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.
42 TF/01 Failure to get a lock on a hints database can have serious
43 consequences so log it to the panic log.
45 TF/02 Log LMTP confirmation messages in the same way as SMTP,
46 controlled using the smtp_confirmation log selector.
48 TF/03 Include the error message when we fail to unlink a spool file.
50 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
51 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
52 for maintaining out-of-tree patches for some time.
54 PP/01 Bugzilla 139: Documentation and portability issues.
55 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
56 Handle per-OS dynamic-module compilation flags.
58 PP/02 Let /dev/null have normal permissions.
59 The 4.73 fixes were a little too stringent and complained about the
60 permissions on /dev/null. Exempt it from some checks.
61 Reported by Andreas M. Kirchwitz.
63 PP/03 Report version information for many libraries, including
64 Exim version information for dynamically loaded libraries. Created
65 version.h, now support a version extension string for distributors
66 who patch heavily. Dynamic module ABI change.
68 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
69 privilege escalation vulnerability whereby the Exim run-time user
70 can cause root to append content of the attacker's choosing to
73 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
76 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
77 If dropping privileges for untrusted macros, we disabled normal logging
78 on the basis that it would fail; for the Exim run-time user, this is not
79 the case, and it resulted in successful deliveries going unlogged.
80 Fixed. Reported by Andreas Metzler.
86 PP/01 Date: & Message-Id: revert to normally being appended to a message,
87 only prepend for the Resent-* case. Fixes regression introduced in
88 Exim 4.70 by NM/22 for Bugzilla 607.
90 PP/02 Include check_rfc2047_length in configure.default because we're seeing
91 increasing numbers of administrators be bitten by this.
93 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
95 PP/03 Bugzilla 994: added openssl_options main configuration option.
97 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
99 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
101 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
103 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
104 without a peer certificate, leading to a segfault because of an
105 assumption that peers always have certificates. Be a little more
106 paranoid. Problem reported by Martin Tscholak.
108 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
109 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
110 NB: ClamAV planning to remove STREAM in "middle of 2010".
111 CL also introduces -bmalware, various -d+acl logging additions and
112 more caution in buffer sizes.
114 PP/09 Implemented reverse_ip expansion operator.
116 PP/10 Bugzilla 937: provide a "debug" ACL control.
118 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
120 PP/12 Bugzilla 973: Implement --version.
122 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
124 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
126 PP/15 Bugzilla 816: support multiple condition rules on Routers.
128 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
129 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
130 ignore trailing whitespace.
132 JJ/02 prevent non-panic DKIM error from being sent to paniclog
134 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
137 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
138 Notification from Dr Andrew Aitchison.
140 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
141 ExtendedDetectionInfo response format.
142 Notification from John Horne.
144 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
147 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
148 XSL and documented dependency on system catalogs, with examples of how
151 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
154 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
155 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
156 configuration file which is writeable by the Exim user or group.
158 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
159 of configuration files to cover files specified with the -C option if
160 they are going to be used with root privileges, not just the default
163 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
164 option (effectively making it always true).
166 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
167 files to be used while preserving root privileges.
169 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
170 that rogue child processes cannot use them.
172 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
173 run-time user, instead of root.
175 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
176 Exim run-time user without dropping privileges.
178 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
179 result string, instead of calling string_vformat() twice with the same
182 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
183 for other users. Others should always drop root privileges if they use
184 -C on the command line, even for a whitelisted configure file.
186 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
188 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
194 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
195 $data_path, and $header_path variables; fixed documentation bugs and
198 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
199 exipick to access non-standard spools, including the "frozen" queue
202 NM/01 Bugzilla 965: Support mysql stored procedures.
203 Patch from Alain Williams
205 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
207 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
208 Patch from Andreas Metzler
210 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
211 Patch from Kirill Miazine
213 NM/05 Bugzilla 671: Added umask to procmail example.
215 JJ/03 installed exipick 20100323.0, fixing doc bug
217 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
218 directory. Notification and patch from Dan Rosenberg.
220 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
222 TK/02 Improve log output when DKIM signing operation fails.
224 MH/01 Treat the transport option dkim_domain as a colon separated
225 list, not as a single string, and sign the message with each element,
226 omitting multiple occurences of the same signer.
228 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
229 Bugzilla 985, 986. Patch by Simon Arlott
231 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
232 Patch by Simon Arlott
234 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
235 MBX locking. Notification from Dan Rosenberg.
241 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
243 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
245 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
247 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
249 NM/04 Bugzilla 915: Flag broken perl installation during build.
255 TK/01 Added patch by Johannes Berg that expands the main option
256 "spamd_address" if it starts with a dollar sign.
258 TK/02 Write list of recipients to X-Envelope-Sender header when building
259 the mbox-format spool file for content scanning (suggested by Jakob
262 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
263 (http://www.dcc-servers.net/) support via dccifd. Activated by
264 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
266 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
267 by Mark Daniel Reidel <mr@df.eu>.
269 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
270 When building exim an external PCRE library is now needed -
271 PCRE is a system library on the majority of modern systems.
272 See entry on PCRE_LIBS in EDITME file.
274 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
275 conversation. Added nologin parameter to request.
276 Patch contributed by Kirill Miazine.
278 TF/01 Do not log submission mode rewrites if they do not change the address.
280 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
282 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
283 log files in place. Contributed by Roberto Lima.
285 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
287 TF/03 Bugzilla 615: When checking the local_parts router precondition
288 after a local_part_suffix or local_part_prefix option, Exim now
289 does not use the address's named list lookup cache, since this
290 contains cached lookups for the whole local part.
292 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
293 Robert Millan. Documentation is in experimental-spec.txt.
295 TF/04 Bugzilla 668: Fix parallel build (make -j).
297 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
299 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
300 Patch provided by Jan Srzednicki.
302 TF/05 Leading white space used to be stripped from $spam_report which
303 wrecked the formatting. Now it is preserved.
305 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
306 that they are available at delivery time.
308 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
310 TF/08 TLS error reporting now respects the incoming_interface and
311 incoming_port log selectors.
313 TF/09 Produce a more useful error message if an SMTP transport's hosts
314 setting expands to an empty string.
316 NM/06 Bugzilla 744: EXPN did not work under TLS.
317 Patch provided by Phil Pennock.
319 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
320 Patch provided by Richard Godbee.
322 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
323 acl_smtp_notquit, added index entry.
325 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
326 Patch provided by Eugene Bujak.
328 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
329 accept(). Patch provided by Maxim Dounin.
331 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
332 Patch provided by Phil Pennock.
334 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
336 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
337 Patch provided by Brad "anomie" Jorsch.
339 NM/14 Bugzilla 622: Added timeout setting to transport filter.
340 Patch provided by Dean Brooks.
342 TK/05 Add native DKIM support (does not depend on external libraries).
344 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
345 Patch provided by Graeme Fowler.
347 NM/16 Bugzilla 851: Documentation example syntax fix.
349 NM/17 Changed NOTICE file to remove references to embedded PCRE.
351 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
354 NM/19 Bugzilla 745: TLS version reporting.
355 Patch provided by Phil Pennock.
357 NM/20 Bugzilla 167: bool: condition support.
358 Patch provided by Phil Pennock.
360 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
361 clients. Patch provided by Phil Pennock.
363 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
364 Patch provided by Brad "anomie" Jorsch.
366 NM/23 Bugzilla 687: Fix misparses in eximstats.
367 Patch provided by Heiko Schlittermann.
369 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
370 Patch provided by Heiko Schlittermann.
372 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
373 plus update to original patch.
375 NM/26 Bugzilla 799: Documentation correction for ratelimit.
377 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
378 Patch provided by David Brownlee.
380 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
382 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
384 NM/30 Bugzilla 888: TLS documentation bugfixes.
386 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
388 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
389 Unlike the original bugzilla I have changed all shell scripts in src tree.
391 NM/33 Bugzilla 898: Transport filter timeout fix.
392 Patch by Todd Rinaldo.
394 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
395 Patch by Serge Demonchaux.
397 NM/35 Bugzilla 39: Base64 decode bug fixes.
398 Patch by Jakob Hirsch.
400 NM/36 Bugzilla 909: Correct connect() call in dcc code.
402 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
404 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
406 NM/39 Bugzilla 911: Fixed MakeLinks build script.
412 TK/01 Add preliminary DKIM support. Currently requires a forked version of
413 ALT-N's libdkim that I have put here:
414 http://duncanthrax.net/exim-experimental/
416 Note to Michael Haardt: I had to rename some vars in sieve.c. They
417 were called 'true' and it seems that C99 defines that as a reserved
418 keyword to be used with 'bool' variable types. That means you could
419 not include C99-style headers which use bools without triggering
420 build errors in sieve.c.
422 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
423 as mailq or other aliases. Changed the --help handling significantly
424 to do whats expected. exim_usage() emits usage/help information.
426 SC/01 Added the -bylocaldomain option to eximstats.
428 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
430 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
432 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
438 PH/01 Another patch from the Sieve maintainer.
440 PH/02 When an IPv6 address is converted to a string for single-key lookup
441 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
442 dots are used instead of colons so that keys in lsearch files need not
443 contain colons. This was done some time before quoting was made available
444 in lsearch files. However, iplsearch files do require colons in IPv6 keys
445 (notated using the quote facility) so as to distinguish them from IPv4
446 keys. This meant that lookups for IP addresses in host lists did not work
447 for iplsearch lookups.
449 This has been fixed by arranging for IPv6 addresses to be expressed with
450 colons if the lookup type is iplsearch. This is not incompatible, because
451 previously such lookups could never work.
453 The situation is now rather anomolous, since one *can* have colons in
454 ordinary lsearch keys. However, making the change in all cases is
455 incompatible and would probably break a number of configurations.
457 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
460 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
461 conversion specification without a maximum field width, thereby enabling
462 a rogue spamd server to cause a buffer overflow. While nobody in their
463 right mind would setup Exim to query an untrusted spamd server, an
464 attacker that gains access to a server running spamd could potentially
465 exploit this vulnerability to run arbitrary code as the Exim user.
467 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
468 $primary_hostname instead of what libspf2 thinks the hosts name is.
470 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
471 a directory entry by the name of the lookup key. Previously, if a
472 symlink pointed to a non-existing file or a file in a directory that
473 Exim lacked permissions to read, a lookup for a key matching that
474 symlink would fail. Now it is enough that a matching directory entry
475 exists, symlink or not. (Bugzilla 503.)
477 PH/03 The body_linecount and body_zerocount variables are now exported in the
480 PH/04 Added the $dnslist_matched variable.
482 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
483 This means they are set thereafter only if the connection becomes
486 PH/06 Added the client_condition to authenticators so that some can be skipped
487 by clients under certain conditions.
489 PH/07 The error message for a badly-placed control=no_multiline_responses left
490 "_responses" off the end of the name.
492 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
494 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
495 (without spaces) instead of just copying the configuration text.
497 PH/10 Added the /noupdate option to the ratelimit ACL condition.
499 PH/11 Added $max_received_linelength.
501 PH/12 Added +ignore_defer and +include_defer to host lists.
503 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
504 way in which PCRE > 7.0 is built.
