock,
3772 there is the potential for a second process to recreate the file and also
3773 acquire a lock. This could lead to two Exim processes writing to the file
3774 at the same time. The added code performs the same test as UW imapd; it
3775 checks after acquiring the lock that its file descriptor still refers to
3776 the same named file.
3778 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3779 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3782 11. Added the smtp_active_hostname option. If used, this will typically be made
3783 to depend on the incoming interface address. Because $interface_address is
3784 not set up until the daemon has forked a reception process, error responses
3785 that can happen earlier (such as "too many connections") no longer contain
3788 12. If an expansion in a condition on a "warn" statement fails because a lookup
3789 defers, the "warn" statement is abandoned, and the next ACL statement is
3790 processed. Previously this caused the whole ACL to be aborted.
3792 13. Added the iplsearch lookup type.
3794 14. Added ident_timeout as a log selector.
3796 15. Added tls_certificate_verified as a log selector.
3798 16. Added a global option tls_require_ciphers (compare the smtp transport
3799 option of the same name). This controls incoming TLS connections.
3801 17. I finally figured out how to make tls_require_ciphers do a similar thing
3802 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3803 before starting the TLS session.
3805 18. Tabs are now shown as \t in -bP output.
3807 19. If the log selector return_path_on_delivery was set, Exim crashed when
3808 bouncing a message because it had too many Received: header lines.
3810 20. If two routers both had headers_remove settings, and the first one included
3811 a superfluous trailing colon, the final name in the first list and the
3812 first name in the second list were incorrectly joined into one item (with a
3813 colon in the middle).
3819 1. Added -C and -D options to the exinext utility, mainly to make it easier
3820 to include in the automated testing, but these could be helpful when
3821 multiple configurations are in use.
3823 2. The exinext utility was not formatting the output nicely when there was
3824 an alternate port involved in the retry record key, nor when there was a
3825 message id as well (for retries that were specific to a specific message
3826 and a specific host). It was also confused by IPv6 addresses, because of
3827 the additional colons they contain. I have fixed the IPv4 problem, and
3828 patched it up to do a reasonable job for IPv6.
3830 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3831 delivery, the log line now contains "pipelined" if PIPELINING was used.
3833 4. An SMTP transport process used to panic and die if the bind() call to set
3834 an explicit outgoing interface failed. This has been changed; it is now
3835 treated in the same way as a connect() failure.
3837 5. A reference to $sender_host_name in the part of a conditional expansion
3838 that was being skipped was still causing a DNS lookup. This no longer
3841 6. The def: expansion condition was not recognizing references to header lines
3842 that used bh_ and bheader_.
3844 7. Added the _cache feature to named lists.
3846 8. The code for checking quota_filecount in the appendfile transport was
3847 allowing one more file than it should have been.
3849 9. For compatibility with Sendmail, the command line option
3857 and sets the incoming protocol and host name (for trusted callers). The
3858 host name and its colon can be omitted when only the protocol is to be set.
3859 Note the Exim already has two private options, -pd and -ps, that refer to
3860 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3861 "s", but I don't think that's a major issue.
3863 10. A number of refactoring changes to the code, none of which should affect
3866 (a) The number of logging options was getting close to filling up the
3867 32-bit word that was used as a bit map. I have split them into two classes:
3868 those that are passed in the argument to log_write(), and those that are
3869 only ever tested independently outside of that function. These are now in
3870 separate 32-bit words, so there is plenty of room for expansion again.
3871 There is no change in the user interface or the logging behaviour.
3873 (b) When building, for example, log lines, the code previously used a
3874 macro that called string_cat() twice, in order to add two strings. This is
3875 not really sufficiently general. Furthermore, there was one instance where
3876 it was actually wrong because one of the argument was used twice, and in
3877 one call a function was used. (As it happened, calling the function twice
3878 did not affect the overall behaviour.) The macro has been replaced by a
3879 function that can join an arbitrary number of extra strings onto a growing
3882 (c) The code for expansion conditions now uses a table and a binary chop
3883 instead of a serial search (which was left over from when there were very
3884 few conditions). Also, it now recognizes conditions like "pam" even when
3885 the relevant support is not compiled in: a suitably worded error message is
3886 given if an attempt is made to use such a condition.
3888 11. Added ${time_interval:xxxxx}.
3890 12. A bug was causing one of the ddress fields not to be passed back correctly
3891 from remote delivery subprocesses. The field in question was not being
3892 subsequently used, so this caused to problems in practice.
3894 13. Added new log selectors queue_time and deliver_time.
3896 14. Might have fixed a bug in maildirsizefile handling that threw up
3897 "unexpected character" debug warnings, and recalculated the data
3898 unnecessarily. In any case, I expanded the warning message to give more
3901 15. Added the message "Restricted characters in address" to the statements in
3902 the default ACL that block characters like @ and % in local parts.
3904 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3905 Three changes have been made:
3907 (a) There was a serious bug; a negative response to MAIL caused the whole
3908 recipient domain to be cached as invalid, thereby blocking all messages
3909 to all local parts at the same domain, from all senders. This bug has
3910 been fixed. The domain is no longer cached after a negative response to
3911 MAIL if the sender used is not empty.
3913 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3916 (c) A new callout option, "use_sender" has been added for people who want
3917 the modified behaviour.
3923 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3926 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3927 indeed breaks things for older releases.
3929 3. Added additional logging to the case where there is a problem reading data
3930 from a filter that is running in a subprocess using a pipe, in order to
3931 try to track down a specific problem.
3933 4. Testing facility fudge: when running in the test harness and attempting
3934 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3935 getting "No route to host". Convert this to a timeout.
3937 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3940 6. Some OS don't have socklen_t but use size_t instead. This affects the
3941 fifth argument of getsockopt() amongst other things. This is now
3942 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3943 can be set for individual OS. I have set it for SunOS5, OSF1, and
3944 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3945 some earlier ones do not.
3947 7. Change 4.30/15 was not doing the test caselessly.
3949 8. The standard form for an IPv6 address literal was being rejected by address
3950 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3951 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3952 this, as well as the form without the "IPv6" on the front (but only when
3953 address literals are enabled, of course).
3955 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3957 10. Exim crashed if a message with an empty sender address specified by -f
3958 encountered a router with an errors_to setting. This could be provoked only
3959 by a command such as
3963 where an empty string was supplied; "<>" did not hit this bug.
3965 11. Installed PCRE release 4.5.
3967 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3968 remained set. It is now erased.
3970 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3971 times from message ids (which are base 36 rather than the normal 62).
3973 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3974 were being counted as actual protocol errors, and logged if the log
3975 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3976 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3977 and DATA following a set of rejected RCPTs do not count as protocol errors.