506 PH/14 Implemented queue_only_load_latch.
508 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
509 MAIL command. The effect was to mangle the value on 64-bit systems.
511 PH/16 Another patch from the Sieve maintainer.
513 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
515 PH/18 If a system quota error occurred while trying to create the file for
516 a maildir delivery, the message "Mailbox is full" was not appended to the
517 bounce if the delivery eventually timed out. Change 4.67/27 below applied
518 only to a quota excession during the actual writing of the file.
520 PH/19 It seems that peer DN values may contain newlines (and other non-printing
521 characters?) which causes problems in log lines. The DN values are now
522 passed through string_printing() before being added to log lines.
524 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
525 and InterBase are left for another time.)
527 PH/21 Added message_body_newlines option.
529 PH/22 Guard against possible overflow in moan_check_errorcopy().
531 PH/23 POSIX allows open() to be a macro; guard against that.
533 PH/24 If the recipient of an error message contained an @ in the local part
534 (suitably quoted, of course), incorrect values were put in $domain and
535 $local_part during the evaluation of errors_copy.
541 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
542 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
545 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
546 issue a MAIL command.
548 PH/02 In an ACL statement such as
550 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
552 if a client was not listed at all, or was listed with a value other than
553 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
554 the condition was not true (as it should be), so access was not denied.
555 The bug was that the ! inversion was incorrectly passed on to the second
556 item. This has been fixed.
558 PH/03 Added additional dnslists conditions == and =& which are different from
559 = and & when the dns lookup returns more than one IP address.
561 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
562 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
564 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
565 FSYNC, which compiles an option called disable_fsync that allows for
566 bypassing fsync(). The documentation is heavily laced with warnings.
568 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
570 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
571 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
572 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
573 including adding "make clean"; (3) Added -fPIC when compiling the test
574 dynamically loaded module, to get rid of a warning.
576 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
577 message fails, move_frozen_messages = true and ignore_bounce_errors_after
578 = 0s. The bug is otherwise harmless.
580 PH/07 There was a bug in the dovecot authenticator such that the value of
581 $auth1 could be overwritten, and so not correctly preserved, after a
582 successful authentication. This usually meant that the value preserved by
583 the server_setid option was incorrect.
585 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
587 PH/09 Installed PCRE release 7.0.
589 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
590 run for batched SMTP input. It is now run at the start of every message
591 in the batch. While fixing this I discovered that the process information
592 (output by running exiwhat) was not always getting set for -bs and -bS
593 input. This is fixed, and it now also says "batched" for BSMTP.
595 PH/11 Added control=no_pipelining.
597 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
598 patch, slightly modified), and move the expansion of helo_data till after
599 the connection is made in the smtp transport (so it can use these
602 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
604 PH/14 Added log_selector = +pid.
606 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
608 PH/16 Add ${if forany and ${if forall.
610 PH/17 Added dsn_from option to vary the From: line in DSNs.
612 PH/18 Flush SMTP output before performing a callout, unless control =
613 no_callout_flush is set.
615 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
616 was true (the default) a successful delivery failed to delete the retry
617 item, thus causing premature timeout of the address. The bug is now
620 PH/20 Added hosts_avoid_pipelining to the smtp transport.
622 PH/21 Long custom messages for fakedefer and fakereject are now split up
623 into multiline reponses in the same way that messages for "deny" and
624 other ACL rejections are.
626 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
627 with slight modification.
629 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
630 draft, changing the syntax and factoring some duplicate code".
632 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
633 for deliveries of the second and subsequent messages over the same SMTP
636 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
637 ${reduce, with only minor "tidies".
639 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
641 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
642 expansion side effects.
644 PH/27 When a message times out after an over-quota error from an Exim-imposed
645 quota, the bounce message says "mailbox is full". This message was not
646 being given when it was a system quota that was exceeded. It now should
649 MH/03 Made $recipients available in local_scan(). local_scan() already has
650 better access to the recipient list through recipients_list[], but
651 $recipients can be useful in postmaster-provided expansion strings.
653 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
654 in the case of a MAIL command with additional options following the
655 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
656 were accidentally chopped off.
658 PH/29 SMTP synchronization checks are implemented when a command is read -
659 there is a check that no more input is waiting when there shouldn't be
660 any. However, for some commands, a delay in an ACL can mean that it is
661 some time before the response is written. In this time, more input might
662 arrive, invalidly. So now there are extra checks after an ACL has run for
663 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
664 pipelining has not been advertised.
666 PH/30 MH's patch to allow iscntrl() characters to be list separators.
668 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
669 returned in the SMTP response when smtp_return_error_details was false.
672 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
673 instead of the C I/O that was originally supplied, because problems were
676 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
677 Exim which did not show up earlier: it was assuming that a call to
678 SSL_CTX_set_info_callback() might give an error value. In fact, there is
679 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
680 was a macro that became an assignment, so it seemed to work. This has
681 changed to a proper function call with a void return, hence the compile
682 error. Exim's code has been fixed.
684 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
687 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
689 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
691 PH/37 If a message is not accepted after it has had an id assigned (e.g.
692 because it turns out to be too big or there is a timeout) there is no
693 "Completed" line in the log. When some messages of this type were
694 selected by exigrep, they were listed as "not completed". Others were
695 picked up by some special patterns. I have improved the selection
696 criteria to be more general.
698 PH/38 The host_find_failed option in the manualroute router can now be set
699 to "ignore", to completely ignore a host whose IP address cannot be
700 found. If all hosts are ignored, the behaviour is controlled by the new
701 host_all_ignored option.
703 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
704 homing or because of multiple MX records with /mx) generated more than
705 one IP address, and the following item turned out to be the local host,
706 all the secondary addresses of the first item were incorrectly removed
707 from the list, along with the local host and any following hosts (which
708 is what is supposed to happen).
710 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
711 whoever called Exim into the -H file. In the case of the daemon it was
712 behaving confusingly. When first started, it used values for whoever
713 started the daemon, but after a SIGHUP it used the Exim user (because it
714 calls itself on a restart). I have changed the code so that it now always
717 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
718 message are rejected with the same error (e.g. no authentication or bad
719 sender address), and a DATA command is nevertheless sent (as can happen
720 with PIPELINING or a stupid MUA), the error message that was given to the
721 RCPT commands is included in the rejection of the DATA command. This is
722 intended to be helpful for MUAs that show only the final error to their
725 PH/42 Another patch from the Sieve maintainer.
727 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
728 Eximstats - Fixed some broken HTML links and added missing column headers
730 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
731 columns for Rejects, Temp Rejects, Ham, and Spam rows.
733 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
735 PH/43 Yet another patch from the Sieve maintainer.
737 PH/44 I found a way to check for a TCP/IP connection going away before sending
738 the response to the final '.' that terminates a message, but only in the
739 case where the client has not sent further data following the '.'
740 (unfortunately, this is allowed). However, in many cases there won't be
741 any further data because there won't be any more messages to send. A call
742 to select() can be used: if it shows that the input is "ready", there is
743 either input waiting, or the socket has been closed. An attempt to read
744 the next input character can distinguish the two cases. Previously, Exim
745 would have sent an OK response which the client would never have see.
746 This could lead to message repetition. This fix should cure that, at
747 least in a lot of common cases.
749 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
750 advertised in response to EHLO.
756 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
757 fixed by 4.65/MH/01 (is this a record?) are fixed:
759 (i) An empty string was always treated as zero by the numeric comparison
760 operators. This behaviour has been restored.
762 (ii) It is documented that the numeric comparison operators always treat
763 their arguments as decimal numbers. This was broken in that numbers
764 starting with 0 were being interpreted as octal.
766 While fixing these problems I realized that there was another issue that
767 hadn't been noticed. Values of message_size_limit (both the global option
768 and the transport option) were treated as octal if they started with 0.
769 The documentation was vague. These values are now always treated as
770 decimal, and I will make that clear in the documentation.
776 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
777 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
780 MH/01 Don't check that the operands of numeric comparison operators are
781 integers when their expansion is in "skipping" mode (fixes bug
782 introduced by 4.64-PH/07).
784 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
785 child addresses, Exim now panics and dies. Previously, because the count
786 is held in a short int, deliveries were likely to be lost. As such a
787 large number of recipients for a single message is ridiculous
788 (performance will be very, very poor), I have chosen to impose a limit
789 rather than extend the field.
795 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
796 leftover -K file (the existence of which was triggered by #402).
797 While we were at it, introduced process PID as part of the -K
798 filename. This should rule out race conditions when creating
801 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
802 processing considerably. Previous code took too long for large mails,
803 triggering a timeout which in turn triggers #401.
805 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
806 in the DK code in transports.c. sendfile() is not really portable,
807 hence the _LINUX specificness.
809 TF/01 In the add_headers option to the mail command in an Exim filter,
810 there was a bug that Exim would claim a syntax error in any
811 header after the first one which had an odd number of characters
814 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
815 callout verification, Exim cached a "reject" for the entire domain. This
816 is correct for most verifications, but it is not correct for a recipient
817 verification with use_sender or use_postmaster set, because in that case
818 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
819 case of MAIL FROM:<> rejection from other early rejections (e.g.
820 rejection of HELO). When verifying a recipient using a non-null MAIL
821 address, the cache is ignored if it shows MAIL FROM:<> rejection.
822 Whatever the result of the callout, the value of the domain cache is
823 left unchanged (for any other kind of callout, getting as far as trying
824 RCPT means that the domain itself is ok).
826 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
829 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
830 manifest itself as EPIPE rather than ECONNECT. When tidying away a
831 session, the daemon ignores ECONNECT errors and logs others; it now
832 ignores EPIPE as well.
834 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
835 (quoted-printable decoding).
837 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
838 later the small subsequent patch to fix an introduced bug.
840 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
842 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
844 PH/08 An error is now given if message_size_limit is specified negative.
846 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
847 to be given (somewhat) arbitrary names.
849 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
852 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
853 miscellaneous code fixes
855 PH/10 Added the log_reject_target ACL modifier to specify where to log
858 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
859 hostname. This is wrong, because it relates to the incoming message (and
860 probably the interface on which it is arriving) and not to the outgoing
861 callout (which could be using a different interface). This has been
862 changed to use the value of the helo_data option from the smtp transport
863 instead - this is what is used when a message is actually being sent. If
864 there is no remote transport (possible with a router that sets up host
865 addresses), $smtp_active_hostname is used.
867 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
868 tweaks were necessary in order to get it to work (see also 21 below):
869 (a) The code assumed that strncpy() returns a negative number on buffer
870 overflow, which isn't the case. Replaced with Exim's string_format()
872 (b) There were several signed/unsigned issues. I just did the minimum
873 hacking in of casts. There is scope for a larger refactoring.
874 (c) The code used strcasecmp() which is not a standard C function.
875 Replaced with Exim's strcmpic() function.
876 (d) The code set only $1; it now sets $auth1 as well.
877 (e) A simple test gave the error "authentication client didn't specify
878 service in request". It would seem that Dovecot has changed its
879 interface. Fortunately there's a specification; I followed it and
880 changed what the client sends and it appears to be working now.
882 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
885 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
886 address A is aliased to B and C, where B exists and C does not. Without
887 -v the output is "A verified" because verification stops after a
888 successful redirection if more than one address is generated. However,
889 with -v the child addresses are also verified. Exim was outputting "A
890 failed to verify" and then showing the successful verification for C,
891 with its parentage. It now outputs "B failed to verify", showing B's
892 parentage before showing the successful verification of C.