3978 In other words, Exim assumes they were pipelined, though this may not
3979 actually be the case. Of course, in all cases the client gets an
3980 appropriate error code.
3982 15. If a lookup fails in an ACL condition, a message about the failure may
3983 be available; it is used if testing the ACL cannot continue, because most
3984 such messages specify what the cause of the deferral is. However, some
3985 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3986 that caused an old message to be retained and used if a later statement
3987 caused a defer, replacing the real cause of the deferral.
3989 16. If an IP address had so many PTR records that the DNS lookup buffer
3990 was not large enough to hold them, Exim could crash while trying to process
3991 the truncated data. It now detects and logs this case.
3993 17. Further to 4.21/58, another change has been made: if (and only if) the
3994 first line of a message (the first header line) ends with CRLF, a bare LF
3995 in a subsequent header line has a space inserted after it, so as not to
3996 terminate the header.
3998 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3999 unnecessarily, used atoi() instead of strtol(), and didn't check the
4000 termination when getting file sizes from file names by regex.
4002 19. Completely re-implemented the support for maildirsize files, in the light
4003 of a number of problems with the previous contributed implementation
4004 (4.30/29). In particular:
4006 . If the quota is zero, the maildirsize file is maintained, but no quota is
4009 . If the maildir directory does not exist, it is created before any attempt
4010 to write a maildirsize file.
4012 . The quota value in the file is just a cache; if the quota is changed in
4013 the transport, the new value overrides.
4015 . A regular expression is available for excluding directories from the
4018 20. The autoreply transport checks the characters in options that define the
4019 message's headers; it allows continued headers, but it was checking with
4020 isspace() after an embedded newline instead of explicitly looking for a
4023 21. If all the "regular" hosts to which an address was routed had passed their
4024 expiry times, and had not reached their retry times, the address was
4025 bounced, even if fallback hosts were defined. Now Exim should go on to try
4028 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4029 equivalent code in the SMTP transport. Some hosts send humungous responses
4030 to HELO/EHLO, more than 1024 it seems.
4032 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4033 gives compiler warnings in some environments. I've now done it "properly",
4036 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4037 (because of problems with the built-in one) was declared to return uschar *
4038 instead of char *, causing compiler failure.
4040 25. Fixed a file descriptor leak when processing alias/forward files.
4042 26. Fixed a minor format string issue in dbfn.c.
4044 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4046 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4047 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4050 29. When a sender address is verified, it is cached, to save repeating the test
4051 when there is more than one recipient in a message. However, when the
4052 verification involves a callout, it is possible for different callout
4053 options to be set for different recipients. It is too complicated to keep
4054 track of this in the cache, so now Exim always runs a verification when a
4055 callout is required, relying on the callout cache for the optimization.
4056 The overhead is duplication of the address routing, but this should not be
4059 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4060 to be verified with callout=postmaster, and the main callout worked but the
4061 postmaster check failed, the verification correctly failed. However, if a
4062 subsequent RCPT command asked for sender verification *without* the
4063 postmaster check, incorrect caching caused this verification also to fail,
4066 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4067 it was not caching the DNS options (qualify_single, search_parents) that
4068 were used when the lookup failed. A subsequent lookup with different
4069 options therefore always gave the same answer, though there were cases
4070 where it should not have. (Example: a "domains = !$mx_any" option on a
4071 dnslookup router: the "domains" option is always processed without any
4072 widening, but the router might have qualify_single set.) Now Exim uses the
4073 cached value only when the same options are set.
4075 32. Added John Jetmore's "exipick" utility to the distribution.
4077 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4078 than a timeout (e.g. a certificate is required, and is not provided), an
4079 Exim server now closes the connection immediately. Previously it waited for
4080 the client to close - but if the client is SSL, it seems that they each
4081 wait for each other, leading to a delay before one of them times out.
4083 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4084 maintained 0.8.x compatibility because I don't think many are using it, and
4085 it is clearly obsolete.
4087 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4090 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4091 client certificate was expired. A simple patch fixes this, though I don't
4092 understand the full logic of why the verify callback is called multiple
4095 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4096 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4097 which causes problems with some clients (such as the Certicom SSL Plus
4098 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4099 disables the coutermeasure allowing Eudora to connect."
4101 38. Exim was not checking that a write() to a log file succeeded. This could
4102 lead to Bad Things if a log got too big, in particular if it hit a file
4103 size limit. Exim now panics and dies if it cannot write to a log file, just
4104 as it does if it cannot open a log file.
4106 39. Modified OS/Makefile-Linux so that it now contains
4108 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4110 The two -D definitions ensure that Exim is compiled with large file
4111 support, which makes it possible to handle log files that are bigger than
4114 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4115 instance) a domain was checked against a named list that involved a lookup,
4116 causing $domain_data to be set, then another domain was checked against the
4117 same list, then the first domain was re-checked, the value of $domain_data
4118 after the final check could be wrong. In particular, if the second check
4119 failed, it could be set empty. This bug probably also applied to
4122 41. The strip_trailing_dot option was not being applied to the address given
4123 with the -f command-line option.
4125 42. The code for reading a message's header from the spool was incrementing
4126 $received_count, but never initializing it. This meant that the value was
4127 incorrect (doubled) while delivering a message in the same process in which
4128 it was received. In the most common configuration of Exim, this never
4129 happens - a fresh exec is done - but it can happen when
4130 deliver_drop_privilege is set.
4132 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4133 it now includes up to 150 characters of the unexpected data in the log
4136 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4137 and building data strings. The size of both of these buffers was 10 000
4138 bytes - far larger than anybody would *ever* want, thought I. Needless to
4139 say, somebody hit the limit. I have increased the maximum line length to
4140 20 000 and the maximum data length of concatenated lines to 100 000. I have
4141 also fixed two bugs, because there was no checking on these buffers. Tsk,
4142 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4143 buffer is too small.
4145 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4146 lsearch lookups. Now it does.
4148 46. When parsing a route_list item in a manualroute router, a fixed-length
4149 buffer was used for the list of hosts. I made this 1024 bytes long,
4150 thinking that nobody would ever have a list of hosts that long. Wrong.
4151 Somebody had a whole pile of complicated expansion conditions, and the
4152 string was silently truncated, leading to an expansion error. It turns out
4153 that it is easier to change to an unlimited length (owing to other changes
4154 that have happened since this code was originally written) than to build
4155 structure for giving a limitation error. The length of the item that
4156 expands into the list of hosts is now unlimited.
4158 47. The lsearch lookup could not handle data where the length of text line was
4159 more than 4095 characters. Such lines were truncated, leading to shortened
4160 data being returned. It should now handle lines of any length.
4162 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4163 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4166 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4167 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4168 no longer happens in comment lines. A list of replacements is now placed
4169 at the head of all of the source files, except those whose only change is
4170 to replace PERL_COMMAND in the very first #! line.
4172 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4173 messages on the queue, with a bottom-up merge sort, using code contributed
4174 by Michael Haardt. This should make operations like -bp somewhat faster on
4175 large queues. It won't affect queue runners, except when queue_run_in_order
4178 51. Installed eximstats 1.31 in the distribution.
4180 52. Added support for SRV lookups to the dnslookup router.
4182 53. If an ACL referred to $message_body or $message_body_end, the value was not
4183 reset for any messages that followed in the same SMTP session.
4185 54. The store-handling optimization for building very long strings was not
4186 differentiating between the different store pools. I don't think this
4187 actually made any difference in practice, but I've tidied it.
4189 55. While running the routers to verify a sender address, $sender_address
4190 was still set to the sender address. This is wrong, because when routing to
4191 send a bounce to the sender, it would be empty. Therefore, I have changed
4192 it so that, while verifying a sender address, $sender_address is set to <>.
4193 (There is no change to what happens when verifying a recipient address.)
4195 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4196 target A or AAAA records (if not already returned) without resetting the
4197 qualify_single or search_parents options of the DNS resolver. These are
4198 inappropriate in this case because the targets of MX and SRV records must
4199 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4200 target that, when qualified, matched something in the local domain. These
4201 two options are now turned off when doing these lookups.
4203 57. It seems that at least some releases of Reiserfs (which does not have the
4204 concept of a fixed number of inodes) returns zero and not -1 for the
4205 number of available inodes. This interacted badly with check_spool_inodes,
4206 which assumed that -1 was the "no such thing" setting. What I have done is
4207 to check that the total number of inodes is greater than zero before doing
4208 the test of how many are available.
4210 58. When a "warn" ACL statement has a log_message modifier, the message is
4211 remembered, and not repeated. This is to avoid a lot of repetition when a
4212 message has many recipients that cause the same warning to be written.
4213 Howewer, Exim was preserving the list of already written lines for an
4214 entire SMTP session, which doesn't seem right. The memory is now reset if a
4215 new message is started.
4217 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4218 debugging output unless log_rewrite was also set.
4220 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4221 of (int)(handle) when we know that handle contains (void *)(-1).
4223 61. The Exim daemon panic-logs an error return when it closes the incoming
4224 connection. However "connection reset by peer" seems to be common, and
4225 isn't really an error worthy of noting specially, so that particular error
4228 62. When Exim is trying to find all the local interfaces, it used to panic and
4229 die if the ioctl to get the interface flags failed. However, it seems that
4230 on at least one OS (Solaris 9) it is possible to have an interface that is
4231 included in the list of interfaces, but for which you get a failure error
4232 for this call. This happens when the interface is not "plumbed" into a
4233 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4234 failure of the "get flags" call assumes that the interface is down.
4236 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4237 makes life easier for people who are doing arithmetic on fields extracted
4238 from dates, where you often get leading zeros that should not be
4239 interpreted as octal.
4241 64. Added qualify_domain to the redirect router, to override the global
4244 65. If a pathologically long header line contained very many addresses (the
4245 report of this problem mentioned 10 000) and each of them was rewritten,
4246 Exim could use up a very large amount of memory. (It kept on making new
4247 copies of the header line as it rewrote, and never released the old ones.)
4248 At the expense of a bit more processing, the header rewriting function has
4249 been changed so that it no longer eats memory in this way.
4251 66. The generation of the Received: header has been moved from the time that a
4252 message starts to be received, to the time that it finishes. The timestamp
4253 in the Received: header should now be very close to that of the <= log
4254 line. There are two side-effects of this change:
4256 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4257 logged header lines no longer include the local Received: line, because
4258 it has not yet been created. The same applies to a copy of the message
4259 that is returned to a non-SMTP sender when a message is rejected.
4261 (b) When a filter file is tested using -bf, no additional Received: header
4262 is added to the test message. After some thought, I decided that this
4265 This change does not affect the value of $received_for. It is still set
4266 after address rewriting, but before local_scan() is called.
4268 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4270 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4271 gave an unhelpful panic error message, and a defer error. I have managed to
4272 change this behaviour so that it now rejects any supplied certificate,
4273 which seems right, as the list of acceptable certificates is empty.
4275 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4276 gave an unhelpful defer error. I have not managed to make this reject any
4277 supplied certificates, but the error message it gives is "no certificate
4278 supplied", which is not helpful.
4280 70. exigrep's output now also includes lines that are not associated with any
4281 message, but which match the given pattern. Implemented by a patch from
4282 Martin Sluka, which also tidied up the Perl a bit.
4284 71. Recipient callout verification, like sender verification, was using <> in
4285 the MAIL FROM command. This isn't really the right thing, since the actual
4286 sender may affect whether the remote host accepts the recipient or not. I
4287 have changed it to use the actual sender in the callout; this means that
4288 the cache record is now keyed on a recipient/sender pair, not just the
4289 recipient address. There doesn't seem to be a real danger of callout loops,
4290 since a callout by the remote host to check the sender would use <>.
4291 [SEE ABOVE: changed after hitting problems.]
4293 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4294 temporary errors. However, in the case of such a code being given after
4295 the end of a data transmission (i.e. after ".") Exim was failing to write
4296 a retry record for the message. (Yes, there was some broken host that was
4297 actually sending 8xx at this point.)
4299 73. An unknown lookup type in a host list could cause Exim to panic-die when
4300 the list was checked. (An example that provoked this was putting <; in the
4301 middle of a list instead of at the start.) If this happened during a DATA
4302 ACL check, a -D file could be left lying around. This kind of configuration
4303 error no longer causes Exim to die; instead it causes a defer errror. The
4304 incident is still logged to the main and panic logs.
4306 74. Buglet left over from Exim 3 conversion. The message "too many messages
4307 in one connection" was written to the rejectlog but not the mainlog, except
4308 when address rewriting (yes!) was being logged.
4310 75. Added write_rejectlog option.
4312 76. When a system filter was run not as root (that is, when system_filter_user
4313 was set), the values of the $n variables were not being returned to the
4314 main process; thus, they were not subsequently available in the $sn
4317 77. Added +return_path_on_delivery log selector.
4319 78. A connection timeout was being treated differently from recipients deferred
4320 when testing hosts_max_try with a message that was older than the host's
4321 retry timeout. (The host should not be counted, thus allowing all hosts to
4322 be tried at least once before bouncing.) This may have been the cause of an
4323 occasionally reported bug whereby a message would remain on the queue
4324 longer than the retry timeout, but would be bounced if a delivery was
4325 forced. I say "may" because I never totally pinned down the problem;
4326 setting up timeout/retry tests is difficult. See also the next item.