894 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
895 look up a TXT record in a specific list after matching in a combined
898 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
899 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
900 they consult the DNS. I had assumed they would set it the way they
901 wanted; and indeed my experiments on Linux seem to show that in some
902 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
903 To be on the safe side, however, I have now made the interface to
904 host_find_byname() similar to host_find_bydns(), with an argument
905 containing the DNS resolver options. The host_find_byname() function now
906 sets these options at its start, just as host_find_bydns() does. The smtp
907 transport options dns_qualify_single and dns_search_parents are passed to
908 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
909 of host_find_byname() use the default settings of RES_DEFNAMES
910 (qualify_single) but not RES_DNSRCH (search_parents).
912 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
913 spool_read_header() do less string testing, by means of a preliminary
914 switch on the second character of optional "-foo" lines. (This is
915 overdue, caused by the large number of possibilities that now exist.
916 Originally there were few.) While I was there, I also converted the
917 str(n)cmp tests so they don't re-test the leading "-" and the first
918 character, in the hope this might squeeze out yet more improvement.
920 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
921 flag allowing group syntax was set by the header_syntax check but not
922 turned off, possible causing trouble later; (2) The flag was not being
923 set at all for the header_verify test, causing "group"-style headers to
924 be rejected. I have now set it in this case, and also caused header_
925 verify to ignore an empty address taken from a group. While doing this, I
926 came across some other cases where the code for allowing group syntax
927 while scanning a header line wasn't quite right (mostly, not resetting
928 the flag correctly in the right place). These bugs could have caused
929 trouble for malformed header lines. I hope it is now all correct.
931 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
932 with the "reply" argument non-NULL. The code, however (which originally
933 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
934 but it didn't always do it. This confused somebody who was copying the
935 code for some other use. I have removed all the tests.
937 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
938 feature that was used to support insecure browsers during the U.S. crypto
939 embargo. It requires special client support, and Exim is probably the
940 only MTA that supported it -- and would never use it because real RSA is
941 always available. This code has been removed, because it had the bad
942 effect of slowing Exim down by computing (never used) parameters for the
943 RSA_EXPORT functionality.
945 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
946 authenticator to fail if there's a tab character in the incoming data
947 (there should never be unless someone is messing about, as it's supposed
948 to be base64-encoded). Also added, on Timo's advice, the "secured" option
949 if the connection is using TLS or if the remote IP is the same as the
950 local IP, and the "valid-client-cert option" if a client certificate has
953 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
954 authenticators. This can be used for authorization after authentication
955 succeeds. (In the case of plaintext, it servers for both authentication
958 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
959 if any retry times were supplied.
961 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
962 connection, where there is no client IP address to check. In this
963 situation, the verify now always succeeds.
965 PH/25 Applied John Jetmore's -Mset patch.
967 PH/26 Added -bem to be like -Mset, but loading a message from a file.
969 PH/27 In a string expansion for a processed (not raw) header when multiple
970 headers of the same name were present, leading whitespace was being
971 removed from all of them, but trailing whitespace was being removed only
972 from the last one. Now trailing whitespace is removed from each header
973 before concatenation. Completely empty headers in a concatenation (as
976 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
977 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
979 PH/29 [Removed. This was a change that I later backed out, and forgot to
980 correct the ChangeLog entry (that I had efficiently created) before
981 committing the later change.]
983 PH/30 Exim was sometimes attempting to deliver messages that had suffered
984 address errors (4xx response to RCPT) over the same connection as other
985 messages routed to the same hosts. Such deliveries are always "forced",
986 so retry times are not inspected. This resulted in far too many retries
987 for the affected addresses. The effect occurred only when there were more
988 hosts than the hosts_max_try setting in the smtp transport when it had
989 the 4xx errors. Those hosts that it had tried were not added to the list
990 of hosts for which the message was waiting, so if all were tried, there
991 was no problem. Two fixes have been applied:
993 (i) If there are any address or message errors in an SMTP delivery, none
994 of the hosts (tried or untried) are now added to the list of hosts
995 for which the message is waiting, so the message should not be a
996 candidate for sending over the same connection that was used for a
997 successful delivery of some other message. This seems entirely
998 reasonable: after all the message is NOT "waiting for some host".
999 This is so "obvious" that I'm not sure why it wasn't done
1000 previously. Hope I haven't missed anything, but it can't do any
1001 harm, as the worst effect is to miss an optimization.
1003 (ii) If, despite (i), such a delivery is accidentally attempted, the
1004 routing retry time is respected, so at least it doesn't keep
1005 hammering the server.
1007 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1008 in ${readsocket because some servers need this prod.
1010 PH/32 Added some extra debug output when updating a wait-xxx database.
1012 PH/33 The hint "could be header name not terminated by colon", which has been
1013 given for certain expansion errors for a long time, was not being given
1014 for the ${if def:h_colon_omitted{... case.
1016 PH/34 The spec says: "With one important exception, whenever a domain list is
1017 being scanned, $domain contains the subject domain." There was at least
1018 one case where this was not true.
1020 PH/35 The error "getsockname() failed: connection reset by peer" was being
1021 written to the panic log as well as the main log, but it isn't really
1022 panic-worthy as it just means the connection died rather early on. I have
1023 removed the panic log writing for the ECONNRESET error when getsockname()
1026 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1027 runs only) independently of the message's sender address. This meant
1028 that, if the 4xx error was in fact related to the sender, a different
1029 message to the same recipient with a different sender could confuse
1030 things. In particualar, this can happen when sending to a greylisting
1031 server, but other circumstances could also provoke similar problems.
1032 I have changed the default so that the retry time for these errors is now
1033 based a combination of the sender and recipient addresses. This change
1034 can be overridden by setting address_retry_include_sender=false in the
1037 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1038 remote server are returned as part of bounce messages. This was not
1039 happening for LMTP over a pipe (the lmtp transport), but now it is the
1040 same for both kinds of LMTP.
1042 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1043 in header lines that were in fact CNAMEs. This is no longer the case.
1045 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1046 and queue runs started by the daemon processed all messages. This has
1047 been fixed so that -R and -S can now usefully be given with -q<time>.
1049 PH/40 Import PCRE release 6.7 (fixes some bugs).
1051 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1053 PH/42 Give an error if -q is specified more than once.
1055 PH/43 Renamed the variables $interface_address and $interface_port as
1056 $received_ip_address and $received_port, to make it clear that these
1057 values apply to message reception, and not to the outgoing interface when
1058 a message is delivered. (The old names remain recognized, of course.)
1060 PH/44 There was no timeout on the connect() call when using a Unix domain
1061 socket in the ${readsocket expansion. There now is.
1063 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1064 be meaningful with "accept".
1066 SC/01 Eximstats V1.43
1067 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1069 SC/02 Eximstats V1.44
1070 Use a glob alias rather than an array ref in the generated
1071 parser. This improves both readability and performance.
1073 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1074 Collect SpamAssassin and rejection statistics.
1075 Don't display local sender or destination tables unless
1076 there is data to show.
1077 Added average volumes into the top table text output.
1079 SC/04 Eximstats V1.46
1080 Collect data on the number of addresses (recipients)
1081 as well as the number of messages.
1083 SC/05 Eximstats V1.47
1084 Added 'Message too big' to the list of mail rejection
1085 reasons (thanks to Marco Gaiarin).
1087 SC/06 Eximstats V1.48
1088 Mainlog lines which have GMT offsets and are too short to
1089 have a flag are now skipped.
1091 SC/07 Eximstats V1.49 (Alain Williams)
1092 Added the -emptyok flag.
1094 SC/08 Eximstats V1.50
1095 Fixes for obtaining the IP address from reject messages.
1097 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1098 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1099 whitesspace changes from 4.64-PH/27
1101 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1104 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1105 are found, allow negative numbers in numeric criteria)
1107 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1109 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1110 to match changes made in 4.64-PH/43
1112 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1114 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1115 "domain" is actually an IP address, and force a failure. This locks out
1116 those revolvers/nameservers that support "A-for-A" lookups, in
1117 contravention of the specifications.
1119 PH/48 When a host name was looked up from an IP address, and the subsequent
1120 forward lookup of the name timed out, the host name was left in
1121 $sender_host_name, contrary to the specification.
1123 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1124 restricted to single-key lookups, Exim was not diagnosing an error if
1125 * or *@ was used with a query-style lookup.
1127 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1129 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1130 long ago, but noone interested enough thought of it. Let's just say that
1131 the "1.1" means that there are some new functions that weren't there at
1132 some point in the past.
1134 PH/51 Error processing for expansion failure of helo_data from an smtp
1135 transport during callout processing was broken.
1137 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1138 tested/used via the -bh/-bhc/-bs options.
1140 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1141 bug, fixed in subsequent PCRE releases).
1143 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1144 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1146 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1152 SC/01 Use a glob alias rather than an array ref in eximstats generated
1153 parser. This improves both readability and performance.
1155 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1156 Don't display local sender or destination tables in eximstats unless
1157 there is data to show.
1158 Added average volumes into the eximstats top table text output.
1160 SC/03 Collect data on the number of addresses (recipients) as well
1161 as the number of messages in eximstats.
1163 TF/01 Correct an error in the documentation for the redirect router. Exim
1164 does (usually) call initgroups() when daemonizing.
1166 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1167 with consistent privilege compared to when running as a daemon.
1169 TF/03 Note in the spec that $authenticated_id is not set for local
1170 submissions from trusted users.
1172 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1173 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1175 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1176 by adding some example configuration directives to the default
1177 configuration file. A little bit of work is required to uncomment the
1178 directives and define how usernames and passwords are checked, but
1179 there is now a framework to start from.
1181 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1182 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1183 without this. I don't know how relevant this is to other LDAP libraries.
1185 PH/02 Add the verb name to the "unknown ACL verb" error.
1187 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1189 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1191 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1192 directory not expanded when it should be if an expanded home directory
1193 was set for the address (which is overridden by the transport).
1195 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1198 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1199 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1200 because it is too late at that time, and has no effect.
1202 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1203 security issue with \' (bugzilla #107). I could not use the
1204 PQescapeStringConn() function, because it needs a PGconn value as one of
1207 PH/08 When testing addresses using -bt, indicate those final addresses that
1208 are duplicates that would not cause an additional delivery. At least one
1209 person was confused, thinking that -bt output corresponded to deliveries.
1210 (Suppressing duplicates isn't a good idea as you lose the information
1211 about possibly different redirections that led to the duplicates.)
1213 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1214 systems where poll() doesn't work, in particular OS X.
1216 PH/10 Added more information to debugging output for retry time not reached.
1218 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1219 operations in malware.c.
1221 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1224 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1225 syslog_duplication set false, log lines that would normally be written
1226 both the the main log and to the reject log were not written to syslog at
1229 PH/14 In the default configuration, change the use of "message" in ACL warn
1230 statements to "add_header".
1232 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1233 not followed by a command (e.g. "seen endif").
1235 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1236 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1239 PH/17 Added extra conditions to the default value of delay_warning_condition
1243 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1244 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1245 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1248 The Auto-Submitted: and various List- headers are standardised, whereas I
1249 don't think Precedence: ever was.
1251 PH/18 Refactored debugging code in route_finduser() to show more information,
1252 in particular, the error code if getpwnam() issues one.
1254 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1255 This is apparently needed in addition to the PH/07 change above to avoid
1256 any possible encoding problems.
1258 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1259 but not after initializing Perl.
1261 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1262 output them only if debugging. By default they are written stderr,
1263 apparently, which is not desirable.
1265 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1268 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1271 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1273 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1274 authenticated or an ident call has been made. Suppress the default
1275 values for $authenticated_id and $authenticated_sender (but permit -oMai
1276 and -oMas) when testing with -bh.
1278 PH/24 Re-jigged the order of the tests in the default configuration so that the
1279 tests for valid domains and recipients precede the DNS black list and CSA
1280 tests, on the grounds that those ones are more expensive.
1282 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1283 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1284 This bug exists in every version of Exim that I still have, right back to
1287 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1288 However, an attempt to turn on case-sensitivity in a regex key by
1289 including (?-i) didn't work because the subject string was already
1290 lowercased, and the effects were non-intuitive. It turns out that a
1291 one-line patch can be used to allow (?-i) to work as expected.
1297 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1298 other effects) broke the use of negated acl sub-conditions.
1300 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1303 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1304 "Deny" causes Exim to reject the incoming connection with a 554 error.
1305 Unfortunately, if there is a major crisis, such as a disk failure,
1306 tcp-wrappers gives "deny", whereas what one would like would be some
1307 kind of temporary error. A kludge has been added to help with this.
1308 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1309 554 error is used if errno is still zero or contains ENOENT (which occurs
1310 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1313 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1315 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1316 errors. Otherwise a message that provokes a temporary error (when other
1317 messages do not) can cause a whole host to time out.
1319 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1320 addresses were routed directly to files or pipes from a redirect router.
1321 File deliveries just didn't batch; pipe deliveries might have suffered
1324 PH/06 A failure to get a lock for a hints database would erroneously always say
1325 "Failed to get write lock", even when it was really a read lock.
1327 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1328 of 0600. This has been changed to use the value of the lockfile_mode
1329 option (which defaults to 0600).
1331 PH/08 Applied small patch from the Sieve maintainer.
1333 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1334 folder from quota calculations, a direct delivery into this folder messed
1335 up the contents of the maildirsize file. This was because the regex was
1336 used only to exclude .Trash (or whatever) when the size of the mailbox
1337 was calculated. There was no check that a delivery was happening into an
1338 excluded directory. This bug has been fixed by ignoring all quota
1339 processing for deliveries into excluded directories.
1341 PH/10 Added the maildirfolder_create_regex option to appendfile.
1347 PH/01 The code for finding all the local interface addresses on a FreeBSD
1348 system running IPv6 was broken. This may well have applied to all BSD
1349 systems, as well as to others that have similar system calls. The broken
1350 code found IPv4 interfaces correctly, but gave incorrect values for the
1351 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1352 that it would not match correctly against @[] and not recognize the IPv6
1355 PH/02 The ipliteral router was not recognizing addresses of the form user@
1356 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1358 PH/03 Added disable_ipv6.
1360 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1361 decoded form, because it is most often used to construct To: headers
1362 lines in autoreplies, and the decoded form may well be syntactically
1363 invalid. However, $reply_address has leading white space removed, and all
1364 newlines turned into spaces so that the autoreply transport does not
1367 PH/05 If group was specified without a user on a router, and no group or user
1368 was specified on a transport, the group from the router was ignored.
1370 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1371 for visible compile-time settings that can be used to change these
1372 numbers, for those that want even more. Backwards compatibility with old
1373 spool files has been maintained. However, going back to a previous Exim
1374 release will lost any variables that are in spool files.
1376 PH/07 Two small changes when running in the test harness: increase delay when
1377 passing a TCP/IP connection to a new process, in case the original
1378 process has to generate a bounce, and remove special handling of
1379 127.0.0.2 (sic), which is no longer necessary.
1381 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1382 be the same on different OS.
1384 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1387 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1388 whether --show-vars was specified or not
1390 JJ/02 exipick: Added support for new ACL variable spool format introduced
1393 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1394 syntactically invalid From: or Reply-to: line, and a filter used this to
1395 generate an autoreply, and therefore failed to obtain an address for the
1396 autoreply, Exim could try to deliver to a non-existent relative file
1397 name, causing unrelated and misleading errors. What now happens is that
1398 it logs this as a hard delivery error, but does not attempt to create a
1401 PH/11 The exinext utility has a -C option for testing purposes, but although
1402 the given file was scanned by exinext itself; it wasn't being passed on
1403 when Exim was called.
1405 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1406 an end-of-file indication when reading a command response.
1408 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1409 compiled. In many other places in Exim, IPv6 addresses are always
1410 recognized, so I have changed this. It also means that IPv4 domain
1411 literals of the form [IPV4:n.n.n.n] are now always recognized.
1413 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1414 used if the router is not running as root, for example, when verifying at
1415 ACL time, or when using -bh. The debugging output from this situation was
1416 non-existent - all you got was a failure to exec. I have made two
1419 (a) Failures to set uid/gid, the current directory, or a process leader
1420 in a subprocess such as that created by queryprogram now generate
1421 suitable debugging ouput when -d is set.
1423 (b) The queryprogram router detects when it is not running as root,
1424 outputs suitable debugging information if -d is set, and then runs
1425 the subprocess without attempting to change uid/gid.
1427 PH/15 Minor change to Makefile for building test_host (undocumented testing
1430 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1431 additional section of a DNS packet that returns MX or SRV records.
1432 Instead, it always explicitly searches for A/AAAA records. This avoids
1433 major problems that occur when a DNS server includes only records of one
1434 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1435 fixed another bug: if SRV records were looked up and the corresponding
1436 address records were *not* found in the additional section, the port
1437 values from the SRV records were lost.
1439 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1440 using the correct key (the original address) when searching the retry
1441 rules in order to find which one to use for generating the retry hint.
1443 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1444 adding the default one. Similarly, if it contains a Reply-To: header, the
1445 errors_reply_to option, if set, is not used.
1447 PH/19 When calculating a retry time, Exim used to measure the "time since
1448 failure" by looking at the "first failed" field in the retry record. Now
1449 it does not use this if it is later than than the arrival time of the
1450 message. Instead it uses the arrival time. This makes for better
1451 behaviour in cases where some deliveries succeed, thus re-setting the
1452 "first failed" field. An example is a quota failure for a huge message
1453 when small messages continue to be delivered. Without this change, the
1454 "time since failure" will always be short, possible causing more frequent
1455 delivery attempts for the huge message than are intended.
1456 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1458 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1459 $1, $2, $3) because the numerical variables can be reset during some
1460 expansion items (e.g. "match"), thereby losing the authentication data.
1462 PH/21 Make -bV show the size of off_t variables so that the test suite can
1463 decide whether to run tests for quotas > 2G.
1465 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1466 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1467 filecount value is greater than 2G or if a quota value is greater than 2G
1468 on a system where the size of off_t is not greater than 4, a panic error
1471 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1472 never match. The debug and -bh output now contains an explicit error
1473 message indicating a malformed IPv4 address or mask.
1475 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1476 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1477 PH/23 above applies.
1479 PH/25 Do not write to syslog when running in the test harness. The only
1480 occasion when this arises is a failure to open the main or panic logs
1481 (for which there is an explicit test).
1483 PH/26 Added the /no_tell option to "control=freeze".
1485 PH/27 If a host name lookup failed very early in a connection, for example, if
1486 the IP address matched host_lookup and the reverse lookup yielded a name
1487 that did not have a forward lookup, an error message of the form "no IP
1488 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1489 could be logged. Now it outputs the IP address instead of "NULL".
1491 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1492 allows the sender and the authenticated sender to be set when
1493 submitting a message from within Exim. Since child_open_exim() is
1494 documented for local_scan(), the new function should be too.
1496 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1497 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1498 results in an empty string is now treated as unset.
1500 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1502 PH/31 Added sender_verify_fail logging option.
1504 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1505 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1506 tidied the source and removed it altogether.
1508 PH/33 When a queue run was abandoned because the load average was too high, a
1509 log line was always written; now it is written only if the queue_run log
1510 selector is set. In addition, the log line for abandonment now contains
1511 information about the queue run such as the pid. This is always present
1512 in "start" and "stop" lines but was omitted from the "abandon" line.
1514 PH/34 Omit spaces between a header name and the colon in the error message that
1515 is given when verify = headers_syntax fails (if there are lots of them,
1516 the message gets confusing).
1518 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1519 names, as there are now some PTR records that contain slashes. This check
1520 is only to protect against broken name servers that fall over on strange
1521 characters, so the fact that it applies to all lookups doesn't matter.
1523 PH/36 Now that the new test suite is complete, we can remove some of the
1524 special code in Exim that was needed for the old test suite. For example,
1525 sorting DNS records because real resolvers return them in an arbitrary
1526 order. The new test suite's fake resolver always returns records in the
1529 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1530 bounces) except when queue_only is set, to avoid logging races between
1531 the different processes.
1533 PH/38 Panic-die if .include specifies a non-absolute path.
1535 PH/39 A tweak to the "H" retry rule from its user.
1537 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1538 a label. They prevented compilation on older perls.
1540 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1541 a warning to be raised on newish perls.
1543 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1544 on queue. Changes to match documented behaviour of showing count of
1545 messages matching specified criteria.
1547 PH/40 Changed the default ident timeout from 30s to 5s.
1549 PH/41 Added support for the use of login_cap features, on those BSD systems
1550 that have them, for controlling the resources used by pipe deliveries.
1552 PH/42 The content-scanning code uses fopen() to create files in which to put
1553 message data. Previously it was not paying any attention to the mode of
1554 the files. Exim runs with umask(0) because the rest of the code creates
1555 files with open(), and sets the required mode explicitly. Thus, these
1556 files were ending up world-writeable. This was not a big issue, because,
1557 being within the spool directory, they were not world-accessible. I have
1558 created a function called modefopen, which takes an additional mode
1559 argument. It sets umask(777), creates the file, chmods it to the required
1560 mode, then resets the umask. All the relevant calls to fopen() in the
1561 content scanning code have been changed to use this function.
1563 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1564 to 24 hours. This avoids potential overflow problems when processing G
1565 and H retry rules. I suspect nobody ever tinkers with this value.
1567 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1569 PH/45 When the plaintext authenticator is running as a client, the server's
1570 challenges are checked to ensure they are valid base64 strings. By
1571 default, the authentication attempt is cancelled if an invalid string is
1572 received. Setting client_ignore_invalid_base64 true ignores these errors.
1573 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1574 they are received. Thus, the responses can be made to depend on the
1575 challenges. If an invalid string is ignored, an empty string is placed in
1578 PH/46 Messages that are created by the autoreply transport now contains a
1579 References: header, in accordance with RFCs 2822 and 3834.
1581 PH/47 Added authenticated_sender_force to the smtp transport.
1583 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1585 PH/49 Installed latest patch from the Sieve maintainer.
1587 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1588 was also set, the appendfile transport was unnecessarily scanning a
1589 directory of message files (e.g. for maildir delivery) to find the count
1590 of files (along with the size), even though it did not need this
1591 information. It now does the scan only if it needs to find either the
1592 size of the count of files.