4328 79. The ultimate address timeout was not being applied to errors that involved
4329 a combination of host plus message (for example, a timeout on a MAIL
4330 command). When an address resolved to a number of possible hosts, and they
4331 were not all tried for each delivery (e.g. because of hosts_max_try), a
4332 message could remain on the queue longer than the retry timeout.
4334 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4337 81. Fixed an obscure SMTP outgoing bug which required at least the following
4338 conditions: (a) there was another message waiting for the same server;
4339 (b) the server returned 5xx to all RCPT commands in the first message so
4340 that the message was not completed; (c) the server dropped the connection
4341 or gave a negative response to the RSET that Exim sends to abort the
4342 transaction. The observed case was a dropped connection after DATA that had
4343 been sent in pipelining mode. That is, the server had advertised PIPELINING
4344 but was not implementing it correctly. The effect of the bug was incorrect
4345 behaviour, such as trying another host, and this could lead to a crash.
4351 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4352 and daemon.c were passed as pointers to ints; they should have been
4353 pointers to socklen_t variables (which are typically unsigned ints).
4355 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4358 3. Fixed a really odd bug that affected only the testing scheme; patching a
4359 certain fixed string in the binary changed the value of another string that
4360 happened to be identical to the end of the original first string.
4362 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4363 name", it returns that address as the IP address. On some operating
4364 systems (e.g. Solaris), it also passes back the IP address string as the
4365 "host name". However, on others (e.g. Linux), it passes back an empty
4366 string. Exim wasn't checking for this, and was changing the host name to an
4367 empty string, assuming it had been canonicized.
4369 5. Although rare, it is permitted to have more than one PTR record for a given
4370 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4371 all the names associated with an address, because they do in Solaris.
4372 However, it seems that they do not in Linux for data that comes from the
4373 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4374 I found this out when I moved to a new Linux workstation and tried to run
4375 the Exim test suite.
4377 To get round this problem I have changed the code so that it now does its
4378 own call to the DNS to look up PTR records when searching for a host name.
4379 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4380 addresses that are only in /etc/hosts are still found.
4382 This behaviour is, however, controlled by an option called host_lookup_
4383 order, which defaults to "bydns:byaddr". If people want to use the other
4384 order, or indeed, just use one or the other means of lookup, they can
4385 specify it in this variable.
4387 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4388 some operating systems, this comes back from gethostbyaddr() as an empty
4389 string, and this is what Exim used to test for. However, it seems that in
4390 other systems, "." is yielded. Exim now tests for this case too.
4392 7. The values of check_spool_space and check_log_space are now held internally
4393 as a number of kilobytes instead of an absolute number of bytes. If a
4394 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4395 kilobyte. This means that much larger values can be stored.
4397 8. Exim monitor: an attempt to get the action menu when not actually pointing
4398 at a message produces an empty menu entitled "No message selected". This
4399 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4400 no entries in it ("Shell widget menu has zero width and/or height"). So I
4401 have added a single, blank menu entry in this case.
4403 9. Added ${quote_local_part.
4405 10. MIME decoding is now applied to the contents of Subject: header lines when
4408 11. Now that a reference to $sender_host_address automatically causes a reverse
4409 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4410 host lookup before query-style lookups in lists that might use this
4411 variable. This has therefore been abolished, and the "net-" prefix is no
4412 longer necessary for query-style lookups.
4414 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4415 have been a typo for LFLAGS, so it has been changed.
4417 13. The install script calls Exim with "-C /dev/null" in order to find the
4418 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4419 to be output. Howeve, since Exim outputs its version number before the
4420 error, it didn't break the script. It just looked ugly. I fixed this by
4421 always allowing "-C /dev/null" if the caller is root.
4423 14. Ignore overlarge ACL variable number when reading spool file - insurance
4424 against a later release with more variables having written the file.
4426 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4427 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4428 this, as well as the form without the "IPv6" on the front.
4430 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4431 OS/Makefile-Darwin file.
4433 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4434 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4436 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4437 variable. However, some OS use ints for inodes. I've added cast to long int
4438 to get rid of the compiler warning.
4440 19. I had forgotten to lock out "/../" in configuration file names when
4441 ALT_CONFIG_PREFIX was set.
4443 20. Routers used for verification do not need to specify transports. However,
4444 if such a router generated a host list, and callout was configured, Exim
4445 crashed, because it could not find a port number from the (non-existent)
4446 transport. It now assumes port 25 in this circumstance.
4448 21. Added the -t option to exigrep.
4450 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4451 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4452 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4453 them is set, LOOKUP_LSEARCH is forced.
4455 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4457 24. Added sender and host information to the "rejected by local_scan()" log
4458 line; previously there was no indication of these.
4460 25. Added .include_if_exists.
4462 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4463 Linux. It turns out that not all file systems support this. Apparently some
4464 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4465 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4466 not supported on the file descriptor, is now ignored when Exim is trying to
4467 sync a directory. This applies only to Linux.
4469 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4471 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4472 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4475 29. Added support for maildirsize files from supplied patch (modified a bit).
4477 30. The use of :fail: followed by an empty string could lead Exim to respond to
4478 sender verification failures with (e.g.):
4480 550 Verification failed for <xxx>
4481 550 Sender verify failed
4483 where the first response line was missing the '-' that indicates it is not
4484 the final line of the response.
4486 31. The loop for finding the name of the user that called Exim had a hardwired
4487 limit of 10; it now uses the value of finduser_retries, which is used for
4488 all other user lookups.
4490 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4493 33. Exim was neglecting to zero errno before one call of strtol() when
4494 expanding a string and expecting an integer value. On some systems this
4495 resulted in spurious "integer overflow" errors. Also, it was casting the
4496 result into an int without checking.
4498 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4499 did not work. The code looks as if it has *never* worked, though it appears
4500 to have been documented since at least releast 1.62. I have made it work.
4502 35. The "timeout_DNS" error in retry rules, also documented since at least
4503 1.62, also never worked. As it isn't clear exactly what this means, and
4504 clearly it isn't a major issue, I have abolished the feature by treating it
4505 as "timeout", and writing a warning to the main and panic logs.
4507 36. The display of retry rules for -brt wasn't always showing the error code
4510 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4511 timeout_connect_A, timeout_connect_MX.
4513 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4514 to the empty sender.