1594 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1596 PH/52 Two bugs concerned with error handling when the smtp transport is
1599 (i) Exim was not creating retry information for temporary errors given
1600 for individual recipients after the DATA command when the smtp transport
1601 was used in LMTP mode. This meant that they could be retried too
1602 frequently, and not timed out correctly.
1604 (ii) Exim was setting the flag that allows error details to be returned
1605 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1606 recipients that were returned after the DATA command.
1608 PH/53 This is related to PH/52, but is more general: for any failing address,
1609 when detailed error information was permitted to be returned to the
1610 sender, but the error was temporary, then after the final timeout, only
1611 "retry timeout exceeded" was returned. Now it returns the full error as
1612 well as "retry timeout exceeded".
1614 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1615 do this, and (what is worse) MTAs that accept it.
1617 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1618 will now be deprecated.
1620 PH/56 New os.c-cygwin from the Cygwin maintainer.
1622 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1623 formats (previously only available in exim formats via -bpr, -bpru,
1624 and -bpra. Now also available in native and exiqgrep formats)
1626 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1627 with very large, slow to parse queues
1629 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1631 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1633 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1634 responses to authentication challenges, though it was showing the
1635 challenges; (ii) I've removed the CR characters from the debug output for
1638 PH/58 Allow for the insertion of a newline as well as a space when a string
1639 is turned into more than one encoded-word during RFC 2047 encoding. The
1640 Sieve code now uses this.
1642 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1643 data_4xx, lost_connection, tls_required.
1645 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1646 message was being sent as an SMTP response.
1648 PH/61 Add -l and -k options to exicyclog.
1650 PH/62 When verifying, if an address was redirected to one new address, so that
1651 verification continued, and the new address failed or deferred after
1652 having set something in $address_data, the value of $address_data was not
1653 passed back to the ACL. This was different to the case when no
1654 redirection occurred. The value is now passed back in both cases.
1656 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1657 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1658 use login_cap.h, so on its own it isn't the distinguishing feature. The
1659 new name refers directly to the setclassresources() function.
1661 PH/65 Added configuration files for NetBSD3.
1663 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1665 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1666 is preferred over IPv4.
1668 PH/68 The bounce_return_message and bounce_return_body options were not being
1669 honoured for bounces generated during the reception of non-SMTP messages.
1670 In particular, this applied to messages rejected by the ACL. This bug has
1671 been fixed. However, if bounce_return_message is true and bounce_return_
1672 body is false, the headers that are returned for a non-SMTP message
1673 include only those that have been read before the error was detected.
1674 (In the case of an ACL rejection, they have all been read.)
1676 PH/69 The HTML version of the specification is now built in a directory called
1677 spec_html instead of spec.html, because the latter looks like a path with
1678 a MIME-type, and this confuses some software.
1680 PH/70 Catch two compiler warnings in sieve.c.
1682 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1683 function verify_get_ident() calls ip_connect() to connect a socket, but
1684 if the "connect()" function timed out, ip_connect() used to close the
1685 socket. However, verify_get_ident() also closes the socket later, and in
1686 between Exim writes to the log, which may get opened at this point. When
1687 the socket was closed in ip_connect(), the log could get the same file
1688 descriptor number as the socket. This naturally causes chaos. The fix is
1689 not to close the socket in ip_connect(); the socket should be closed by
1690 the function that creates it. There was only one place in the code where
1691 this was missing, in the iplookup router, which I don't think anybody now
1692 uses, but I've fixed it anyway.
1694 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1695 well as to direct DNS lookups. Otherwise the handling of names in host
1696 lists is inconsistent and therefore confusing.
1702 PH/01 Two changes to the default runtime configuration:
1704 (1) Move the checks for relay_from_hosts and authenticated clients from
1705 after to before the (commented out) DNS black list checks.
1707 (2) Add control=submission to the relay_from_hosts and authenticated
1708 clients checks, on the grounds that messages accepted by these
1709 statements are most likely to be submissions.
1711 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1713 (1) Generate an error if the third argument for the ${prvs expansion is
1716 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1719 (3) Reset the variables that are obtained from the first argument of
1720 ${prvscheck and used in the second argument before leaving the code,
1721 because their memory is reclaimed, so using them afterwards may do
1724 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1725 one (it's much easier than Tom thought :-).
1727 (5) Because of (4), we can now allow for the use of $prvscheck_result
1728 inside the third argument.
1730 PH/03 For some reason, the default setting of PATH when running a command from
1731 a pipe transport was just "/usr/bin". I have changed it to
1734 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1735 anything to be listed in the output from -bV.
1737 PH/05 When a filter generated an autoreply, the entire To: header line was
1738 quoted in the delivery log line, like this:
1740 => >A.N.Other <ano@some.domain> <original@ddress> ...
1742 This has been changed so that it extracts the operative address. There
1743 may be more than one such address. If so, they are comma-separated, like
1746 => >ano@some.domain,ona@other.domain <original@ddress> ...
1748 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1749 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1750 not being looked up in the rDNS to get a host name, Exim was showing the
1751 IP address twice in Received: lines, even though the IP addresses were
1752 identical. For example:
1754 Received: from [1.2.3.4] (helo=[1.2.3.4])
1756 However, if the real host name was known, it was omitting the HELO data
1757 if it matched the actual IP address. This has been tidied up so that it
1758 doesn't show the same IP address twice.
1760 PH/07 When both +timestamp and +memory debugging was on, the value given by
1761 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1762 called by the debug printing, thereby overwriting the timestamp buffer.
1763 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1765 PH/08 When the original message was included in an autoreply transport, it
1766 always said "this is a copy of the message, including all the headers",
1767 even if body_only or headers_only was set. It now gives an appropriate
1770 PH/09 Applied a patch from the Sieve maintainer which:
1772 o fixes some comments
1773 o adds the (disabled) notify extension core
1774 o adds some debug output for the result of if/elsif tests
1775 o points to the current vacation draft in the documentation
1776 and documents the missing references header update
1780 o fixes a bug in processing the envelope test (when testing
1781 multiple envelope elements, the last element determinted the
1784 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1785 Electronic Mail") by including:
1787 Auto-submitted: auto-generated
1789 in the messages that it generates (bounce messages and others, such as
1790 warnings). In the case of bounce messages for non-SMTP mesages, there was
1791 also a typo: it was using "Auto_submitted" (underscore instead of
1792 hyphen). Since every message generated by Exim is necessarily in response
1793 to another message, thes have all been changed to:
1795 Auto-Submitted: auto-replied
1797 in accordance with these statements in the RFC:
1799 The auto-replied keyword:
1801 - SHOULD be used on messages sent in direct response to another
1802 message by an automatic process,
1804 - MUST NOT be used on manually-generated messages,
1806 - MAY be used on Delivery Status Notifications (DSNs) and Message
1807 Disposition Notifications (MDNs),
1809 - MUST NOT be used on messages generated by automatic or periodic
1810 processes, except for messages which are automatic responses to
1813 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1814 to the default Received: header definition.
1816 PH/12 Added log selector acl_warn_skipped (default on).
1818 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1819 variables because (a) they are in the wrong storage pool and (b) even if
1820 they were copied, it wouldn't work properly because of the caching.
1822 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1823 checking when decoding. Apparently there are clients that generate
1824 overlong encoded strings. Why am I not surprised?
1826 PH/15 If the first argument of "${if match_address" was not empty, but did not
1827 contain an "@" character, Exim crashed. Now it writes a panic log message
1828 and treats the condition as false.
1830 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1832 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1833 extension "envelope-auth". The code is finished and in agreement with
1834 other implementations, but there is no documentation so far and in fact,
1835 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1836 not changing the active code.
1838 Print executed "if" and "elsif" statements when debugging is used. This
1839 helps a great deal to understand what a filter does.
1841 Document more things not specified clearly in RFC3028. I had all this
1842 sorted out, when out of a sudden new issues came to my mind. Oops."
1844 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1847 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1848 canonical form (as documented). However, after a host name lookup from
1849 the IP address, check_host() was doing a simple string comparison with
1850 addresses acquired from the DNS when checking that the found name did
1851 have the original IP as one of its addresses. Since any found IPv6
1852 addresses are likely to be in abbreviated form, the comparison could
1853 fail. Luckily, there already exists a function for doing the comparison
1854 by converting both addresses to binary, so now that is used instead of
1855 the text comparison.
1857 PH/20 There was another similar case to PH/19, when a complete host name was
1858 given in a host list; looking up its IP address could give an abbreviated
1859 form, whereas the current host's name might or might not be abbreviated.
1860 The same fix has been applied.
1866 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1867 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1870 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1871 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1873 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1875 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1876 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1877 Also, for "Auto-Submitted", treat anything other than "no" as
1878 non-personal, in accordance with RFC 3834. (Previously it treated
1879 anything starting "auto-" as non-personal.)
1881 TF/01 The control=submission/name=... option had a problem with syntax
1882 errors if the name included a slash character. The /name= option
1883 now slurps the rest of the string, so it can include any characters
1884 but it must come last in the list of options (after /sender_retain
1887 PH/05 Some modifications to the interface to the fake nameserver for the new
1895 TK/01 Added the "success_on_redirect" address verification option. See
1896 NewStuff for rationale and an example.
1898 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1900 PH/02 Patch to exigrep to allow it to work on syslog lines.
1902 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1903 fread() to skip over the body file's header line, because in Cygwin the
1904 header line is locked and is inaccessible.
1906 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1907 co-exist for some time) to make it clear that it is the Exim ID that is
1908 referenced, not the Message-ID: header line.
1910 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1911 string_format() function, because snprintf() does not exist on all
1914 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1915 ${stat: expansion item.
1917 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1918 protocol synchronization error", to keep the pedants happy.
1920 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1921 well as for IRIX systems, when gcc is being used. See the host.c source
1924 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1926 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1929 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1930 transport and to the smtp transport in LMTP mode.
1932 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1934 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1935 run a filter in a subprocess. This could lead to confusion in subsequent
1936 lookups in the parent process. There should also be a search_tidyup() at
1937 the end of the subprocess.
1939 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1940 only if the host matched helo_try_verify_hosts, which caused the
1941 verification to occur when the EHLO/HELO command was issued. The ACL just
1942 tested the remembered result. Now, if a previous verification attempt has
1943 not happened, "verify = helo" does it there and then.
1945 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1947 TK/03 Fix log output including CR from clamd.
1949 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1950 exist provoked a memory error which could cause a segfault.
1952 PH/15 Installed PCRE 6.2
1954 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1956 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1957 of the problem. Specifically, suggested +O2 rather than +O1 for the
1960 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1962 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1963 record, the port was not correctly logged when the outgoing_port log
1964 selector was set (it logged the transort's default port).
1966 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1967 fallback_hosts, and "hosts" in the smtp transport.
1969 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1970 host errors such as "Connection refused".
1972 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1973 authentication with radiusclient 0.4.9:
1975 - Error returned from rc_read_config was caught wrongly
1976 - Username/password not passed on to radius server due to wrong length.
1978 The presumption is that some radiusclient API changes for 4.51/PH/17
1979 were not taken care of correctly. The code is still untested by me (my
1980 Linux distribution still has 0.3.2 of radiusclient), but it was
1981 contributed by a Radius user.
1983 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1984 expanding the "port" option of the smtp transport.
1986 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1987 while reading a MIME header. Thanks to Tom Hughes for a patch.
1989 PH/24 Include config.h inside local_scan.h so that configuration settings are
1992 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1993 that in an ACL for RCPT (for example), you can examine exactly what was
1996 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1997 commands, but it was not correctly comparing the address with the actual
1998 client host address. Thus, it would show the EHLO address in Received:
1999 header lines when this was not necessary.
2001 PH/27 Added the % operator to ${eval:}.
2003 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2004 it should be ignoring failures (because with -C, for example, it has lost
2005 privilege). It wasn't ignoring creation failures other than "already
2008 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2011 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2012 because an input line was too long, either on its own, or by virtue of
2013 too many continuations, the temporary file was not being removed, and the
2014 return code was incorrect.
2016 PH/31 Missing "BOOL" in function definition in filtertest.c.
2018 PH/32 Applied Sieve patches from the maintainer.
2020 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2022 PH/33 Added "verify = not_blind".
2024 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2025 Local/Makefile (with some defaults set). These are used in built scripts
2026 such as exicyclog, but they have never been used in the exim_install
2027 script (though there are many overriding facilities there). I have
2028 arranged that the exim_install script now takes note of these two
2031 PH/35 Installed configuration files for Dragonfly.
2033 PH/36 When a locally submitted message by a trusted user did not contain a
2034 From: header, and the sender address was obtained from -f or from an SMTP
2035 MAIL command, and the trusted user did not use -F to supply a sender
2036 name, $originator_name was incorrectly used when constructing a From:
2037 header. Furthermore, $originator_name was used for submission mode
2038 messages from external hosts without From: headers in a similar way,
2039 which is clearly wrong.
2041 PH/37 Added control=suppress_local_fixups.
2043 PH/38 When log_selector = +received_sender was set, and the addition of the
2044 sender made the log line's construction buffer exactly full, or one byte
2045 less than full, an overflow happened when the terminating "\n" was
2048 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2049 when the result of a list match is failure because a DNS lookup failed.
2051 PH/40 RM_COMMAND is now used in the building process.
2053 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2054 the "build-* directories that it finds.
2056 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2057 address in a domain literal was a prefix of an interface address.
2059 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2060 when verifying a sender address, unless rewrite_headers is false.
2062 PH/44 Wrote a long comment about why errors_to addresses are verified as
2063 recipients, not senders.
2065 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2066 the ratelimit ACL was added.
2068 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2070 PH/46 Added extra information about PostgreSQL errors to the error string.
2072 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2073 avoiding the need to install special zones in a real server. This is
2074 backwards compatible; if it can't find the fake resolver, it drops back.
2075 Thus, both old and new test suites can be run.
2077 TF/02 Added util/ratelimit.pl
2079 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2080 clock is set back in time.
2082 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2083 Candler <B.Candler@pobox.com>.
2085 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2086 OK for addresses that are the result of redirections.
2088 PH/48 A number of further additions for the benefit of the new test suite,
2089 including a fake gethostbyname() that interfaces to the fake DNS resolver
2092 TF/06 The fix for widen_domains has also been applied to qualify_single and
2093 search_parents which are the other dnslookup options that can cause
2096 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2099 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2101 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2102 service" if it encountered an MX record with an empty target hostname.
2103 The message is now "an MX or SRV record indicated no SMTP service".
2105 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2106 if the DNS of the sending site is misconfigured. This is quite a
2107 common situation. This change restores the behaviour of treating a
2108 helo verification defer as a failure.
2110 PH/51 If self=fail was set on a router, the bounce message did not include the
2111 actual error message.
2117 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2119 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2120 command itself ended in error, the underlying message about the transport
2121 filter timeout was being overwritten with the pipe command error. Now the
2122 underlying error message should be appended to the second error message.
2124 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2126 PH/02 Reduce the amount of output that "make" produces by default. Full output
2127 can still be requested.
2129 PH/03 The warning log line about a condition test deferring for a "warn" verb
2130 was being output only once per connection, rather than after each
2131 occurrence (because it was using the same function as for successful
2132 "warn" verbs). This seems wrong, so I have changed it.
2134 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2135 it should not have, which might have caused a crash in the right
2136 circumstances, but probably never did.
2138 PH/04 Installed a modified version of Tony Finch's patch to make submission
2139 mode fix the return path as well as the Sender: header line, and to
2140 add a /name= option so that you can make the user's friendly name appear
2143 TF/03 Added the control = fakedefer ACL modifier.
2145 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2146 Mark Lowes for thorough testing.
2148 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2150 TK/03 Merged latest SRS patch from Miles Wilton.
2152 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2153 with the definition in sysexits.h (which is #included earlier).
2154 Fortunately, Exim does not actually use EX_OK. The code used to try to
2155 preserve the sysexits.h value, by assumimg that macro definitions were
2156 scanned for macro replacements. I have been disabused of this notion,
2157 so now the code just undefines EX_OK before #including unistd.h.
2159 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2160 in the smtp transport. When a block could not be written in a single
2161 write() function, the timeout was being re-applied to each part-write.
2162 This seems wrong - if the receiver was accepting one byte at a time it
2163 would take for ever. The timeout is now adjusted when this happens. It
2164 doesn't have to be particularly precise.
2166 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2167 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2169 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2170 without a domain if the check to <postmaster@domain> fails.
2172 SC/01 Eximstats: added -xls and the ability to specify output files
2173 (patch written by Frank Heydlauf).
2175 SC/02 Eximstats: use FileHandles for outputing results.
2177 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2179 SC/04 Eximstats: fixed display of large numbers with -nvr option
2181 SC/05 Eximstats: fixed merging of reports with empty tables.
2183 SC/06 Eximstats: added the -include_original_destination flag
2185 SC/07 Eximstats: removed tabs and trailing whitespace.
2187 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2189 TK/06 MBOX spool code: Add real "From " MBOX separator line
2190 so the .eml file is really in mbox format (even though
2191 most programs do not really care). Patch from Alex Miller.
2193 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2194 The latter is generated from $received_to and is only set if the
2195 message has one envelope recipient. SA can use these headers,
2196 obviously out-of-the-box. Patch from Alex Miller.
2198 PH/08 The ${def test on a variable was returning false if the variable's
2199 value was "0", contrary to what the specification has always said!
2200 The result should be true unless the variable is empty.
2202 PH/09 The syntax error of a character other than { following "${if
2203 def:variable_name" (after optional whitespace) was not being diagnosed.
2204 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2205 accidental colon was present, for example, could give incorrect results.
2207 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2208 result is used (not including appendfile, where other changes are about
2211 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2212 This involved changing a lot of size variables from int to off_t. It
2213 should work with maildirs and everything.
2215 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2216 spamd dying while we are connected to it.
2218 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2221 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2222 function for BDB 4.3.
2224 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2226 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2227 the sources for PCRE have been re-arranged and more files are now
2230 PH/15 The code I had for printing potentially long long variables in PH/11
2231 above was not the best (it lost precision). The length of off_t variables
2232 is now inspected at build time, and an appropriate printing format (%ld
2233 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2234 to be "long long int" or "long int". This is needed for the internal
2235 formatting function string_vformat().
2237 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2238 the configuration file to be ":syslog", then the script "guesses" where
2239 the logs files are, rather than using the compiled in default. In our
2240 case the guess is not the same as the compiled default, so the script
2241 suddenly stopped working when I started to use syslog. The patch checks
2242 to see if log_file_path is "". If so, it attempts to read it from exim
2243 with no configuration file to get the compiled in version, before it
2244 falls back to the previous guessing code."
2246 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2247 implementing BATV in an Exim configuration. See NewStuff for the gory
2250 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2251 Makefile that are specific to HP-UX.
2253 PH/18 If the "use_postmaster" option was set for a recipient callout together
2254 with the "random" option, the postmaster address was used as the MAIL
2255 FROM address for the random test, but not for the subsequent recipient
2256 test. It is now used for both.
2258 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2259 patch removes a few documentation additions to RFC 3028, because the
2260 latest draft now contains them. It adds the new en;ascii-case comparator
2261 and a new error check for 8bit text in MIME parts. Comparator and
2262 require names are now matched exactly. I enabled the subaddress
2263 extension, but it is not well tested yet (read: it works for me)."
2265 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2266 rework some of the code of TK/09 above to avoid the hardwired use of
2267 "%lld" and "long long". Replaced the call to snprintf() with a call to
2270 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2271 records point to non-existent hosts", "retry timeout exceeded", and
2272 "retry time not reached for any host after a long failure period".
2274 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2275 experimental DomainKeys support:
2277 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2278 (2) On an error such as an illegally used "control", the wrong name for
2279 the control was given.
2281 These problems did NOT occur unless DomainKeys support was compiled.
2283 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2285 PH/24 Added ${if match_ip condition.
2287 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2288 databases so that it will be absolutely obvious if a crash occurs in the
2289 DB library. This is a regular occurrence (often caused by mis-matched
2292 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2293 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2294 picked up on a user's system that detects such things. There doesn't seem
2295 to be a gcc warning option for this - only an attribute that has to be
2296 put on the function's prototype. It seems that in Fedora Core 4 they have
2297 set this on a number of new functions. No doubt there will be more in due
2300 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2301 specify a transport. However, if an address that was verified by such a
2302 router was the subject of a callout, Exim crashed because it tried to
2303 read the rcpt_include_affixes from the non-existent transport. Now it
2304 just assumes that the setting of that option is false. This bug was
2305 introduced by 4.51/PH/31.
2307 PH/28 Changed -d+all to exclude +memory, because that information is very
2308 rarely of interest, but it makes the output a lot bigger. People tend to
2309 do -d+all out of habit.
2311 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2312 code in os-type was giving problems when libc.so lives in lib64, like on
2315 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2316 aren't the modern standard, and it seems that some systems' include files
2317 don't always have them. Exim was already checking for some of the newer
2318 ones like T_AAAA, and defining it itself. I've added checks for all the
2319 record types that Exim uses.
2321 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2322 not automatically generating a new one, as it is supposed to. This
2323 prevented TLS from working. If the file did exist, but contained invalid
2324 data, a new version was generated, as expected. It was only the case of a
2325 non-existent file that was broken.
2327 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2328 with a change in libdomainkeys > 0.64.
2330 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2331 from DNS. If the selector record carries the flag, it now has
2332 precedence over the domain-wide flag.
2334 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2336 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2337 the use of a port name that isn't defined in /etc/services) occurred, the
2338 message was deferred as in a normal delivery, and thus remained on the
2339 spool, instead of being failed because of the mua_wrapper setting. This
2340 is now fixed, and I tidied up some of the mua_wrapper messages at the
2343 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2344 the messages in a hash of arrays rather than using individual hashes.
2345 This is a bit cleaner and results in dramatic memory savings, albeit
2346 at a slight CPU cost.
2348 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2349 as requested by Marc Sherman.
2351 SC/10 Eximstats: added histograms for user specified patterns as requested
2354 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2356 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2357 fopen() in the content-scanning modules that did not already have it.
2363 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2364 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2366 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2368 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2370 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2371 to test Sieve filters that use "vacation".
2373 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2374 that changes the way the GnuTLS parameters are stored in the cache file.