4516 39. The daemon was not analysing the content of -oX till after it had closed
4517 stderr and disconnected from the controlling terminal. This meant that any
4518 syntax errors were only noted on the panic log, and the return code from
4519 the command was 0. By re-arranging the code a little, I've made the
4520 decoding happen first, so such errors now appear on stderr, and the return
4521 code is 1. However, the actual setting up of the sockets still happens in
4522 the disconnected process, so errors there are still only recorded on the
4525 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4526 connections (as happens on some IP stacks) was logged at start up time as
4527 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4528 it from "IPv6 and IPv4", which means that two separate sockets are being
4531 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4532 says whether AF_INET or AF_INET6 was passed as an argument.
4534 42. Exiwhat output was messed up when time zones were included in log
4537 43. Exiwhat now gives more information about the daemon's listening ports,
4538 and whether -tls-on-connect was used.
4540 44. The "port" option of the smtp transport is now expanded.
4542 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4543 silently ignored. Now an error message is written to the main and panic
4546 46. There's a new ACL modifier called "logwrite" which writes to a log file
4547 as soon as it is encountered.
4549 47. Added $local_user_uid and $local_user_gid at routing time.
4551 48. Exim crashed when trying to verify a sender address that was being
4554 49. Exim was recognizing only a space character after ".include". It now also
4555 recognizes a tab character.
4557 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4558 extracting the relevant information from the specification. The man page no
4559 longer contains scrambled data for the -d option, and I've added a section
4560 at the front about calling Exim under different names.
4562 51. Added "extra_headers" argument to the "mail" command in filter files.
4564 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4567 53. Installed eximstats 1.29.
4569 54. Added transport_filter_timeout as a generic transport option.
4571 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4572 Cc: header lines. This was required by RFC 822, but it not required by RFC
4575 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4576 incoming messages that did not have them. Now it does so only if the
4577 message originates locally, that is, if there is no associated remote host
4578 address. When Resent- header lines are present, this applies to the Resent-
4579 lines rather than the non-Resent- lines.
4581 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4582 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4584 58. Messages for configuration errors now include the name of the main
4585 configuration file - useful now that there may be more than one file in a
4586 list (.included file names were always shown).
4588 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4589 for those rare installations that do not start the daemon as root or run it
4590 setuid root. I've cut out the call to initgroups() if the daemon is not
4593 60. The Exim user and group can now be bound into the binary as text strings
4594 that are looked up at the start of Exim's processing.
4596 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4598 62. Added $mailstore_basename variable.
4600 63. Installed patch to sieve.c from Michael Haardt.
4602 64. When Exim failed to open the panic log after failing to open the main log,
4603 the original message it was trying to log was written to stderr and debug
4604 output, but if they were not available (the usual case in production), it
4605 was lost. Now it is written to syslog before the two lines that record the
4606 failures to open the logs.
4608 65. Users' Exim filters run in subprocesses under the user's uid. It is
4609 possible for a "deliver" command or an alias in a "personal" command to
4610 provoke an address rewrite. If logging of address rewriting is configured,
4611 this fails because the process is not running as root or exim. There may be
4612 a better way of dealing with this, but for the moment (because 4.30 needs
4613 to be released), I have disabled address rewrite logging when running a
4614 filter in a non-root, non-exim process.
4620 1. The buildconfig auxiliary program wasn't quoting the value set for
4621 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4622 not defined. This bug was masked in 4.22 by the effect that was fixed in
4625 2. Some messages that were rejected after a message id was allocated were
4626 shown as "incomplete" by exigrep. It no longer does this for messages that
4627 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4629 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4630 have allow_domain_literals set, the ID did not get logged in the <= line.
4631 Domain literals are now always recognized in Message-ID: header lines.
4633 4. The first argument for a ${extract expansion item is the key name or field
4634 number. Leading and trailing spaces in this item were not being ignored,
4635 causing some misleading effects.
4637 5. When deliver_drop_privilege was set, single queue runner processes started
4638 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4639 same command in the process it spins off) were not dropping privilege.
4641 6. When the daemon running as "exim" started a queue runner, it always
4642 re-executed Exim in the spun-off process. This is a waste of effort when
4643 deliver_drop_privilege is set. The new process now just calls the
4644 queue-runner function directly.
4650 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4653 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4654 ignored. Though the use of -oP was forcing the writing of a pid file, it
4655 was always written to the default place.
4657 3. If the message "no IP address found for host xxxx" is generated during
4658 incoming verification, it is now followed by identification of the incoming
4659 connection (so you can more easily find what provoked it).
4661 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4663 5. Added some features to "harden" Exim a bit more against certain attacks:
4665 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4666 be put in Local/Makefile. This is like the never_users runtime option,
4667 but it cannot be overridden. The default setting is "root".
4669 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4670 prefix string with which any file named in a -C command line option
4673 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4674 is retained for -C and -D only if the caller of Exim is root. Without
4675 it, the exim user may also use -C and -D and retain privilege.
4677 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4678 command line option is disabled.
4680 6. Macro names set by the -D option must start with an upper case letter, just
4681 like macro names defined in the configuration file.
4683 7. Added "dereference=" facility to LDAP.
4685 8. Two instances of the typo "uknown" in the source files are fixed.
4687 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4688 the Configure-Makefile script screwed up while processing it.
4690 10. Incorporated PCRE 4.4.
4692 11. The SMTP synchronization check was not operating right at the start of an
4693 SMTP session. For example, it could not catch a HELO sent before the client
4694 waited for the greeting. There is now a check for outstanding input at the
4695 point when the greeting is written. Because of the duplex, asynchronous
4696 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4697 way, but not yet received, when the check is performed.
4699 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4700 on TCP/IP sockets, because this apparently causes some broken clients to
4703 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4704 unchanged) from the Cygwin maintainer.
4706 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4707 was supported instead of testing for mbx. Effectively a typo.
4709 15. The spa authenticator server code was not checking that the input it
4710 received was valid base64.
4712 16. The debug output line for the "set" modifier in ACLs was not showing the
4713 name of the variable that was being set.
4715 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4717 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4718 lookup on its own. Something else was needed to trigger the lookup. For
4719 example, a match in host_lookup or the need for a host name in a host list.
4720 Now, if $sender_host_name is referenced and the host name has not yet been
4721 looked up, a lookup is performed. If the lookup fails, the variable remains
4722 empty, and $host_lookup_failed is set to "1".
4724 19. Added "eqi" as a case-independent comparison operator.
4726 20. The saslauthd authentication condition could segfault if neither service
4727 nor realm was specified.
4729 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4730 error message that was logged was misleading, and incoming SMTP
4731 connections were dropped. The message is now more accurate, and temporary
4732 errors are given to SMTP connections.
4734 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4735 (except RSET) with a 421 error, until QUIT is received. However, it was
4736 failing to send a response to QUIT.
4738 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4739 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4740 ACL is now run after the helo_try_verify_hosts code.