2375 The new format can be generated externally. For backward compatibility,
2376 if the data in the cache doesn't make sense, Exim assumes it has read an
2377 old-format file, and it generates new data and writes a new file. This
2378 means that you can't go back to an older release without removing the
2381 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2382 work if there are any delivery delays because "one_time" forces the
2383 parent to be marked "delivered", so its unseen clone is never tried
2384 again. For this reason, Exim now forbids the simultaneous setting of
2387 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2388 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2389 entry if you want to know the details. The fix, however, neglected to
2390 consider the case when local delivery batching is involved. The test for
2391 "previously delivered" was not happening when checking to see if an
2392 address could be batched with a previous (undelivered) one; under
2393 certain circumstances this could lead to multiple deliveries to the same
2396 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2397 in its include files, and this causes problems building Exim.
2399 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2400 header_syntax) but Exim was just ignoring anything given after a slash.
2401 In particular, this caused confusion with an attempt to use "verify =
2402 reverse_host_lookup/defer_ok". An error is now given when options are
2403 supplied for verify items that do not have them. (Maybe reverse_host_
2404 lookup should have a defer_ok option, but that's a different point.)
2406 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2407 defined by RFC 821) to 2048, because there were problems with some AUTH
2408 commands, and RFC 1869 says the size should be increased for extended
2409 SMTP commands that take arguments.
2411 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2414 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2415 "unknown" error; now it says that the functionality isn't in the binary.
2417 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2418 an address' error message when a string expansion fails (syntax or
2419 whatever). Otherwise the password may appear in the log. Following change
2420 PH/42 below, there is no longer a chance of it appearing in a bounce
2423 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2425 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2426 of its addresses was ever tried. (Bugzilla bug #2.)
2428 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2429 the result incorrectly in the debug output. (It correctly added a newline
2430 to what was transported.)
2432 TF/01 Added $received_time.
2434 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2435 commented out examples of how to interface to a virus scanner and to
2436 SpamAssassin. Also added commented examples of av_scanner and
2437 spamd_address settings.
2439 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2440 and controls are allowed in which ACLs. There were a couple of minor
2441 errors. Some of the entries in the conditions table (which is a table of
2442 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2443 negation of where the condition IS allowed.
2445 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2447 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2448 header file does not have a version number, so I've had to invent a new
2449 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2450 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2451 radiusclient), but it was contributed by a Radius user.
2453 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2454 files or directories, for OpenSSL.
2456 PH/19 When an Exim process that is running as root has to create an Exim log
2457 file, it does so in a subprocess that runs as exim:exim so as to get the
2458 ownership right at creation (otherwise, other Exim processes might see
2459 the file with the wrong ownership). There was no test for failure of this
2460 fork() call, which would lead to the process getting stuck as it waited
2461 for a non-existent subprocess. Forks do occasionally fail when resources
2462 run out. I reviewed all the other calls to fork(); they all seem to check
2465 PH/20 When checking for unexpected SMTP input at connect time (before writing
2466 the banner), Exim was not dealing correctly with a non-positive return
2467 from the read() function. If the client had disconnected by this time,
2468 the result was a log entry for a synchronization error with an empty
2469 string after "input=" when read() returned zero. If read() returned -1
2470 (an event I could not check), uninitialized data bytes were printed.
2471 There were reports of junk text (parts of files, etc) appearing after
2474 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2476 PH/22 Added support for macro redefinition, and (re)definition in between
2477 driver and ACL definitions.
2479 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2480 forgetting to use the resulting value; it was using the unexpanded value.
2482 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2483 hadn't been configured. The fix is from Juergen Kreileder, who
2484 understands it better than I do:
2486 "Here's what I see happening with three configured cyrus_sasl
2487 authenticators configured (plain, login, cram-md5):
2489 On startup auth_cyrus_sasl_init() gets called for each of these.
2490 This means three calls to sasl_listmech() without a specified mech_list.
2491 => SASL tests which mechs of all available mechs actually work
2492 => three warnings about OTP not working
2493 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2495 With the patch, sasl_listmech() also gets called three times. But now
2496 SASL's mech_list option is set to the server_mech specified in the the
2497 authenticator. Or in other words, the answer from sasl_listmech()
2498 gets limited to just the mech you're testing for (which is different
2500 => the return list contains just 'plain' or 'login', 'cram-md5' or
2501 nothing depending on the value of ob->server_mech.
2503 I've just tested the patch: Authentication still works fine,
2504 unavailable mechs specified in the exim configuration are still
2505 caught, and the auth.log warnings about OTP are gone."
2507 PH/25 When debugging is enabled, the contents of the command line are added
2508 to the debugging output, even when log_selector=+arguments is not
2511 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2512 answer is "GNU", and only if the return is "GNU/something" is the answer
2515 PH/27 $acl_verify_message is now set immediately after the failure of a
2516 verification in an ACL, and so is available in subsequent modifiers. In
2517 particular, the message can be preserved by coding like this:
2519 warn !verify = sender
2520 set acl_m0 = $acl_verify_message
2522 Previously, $acl_verify_message was set only while expanding "message"
2523 and "log_message" when a very denied access.
2525 PH/28 Modified OS/os.c-Linux with
2527 -#ifndef OS_LOAD_AVERAGE
2528 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2530 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2531 nomenclature these days.)
2533 PH/29 Installed patch from the Sieve maintainer that adds the options
2534 sieve_useraddress and sieve_subaddress to the redirect router.
2536 PH/30 In these circumstances:
2537 . Two addresses routed to the same list of hosts;
2538 . First host does not offer TLS;
2539 . First host accepts first address;
2540 . First host gives temporary error to second address;
2541 . Second host offers TLS and a TLS session is established;
2542 . Second host accepts second address.
2543 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2544 and peerdn, if requested) that were in fact used only for the second
2547 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2548 attention to any local part prefix or suffix that was matched by the
2549 router that accepted the address. It now behaves in the same way as it
2550 does for delivery: the affixes are removed from the local part unless
2551 rcpt_include_affixes is set on the transport.
2553 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2554 timeout during the DATA phase of an incoming message.
2556 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2557 applied a patch sanctioned by the Sieve maintainer.
2559 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2560 the uid or gid is negative. A case of a negative gid caused this to be
2561 noticed. The fix allows for either to be negative.
2563 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2564 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2567 PH/36 The change PH/12 above was broken. Fixed it.
2569 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2570 the grounds that routing the same address twice would always produce the
2571 same answer. This might have been true once, but it is certainly no
2572 longer true now. Routing a child address may depend on the previous
2573 routing that produced that child. Some complicated redirection strategies
2574 went wrong when messages had multiple recipients, and made Exim's
2575 behaviour dependent on the order in which the addresses were given.
2577 I have moved the duplicate checking until after the routing is complete.
2578 Exim scans the addresses that are assigned to local and remote
2579 transports, and removes any duplicates. This means that more work will be
2580 done, as duplicates will always all be routed, but duplicates are
2581 presumably rare, so I don't expect this is of any significance.
2583 For deliveries to pipes, files, and autoreplies, the duplicate checking
2584 still happens during the routing process, since they are not going to be
2587 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2588 It corrects a timeout issue with spamd. This is Ian's comment: "The
2589 background is that sometimes spamd either never reads data from a
2590 connection it has accepted, or it never writes response data. The exiscan
2591 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2592 blindly assumes that writes won't block so it may never time out."
2594 PH/39 Allow G after quota size as well as K and M.
2596 PH/40 The value set for $authenticated_id in an authenticator may not contain
2597 binary zeroes or newlines because the value is written to log lines and
2598 to spool files. There was no check on this. Now the value is run through
2599 the string_printing() function so that such characters are converted to
2600 printable escape sequences.
2602 PH/41 $message_linecount is a new variable that contains the total number of
2603 lines in the message. Compare $body_linecount, which is the count for the
2606 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2607 bounce and delay warning messages, except in certain special cases, which
2610 (a) An SMTP error message from a remote host;
2611 (b) A message specified in a :fail: redirection;
2612 (c) A message specified in a "fail" command in a system filter;
2613 (d) A message specified in a FAIL return from the queryprogram router;
2614 (e) A message specified by the cannot_route_message router option.
2616 In these cases only, Exim does include the error details in bounce and
2617 warning messages. There are also a few cases where bland messages such
2618 as "unrouteable address" or "local delivery error" are given.
2620 PH/43 $value is now also set for the "else" part of a ${run expansion.
2622 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2623 being worked on, but at least Exim now implements the latest version to
2626 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2627 process to complete was treated as a delivery failure, a timeout while
2628 writing the message to the pipe was logged, but erroneously treated as a
2629 successful delivery. Such timeouts include transport filter timeouts. For
2630 consistency with the overall process timeout, these timeouts are now
2631 treated as errors, giving rise to delivery failures by default. However,
2632 there is now a new Boolean option for the pipe transport called
2633 timeout_defer, which, if set TRUE, converts the failures into defers for
2634 both kinds of timeout. A transport filter timeout is now identified in
2637 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2638 systems where "make" and "gmake" are different, calling "gmake" at top
2639 level broke things. I've arranged for the value of $(MAKE) to be passed
2640 from the Makefile to this script so that it can call the same version of
2644 A note about Exim versions 4.44 and 4.50
2645 ----------------------------------------
2647 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2648 changes of various kinds. As a consequence, a big documentation update was
2649 needed. This delayed the release for rather longer than seemed good, especially
2650 in the light of a couple of (minor) security issues. Therefore, the changes
2651 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2652 release. So 4.44 and 4.50 are in effect two different branches that both start
2655 I have left the 4.50 change log unchanged; it contains all the changes since
2656 4.43. The change log for 4.44 is below; many of its items are identical to
2657 those for 4.50. This seems to be the most sensible way to preserve the
2658 historical information.
2664 1. Minor wording change to the doc/README.SIEVE file.
2666 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2667 computation of the current number of files was incorrect.
2669 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2670 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2673 4. Give more explanation in the error message when the command for a transport
2674 filter fails to execute.
2676 5. There are several places where Exim runs a non-Exim command in a
2677 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2678 was being done only for the command run by the queryprogram router. It is
2679 now done for all such subprocesses. The other cases are: ${run, transport
2680 filters, and the commands run by the lmtp and pipe transports.
2682 6. Added CONFIGURE_GROUP build-time option.
2684 7. Some older OS have a limit of 256 on the maximum number of file
2685 descriptors. Exim was using setrlimit() to set 1000 as a large value
2686 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2687 systems. I've change it so that if it can't get 1000, it tries for 256.
2689 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2690 was an oversight, and furthermore, ever since the addition of extra
2691 controls (e.g. 4.43/32), the checks on when to allow different forms of
2692 "control" were broken. There should now be diagnostics for all cases when a
2693 control that does not make sense is encountered.
2695 9. Added the /retain_sender option to "control=submission".
2697 10. $recipients is now available in the predata ACL (oversight).
2699 11. Tidy the search cache before the fork to do a delivery from a message
2700 received from the command line. Otherwise the child will trigger a lookup
2701 failure and thereby defer the delivery if it tries to use (for example) a
2702 cached ldap connection that the parent has called unbind on.
2704 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2705 of $address_data from the recipient verification was clobbered by the
2706 sender verification.
2708 13. The value of address_data from a sender verification is now available in
2709 $sender_address_data in subsequent conditions in the ACL statement.