4742 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4743 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4744 case-independent, so other case variants are also recognized). Apparently
4745 some systems use these upper case variants.
4747 25. If more than two messages were waiting for the same host, and a transport
4748 filter was specified for the transport, Exim sent two messages over the
4749 same TCP/IP connection, and then failed with "socket operation on non-
4750 socket" when it tried to send the third.
4752 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4754 27. The extern definition of crypt16() in expand.c was not being excluded when
4755 the OS had its own crypt16() function.
4757 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4758 as a preferred synonym for return_size_limit, both as an option and as an
4761 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4763 30. Changed the default configuration ACL to relax the local part checking rule
4764 for addresses that are not in any local domains. For these addresses,
4765 slashes and pipe symbols are allowed within local parts, but the sequence
4766 /../ is explicitly forbidden.
4768 31. SPA server authentication was not clearing the challenge buffer before
4771 32. log_message in a "warn" ACL statement was writing to the reject log as
4772 well as to the main log, which contradicts the documentation and doesn't
4773 seem right (because no rejection is happening). So I have stopped it.
4775 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4776 However, I am unable to do any testing of this.
4778 34. Fixed an infelicity in the appendfile transport. When checking directories
4779 for a mailbox, to see if any needed to be created, it was accidentally
4780 using path names with one or more superfluous leading slashes; tracing
4781 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4783 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4784 contents are added to the log line that is written for every discarded
4785 recipient. (Previously a log_message setting was ignored.)
4787 36. The ${quote: operator now quotes the string if it is empty.
4789 37. The install script runs exim in order to find its version number. If for
4790 some reason other than non-existence or emptiness, which it checks, it
4791 could not run './exim', it was installing it with an empty version number,
4792 i.e. as "exim-". This error state is now caught, and the installation is
4795 38. An argument was missing from the function that creates an error message
4796 when Exim fails to connect to the socket for saslauthd authentication.
4797 This could cause Exim to crash, or give a corrupted message.
4799 39. Added isip, isip4, and isip6 to ${if conditions.
4801 40. The ACL variables $acl_xx are now saved with the message, and can be
4802 accessed later in routers, transports, and filters.
4804 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4805 strings in the file are not string-expanded.
4807 42. If a MAIL command specified a SIZE value that was too large to fit into an
4808 int variable, the check against message_size_limit failed. Such values are
4809 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4810 day this will have to be increased, but I don't think I want to be around
4811 when emails are that large.
4818 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4819 iconv() is not standard in FreeBSD.
4821 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4822 IPv6 enabled. The observed symptom was a segmentation fault on return from
4823 the function os_common_find_running_interfaces() in src/os.c.
4825 3. In the check_special_case() function in daemon.c I had used "errno" as an
4826 argument name, which causes warnings on some systems. This was basically a
4827 typo, since it was named "eno" in the comments!
4829 4. The code that waits for the clock to tick (at a resolution of some fraction
4830 of a second) so as to ensure message-id uniqueness was always waiting for
4831 at least one whole tick, when it could have waited for less. [This is
4832 almost certainly not relevant at current processor speeds, where it is
4833 unlikely to ever wait at all. But we try to future-proof.]
4835 5. The function that sleeps for a time interval that includes fractions of a
4836 second contained a race. It did not block SIGALRM between setting the
4837 timer, and suspending (a couple of lines later). If the interval was short
4838 and the sigsuspend() was delayed until after it had expired, the suspension
4839 never ended. On busy systems this could lead to processes getting stuck for
4842 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4843 process, before it forks any delivery processes. The open lookup caching
4844 mechanism meant that the open file or database connection was passed into
4845 the delivery process. The problem was that delivery processes always tidy
4846 up cached lookup data. This could cause a problem for the next delivery
4847 process started by the queue runner, because the external queue runner
4848 process does not know about the closure. So the next delivery process
4849 still has data in the lookup cache. In the case of a file lookup, there was
4850 no problem because closing a file descriptor in a subprocess doesn't affect
4851 the parent. However, if the lookup was caching a connection to a database,
4852 the connection was closed, and the second delivery process was likely to
4853 see errors such as "PGSQL: query failed: server closed the connection
4854 unexpectedly". The problem has been fixed by closing all cached lookups
4855 in a queue runner before running a delivery process.
4857 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4858 seem to have the "const" qualifier which it has on other OS. I've
4861 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4862 *of the same type* (client or server) with the same public name that an
4863 error should be diagnosed.
4865 9. When Exim looked up a host name for an IP address, but failed to find the
4866 original IP address when looking up the host name (a safety check), it
4867 output the message "<ip address> does not match any IP for NULL", which was
4868 confusing, to say the least. The bug was that the host name should have
4869 appeared instead of "NULL".
4871 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4872 user that is built into the binary, and the -C or -D options is used, root
4873 privilege is dropped before the configuration file is read. In addition,
4874 logging is switched to stderr instead of the normal log files. If the
4875 configuration then re-defines the Exim user, the unprivileged environment
4876 is probably not what is expected, so Exim logs a panic warning message (but
4879 However, if deliver_drop_privilege is set, the unprivileged state may well
4880 be exactly what is intended, so the warning has been cut out in that case,
4881 and Exim is allowed to try to write to its normal log files.
4887 1. smtp_return_error_details was not giving details for temporary sender
4888 or receiver verification errors.
4890 2. Diagnose a configuration error if two authenticators have the same public
4893 3. Exim used not to create the message log file for a message until the first
4894 delivery attempt. This could be confusing when incoming messages were held
4895 for policy or load reasons. The message log file is now created at the time
4896 the message is received, and an initial "Received" line is written to it.
4898 4. The automatically generated man page for command line options had a minor
4899 bug that caused no ill effects; however, a more serious problem was that
4900 the procedure for building the man page automatically didn't always
4901 operate. Consequently, release 4.20 contains an out-of-date version. This
4902 shouldn't happen again.
4904 5. When building Exim with embedded Perl support, the script that builds the
4905 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4906 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4908 6. The freeze_tell option was not being used for messages that were frozen on
4909 arrival, either by an ACL or by local_scan().
4911 7. Added the smtp_incomplete_transaction log selector.
4913 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4914 was accepting AUTH without a new EHLO.
4916 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4917 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4920 10. Logging of TCP/IP connections (when configured) now happens in the main
4921 daemon process instead of the child process, so that the TCP/IP connection
4922 count is more accurate (but it can never be perfect).
4924 11. The use of "drop" in a nested ACL was not being handled correctly in the
4925 outer ACL. Now, if condition failure induced by the nested "drop" causes
4926 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4927 or "require"), the connection is dropped.
4929 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4930 that yield "discard" can now be used with an "accept" or a "discard" verb,
4931 but an error is generated for any others (because I can't see a useful way
4932 to define what should happen).
4934 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4935 are now processed in the same way as lines in the Exim configuration file.
4936 In particular, continuation lines are supported.
4938 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4940 15. Added -ti meaning -t -i.
4942 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4943 domains, and warn by logging if others are found.
4945 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4946 returned from GIFCONF when Exim is trying to find the list of interfaces on
4947 a host. The code in os.c has been modified to copy each ifreq to an aligned
4948 structure in all cases.
4950 Also, in some cases, the returned ifreq's were being copied to a 'struct
4951 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4952 means the last couple of bytes of an IPv6 address could be chopped if the
4953 ifreq contained only a normal sockaddr (14 bytes storage).
4955 18. Named domain lists were not supported in the hosts_treat_as_local option.
4956 An entry such as +xxxx was not recognized, and was treated as a literal
4959 19. Ensure that header lines added by a DATA ACL are included in the reject log
4960 if the ACL subsequently rejects the message.
4962 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4963 MD5 (which is deprecated).
4965 21. When testing a filter file using -bf, Exim was writing a message when it
4966 took the sender from a "From " line in the message, but it was not doing so
4967 when it took $return_path from a Return-Path: header line. It now does.
4969 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4970 with a valid header line field name (a series of printing characters
4971 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4973 23. Changed "disc" in the source to "disk" to conform to the documentation and
4974 the book and for uniformity.
4976 24. Ignore Sendmail's -Ooption=value command line item.
4978 25. When execve() failed while trying to run a command in a pipe transport,
4979 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4980 could be confused with a return value of 69 from the command itself. This
4981 has been changed to 127, the value the shell returns if it is asked to run
4982 a non-existent command. The wording for the related log line suggests a
4983 non-existent command as the problem.
4985 26. If received_header_text expands to an empty string, do not add a Received:
4986 header line to the message. (Well, it adds a token one on the spool, but
4987 marks it "old" so that it doesn't get used or transmitted.)
4989 27. Installed eximstats 1.28 (addition of -nt option).
4991 28. There was no check for failure on the call to getsockname() in the daemon
4992 code. This can fail if there is a shortage of resources on the system, with
4993 ENOMEM, for example. A temporary error is now given on failure.
4995 29. Contrary to the C standard, it seems that in some environments, the
4996 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4997 program. Exim now does this explicitly; it affects the formatting of
4998 timestamps using strftime().
5000 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5001 complaints. I've now initialized all the variables, to avoid this.
5003 32. Header lines added by a system filter were not being "seen" during
5004 transport-time rewrites.
5006 33. The info_callback() function passed to OpenSSL is set up with type void
5007 (*)(SSL *, int, int), as described somewhere. However, when calling the
5008 function (actually a macro) that sets it up, the type void(*)() is
5009 expected. I've put in a cast to prevent warnings from picky compilers.
5011 34. If a DNS black list lookup found a CNAME record, but there were no A
5012 records associated with the domain it pointed at, Exim crashed.
5014 35. If a DNS black list lookup returned more than one A record, Exim ignored
5015 all but the first. It now scans all returned addresses if a particular IP
5016 value is being sought. In this situation, the contents of the
5017 $dnslist_value variable are a list of all the addresses, separated by a
5020 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5021 to accept a host name and all its aliases if the forward lookup for any of
5022 them yielded the IP address of the incoming connection. Now it accepts only
5023 those names whose forward lookup yields the correct IP address. Any other
5024 names are discarded. This closes a loophole whereby a rogue DNS
5025 administrator could create reverse DNS records to break through a
5026 wildcarded host restriction in an ACL.
5028 37. If a user filter or a system filter that ran in a subprocess used any of
5029 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5030 the wrong values were passed to the pipe command ($thisaddress had the
5031 value of $0, $0 had the value of $1, etc). This bug was introduced by
5032 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5034 38. Improved the line breaking for long SMTP error messages from ACLs.
5035 Previously, if there was no break point between 40 and 75 characters, Exim
5036 left the rest of the message alone. Two changes have been made: (a) I've
5037 reduced the minimum length to 35 characters; (b) if it can't find a break
5038 point between 35 and 75 characters, it looks ahead and uses the first one
5039 that it finds. This may give the occasional overlong line, but at least the
5040 remaining text gets split now.
5042 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5043 file descriptors might be low, and that setting 1000 would always raise it.
5044 It turns out that in some environments, the limit is already over 1000 and
5045 that lowering it causes trouble. So now Exim takes care not to decrease it.
5047 40. When delivering a message, the value of $return_path is set to $sender_
5048 address at the start of routing (routers may change the value). By an
5049 oversight, this default was not being set up when an address was tested by
5050 -bt or -bv, which affected the outcome if any router or filter referred to
5053 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5054 log only when "message" or "log_message" are set. However, if one of the
5055 conditions was an address verification, or a call to a nested ACL, the
5056 messages generated by the underlying test were being passed through. This
5057 no longer happens. The underlying message is available in $acl_verify_
5058 message for both "message" and "log_message" expansions, so it can be
5059 passed through if needed.
5061 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5062 new expansion $bh_ to give the encoded byte string without charset
5063 translation. Translation happens only if iconv() is available; HAVE_ICONV
5064 indicates this at build time. HEADERS_CHARSET gives the charset to
5065 translate to; headers_charset can change it in the configuration, and
5066 "headers charset" can change it in an individual filter file.
5068 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5069 that creates RFC 2047 encoded "words" labels them as that charset instead
5070 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5071 expansion operator; (ii) when Exim creates a From: line for a local
5072 message; (iii) when a header line is rewritten to include a "phrase" part.
5074 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5075 buggy, causing it to skip the first lines of messages whose message ID
5076 ended in 'D'. This would not have bitten before Exim release 4.14, because
5077 message IDs were unlikely to end in 'D' before then. The effect was to have
5078 incorrect size information for certain domains.
5080 45. #include "config.h" was missing at the start of the crypt16.c module. This
5081 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5084 46. If there was a timeout during a "random" callout check, Exim treated it as
5085 a failure of the random address, and carried on sending RSET and the real
5086 address. If the delay was just some slowness somewhere, the response to the
5087 original RCPT would be taken as a response to RSET and so on, causing
5088 mayhem of various kinds.
5090 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5091 when I implemented it. It didn't allow for the fact that some option values
5092 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5093 the right test for positive values.