2711 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2713 15. Added a new option "connect=<time>" to callout options, to set a different
2716 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2717 was its contents. (It was OK if the option was not defined at all.)
2719 17. A "Completed" log line is now written for messages that are removed from
2720 the spool by the -Mrm option.
2722 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2723 information about exactly what failed.
2725 19. Added -dd to debug only the daemon process.
2727 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2728 handles timeouts, both on the server side and network timeouts. Renamed the
2729 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2731 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2732 It is now set to "smtps".
2734 22. $host_address is now set to the target address during the checking of
2735 ignore_target_hosts.
2737 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2738 being passed; this would have caused $sender_host_name to have been used if
2739 matching the list had actually called for a host name (not very likely,
2740 since this list is usually IP addresses). A host name is now passed as
2743 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2744 code that specifies a non-restarting handler (typically sigaction() in
2745 modern systems) in an attempt to fix a rare and obscure crash bug.
2747 25. Narrowed the window for a race in the daemon that could cause it to ignore
2748 SIGCHLD signals. This is not a major problem, because they are used only to
2749 wake it up if nothing else does.
2751 26. A malformed maildirsize file could cause Exim to calculate negative values
2752 for the mailbox size or file count. Odd effects could occur as a result.
2753 The maildirsize information is now recalculated if the size or filecount
2756 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2757 support for a long time. Removed HAVE_SYS_VFS_H.
2759 28. Installed the latest version of exipick from John Jetmore.
2761 29. In an address list, if the pattern was not a regular expression, an empty
2762 subject address (from a bounce message) matched only if the pattern was an
2763 empty string. Non-empty patterns were not even tested. This was the wrong
2764 because it is perfectly reasonable to use an empty address as part of a
2765 database query. An empty address is now tested by patterns that are
2766 lookups. However, all the other forms of pattern expect the subject to
2767 contain a local part and a domain, and therefore, for them, an empty
2768 address still always fails if the pattern is not itself empty.
2770 30. Exim went into a mad DNS loop when attempting to do a callout where the
2771 host was specified on an smtp transport, and looking it up yielded more
2772 than one IP address.
2774 31. Re-factored the code for checking spool and log partition space into a
2775 function that finds that data and another that does the check. The former
2776 is then used to implement four new variables: $spool_space, $log_space,
2777 $spool_inodes, and $log_inodes.
2779 32. The RFC2047 encoding function was originally intended for short strings
2780 such as real names; it was not keeping to the 75-character limit for
2781 encoded words that the RFC imposes. It now respects the limit, and
2782 generates multiple encoded words if necessary. To be on the safe side, I
2783 have increased the buffer size for the ${rfc2047: expansion operator from
2786 33. It is now permitted to omit both strings after an "if" condition; if the
2787 condition is true, the result is "true". As before, when the second string
2788 is omitted, a false condition yields an empty string. This makes it less
2789 cumbersome to write custom ACL and router conditions.
2791 34. Failure to deliver a bounce message always caused it to be frozen, even if
2792 there was an errors_to setting on the router. The errors_to setting is now
2795 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2796 canonical form (fully expanded) before being placed in
2797 $sender_host_address.
2799 36. The table in the code that translates DNS record types into text (T_A to
2800 "A" for instance) was missing entries for NS and CNAME. It is just possible
2801 that this could have caused confusion if both these types were looked up
2802 for the same domain, because the text type is used as part of Exim's
2803 per-process caching. But the chance of anyone hitting this buglet seems
2806 37. The dnsdb lookup has been extended in a number of ways.
2808 (1) There is a new type, "zns", which walks up the domain tree until it
2809 finds some nameserver records. It should be used with care.
2811 (2) There is a new type, "mxh", which is like "mx" except that it returns
2812 just the host names, not the priorities.
2814 (3) It is now possible to give a list of domains (or IP addresses) to be
2815 looked up. The behaviour when one of the lookups defers can be
2816 controlled by a keyword.
2818 (4) It is now possible to specify the separator character for use when
2819 multiple records are returned.
2821 38. The dnslists ACL condition has been extended: it is now possible to supply
2822 a list of IP addresses and/or domains to be looked up in a particular DNS
2825 39. Added log_selector=+queue_time_overall.
2827 40. When running the queue in the test harness, wait just a tad after forking a
2828 delivery process, to get repeatability of debugging output.
2830 41. Include certificate and key file names in error message when GnuTLS fails
2831 to set them up, because the GnuTLS error message doesn't include the name
2832 of the failing file when there is a problem reading it.
2834 42. Allow both -bf and -bF in the same test run.
2836 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2838 44. The "Exiscan patch" is now merged into the mainline Exim source.
2840 45. Sometimes the final signoff response after QUIT could fail to get
2841 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2842 < 0 before doing a fflush(). This bug looks as though it goes back to the
2843 introduction of TLS in release 3.20, but "sometimes" must have been rare
2844 because the tests only now provoked it.
2846 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2847 (this can affect the format of dates).
2849 47. exim_tidydb, when checking for the continued existence of a message for
2850 which it has found a message-specific retry record, was not finding
2851 messages that were in split spool directories. Consequently, it was
2852 deleting retry records that should have stayed in existence.
2854 48. Steve fixed some bugs in eximstats.
2856 49. The SPA authentication driver was not abandoning authentication and moving
2857 on to the next authenticator when an expansion was forced to fail,
2858 contradicting the general specification for all authenticators. Instead it
2859 was generating a temporary error. It now behaves as specified.
2861 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2862 (the order specifies the preference for clients). The order is now AES256,
2863 AES128, 3DES, ARCFOUR128.
2865 51. Small patch to Sieve code - explicitly set From: when generating an
2868 52. Exim crashed if a remote delivery caused a very long error message to be
2869 recorded - for instance if somebody sent an entire SpamAssassin report back
2870 as a large number of 550 error lines. This bug was coincidentally fixed by
2871 increasing the size of one of Exim's internal buffers (big_buffer) that
2872 happened as part of the Exiscan merge. However, to be on the safe side, I
2873 have made the code more robust (and fixed the comments that describe what
2876 53. Now that there can be additional text after "Completed" in log lines (if
2877 the queue_time_overall log selector is set), a one-byte patch to exigrep
2878 was needed to allow it to recognize "Completed" as not the last thing in
2881 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2882 patch that reportedly fixes this has been added. I am not expert enough to
2883 create a test for it. This is what the patch creator wrote:
2885 "I found a little strange behaviour of ldap code when working with
2886 Windows 2003 AD Domain, where users was placed in more than one
2887 Organization Units. When I tried to give exim partial DN, the exit code
2888 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2889 But simultaneously result of request was absolutely normal ldap result,
2890 so I produce this patch..."
2892 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2893 REFERENCE, so I have modified the code to exclude the patch when that macro
2896 55. Some experimental protocols are using DNS PTR records for new purposes. The
2897 keys for these records are domain names, not reversed IP addresses. The
2898 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2899 leaves it alone. Component reversal etc. now happens only for IP addresses.
2902 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2904 57. Double the size of the debug message buffer (to 2048) so that more of very
2905 long debug lines gets shown.
2907 58. The exicyclog utility now does better if the number of log files to keep
2908 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2910 59. Two changes related to the smtp_active_hostname option:
2912 (1) $smtp_active_hostname is now available as a variable.
2913 (2) The default for smtp_banner uses $smtp_active_hostname instead
2914 of $primary_hostname.
2916 60. The host_aton() function is supposed to be passed a string that is known
2917 to be a valid IP address. However, in the case of IPv6 addresses, it was
2918 not checking this. This is a hostage to fortune. Exim now panics and dies
2919 if the condition is not met. A case was found where this could be provoked
2920 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2921 components; fortuitously, this particular loophole had already been fixed
2922 by change 4.50/55 above.
2924 If there are any other similar loopholes, the new check in host_aton()
2925 itself should stop them being exploited. The report I received stated that
2926 data on the command line could provoke the exploit when Exim was running as
2927 exim, but did not say which command line option was involved. All I could
2928 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2929 running as the user.
2932 61. There was a buffer overflow vulnerability in the SPA authentication code
2933 (which came originally from the Samba project). I have added a test to the
2934 spa_base64_to_bits() function which I hope fixes it.
2937 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2938 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2940 63. The daemon start-up calls getloadavg() while still root for those OS that
2941 need the first call to be done as root, but it missed one case: when
2942 deliver_queue_load_max is set with deliver_drop_privilege. This is
2943 necessary for the benefit of the queue runner, because there is no re-exec
2944 when deliver_drop_privilege is set.
2946 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2947 This has been fixed.
2949 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2950 was in use, was not putting the data itself into the right store pool;
2951 consequently, it could be overwritten for a subsequent message in the same
2952 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2955 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2957 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2958 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2959 the function were treating the return as a boolean value, which happened to
2960 work because 0=false and not-0=true, but is not correct code.
2962 68. The host_aton() function was not handling scoped IPv6 addresses (those
2963 with, for example, "%eth0" on the end) correctly.
2965 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2966 negated items (that is, ~something) in unsigned ints. Some compilers
2967 apparently mutter when there is no cast.
2969 70. If an address verification called from an ACL failed, and did not produce a
2970 user-specific message (i.e. there was only a "system" message), nothing was
2971 put in $acl_verify_message. In this situation, it now puts the system
2974 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2975 session; change 4.31/43 added the unwanted input to the log line - except
2976 that it did not do this in the start of session case. It now does.
2978 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2979 This is wrong and can cause the other end to generate a synchronization
2980 error if it is another Exim or anything else that does the synchronization
2981 check. A QUIT command is no longer sent after a timeout.
2983 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2984 during host lookups.
2986 74. The defer_ok option of callout verification was not working if it was used
2987 when verifying addresses in header lines, that is, for this case:
2989 verify = header_sender/callout=defer_ok
2991 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2992 those file descriptors could be used for SMTP connections. If anything
2993 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2994 could be sent to the SMTP client, causing chaos. The daemon now opens
2995 stdin, stdout, and stderr to /dev/null when it puts itself into the
2998 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2999 log by default. The user can override this with suitable Perl magic.
3001 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3002 the log message when discard triggers, was not working for the DATA ACL or
3003 for the non-SMTP ACL.
3005 78. Error message wording change in sieve.c.
3007 79. If smtp_accept_max_per_host was set, the number of connections could be
3008 restricted to fewer than expected, because the daemon was trying to set up
3009 a new connection before checking whether the processes handling previous
3010 connections had finished. The check for completed processes is now done
3011 earlier. On busy systems, this bug wouldn't be noticed because something
3012 else would have woken the daemon, and it would have reaped the completed
3015 80. If a message was submitted locally by a user whose login name contained one
3016 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3017 It caused a spool format error. I have fixed the spool reading code. A
3018 related problem was that the "from" clause in the Received: line became
3019 illegal because of the space(s). It is now covered by ${quote_local_part.
3021 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3024 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3026 83. Patch from Sieve maintainer to fix unterminated string problem in
3027 "vacation" handling.
3029 84. Some minor changes to the Linux configuration files to help with other
3030 OS variants using glibc.
3032 85. One more patch for Sieve to update vacation handling to latest spec.
3035 ----------------------------------------------------
3036 See the note above about the 4.44 and 4.50 releases.