5095 48. Domain names in DNS records are case-independent. Exim always looks them up
5096 in lower case. Some resolvers return domain names in exactly the case they
5097 appear in the zone file, that is, they may contain uppercase letters. Not
5098 all resolvers do this - some return always lower case. Exim was treating a
5099 change of case by a resolver as a change of domain, similar to a widening
5100 of a domain abbreviation. This triggered its re-routing code and so it was
5101 trying to route what was effectively the same domain again. This normally
5102 caused routing to fail (because the router wouldn't handle the domain
5103 twice). Now Exim checks for this case specially, and just changes the
5104 casing of the domain that it ultimately uses when it transmits the message
5107 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5110 50. If a filter generated a file delivery with a non-absolute name (possible if
5111 no home directory exists for the router), the forbid_file option was not
5114 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5115 the existing equality matching.
5117 52. Exim was using ints instead of ino_t variables in some places where it was
5118 dealing with inode numbers.
5120 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5121 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5122 called TMPDIR, and if it finds it is different, it changes its value.
5124 54. The smtp_printf() function is now made available to local_scan() so
5125 additional output lines can be written before returning. There is also an
5126 smtp_fflush() function to enable the detection of a dropped connection.
5127 The variables smtp_input and smtp_batched_input are exported to
5130 55. Changed the default runtime configuration: the message "Unknown user"
5131 has been removed from the ACL, and instead placed on the localuser router,
5132 using the cannot_route_message feature. This means that any verification
5133 failures that generate their own messages won't get overridden. Similarly,
5134 the "Unrouteable address" message that was in the ACL for unverifiable
5135 relay addresses has also been removed.
5137 56. Added hosts_avoid_esmtp to the smtp transport.
5139 57. The exicyclog script was not checking for the esoteric option
5140 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5141 will work only if exicyclog is run under the appropriate euid.
5143 58. Following a discussion on the list, the rules by which Exim recognises line
5144 endings on incoming messages have been changed. The -dropcr and drop_cr
5145 options are now no-ops, retained only for backwards compatibility. The
5146 following line terminators are recognized: LF CRLF CR. However, special
5147 processing applies to CR:
5149 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5150 nor a local message in the state where . is a terminator.
5152 (ii) If a bare CR is encountered in a header line, an extra space is added
5153 after the line terminator so as not to end the header. The reasoning
5154 behind this is that bare CRs in header lines are most likely either
5155 to be mistakes, or people trying to play silly games.
5157 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5158 was being incorrectly given as 18 bytes larger than it should have been.
5159 This is a VOB (very old bug).
5161 60. This may never have affected anything current, but just in case it has:
5162 When the local host is found other than at the start of a list of hosts,
5163 the local host, those with the same MX, and any that follow, are discarded.
5164 When the list in question was part of a longer list of hosts, the following
5165 hosts (not currently being processed) were also being discarded. This no
5166 longer happens. I'm not sure if this situation could ever has previously
5169 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5172 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5173 header. This is something that is recommended in a new Internet Draft, and
5174 is something that is documented as being done by Sendmail. There are two
5175 possible values. For messages generated by the autoreply transport, Exim
5178 Auto-Submitted: auto-replied
5180 whereas for all other generated messages (e.g. bounces) it adds
5182 Auto-Submitted: auto-generated
5184 63. The "personal" condition in filters now includes a test for the
5185 Auto-Submitted: header. If it contains the string "auto-" the message it
5186 not considered personal.
5188 64. Added rcpt_include_affixes as a generic transport option.
5190 65. Added queue_only_override (default true).
5192 66. Added the syslog_duplication option.
5194 67. If what should have been the first header line of a message consisted of
5195 a space followed by a colon, Exim was mis-interpreting it as a header line.
5196 It isn't of course - it is syntactically invalid and should therefore be
5197 treated as the start of the message body. The misbehaviour could have
5198 caused a number of strange effects, including loss of data in subsequent
5199 header lines, and spool format errors.
5201 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5202 client host had authenticated. This control can now be exercised by an ACL
5203 for more flexibility.
5205 69. By default, callouts do not happen when testing with -bh. There is now a
5206 variant, -bhc, which does actually run the callout code, including
5207 consulting and updating the callout cache.
5209 70. Added support for saslauthd authentication, courtesy of Alexander
5212 71. If statvfs() failed on the spool or log directories while checking their
5213 size for availability, Exim confusingly gave the error "space shortage".
5214 Furthermore, in debugging mode it crashed with a floating point exception.
5215 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5216 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5217 serious problem, Exim now writes to the main and panic logs when this
5218 happens, with details of the failure. It then refuses to accept the
5219 incoming message, giving the message "spool directory problem" or "log
5220 directory problem" with a 421 code for SMTP messages.
5222 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5223 0, 1, and 2 exist, because some OS complain for execs without them (see
5224 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5225 descriptors. However, the code omitted to check that the open succeeded,
5226 causing mysterious errors if for some reason the permissions on /dev/null
5227 got screwed. Now Exim writes a message to the main and panic logs, and
5228 bombs out if it can't open /dev/null.
5230 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5231 interact so that it is all more flexible. It is supposed to remain
5232 backwards compatible. Also added extra_local_interfaces.
5234 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5235 to bomb out with an assertion failure - to the client this appears as a
5236 connection drop. This problem occurs in the part of the code that was taken
5237 from the Samba project. Fortunately, the assertion is in a very simple
5238 function, so I have fixed this by reproducing the function inline in the
5239 one place where it is called, and arranging for authentication to fail
5240 instead of killing the process with assert().
5242 75. The SPA client code was not working when the server requested OEM rather
5243 than Unicode encoding.
5245 76. Added code to make require_files with a specific uid setting more usable in
5246 the case where statting the file as root fails - usually a non-root-mounted
5247 NFS file system. When this happens and the failure is EACCES, Exim now
5248 forks a subprocess and does the per-uid checking as the relevant uid.
5250 77. Added process_log_path.
5252 78. If log_file_path was not explicitly set, a setting of check_log_space or
5253 check_log_inodes was ignored.
5255 79. If a space check for the spool or log partitions fails, the incident is now
5256 logged. Of course, in the latter case the data may get lost...
5258 80. Added the %p formatting code to string_format() so that it can be used to
5259 print addresses in debug_print(). Adjusted all the address printing in the
5260 debugging in store.c to use %p rather than %d.
5262 81. There was a concern that a line of code in smtp_in.c could overflow a
5263 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5264 initially expressed, the concern was not well-founded, because trailing
5265 spaces are removed early. However, if the trailing spaces were followed by
5266 a NULL, they did not get removed, so the overflow was possible. Two fixes
5269 (a) I re-wrote the offending code in a cleaner fashion.
5270 (b) If an incoming SMTP command contains a NULL character, it is rejected
5273 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5274 runs initgroups(), so that if the Exim user is in any additional groups,
5275 they will be used during message reception.
5281 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob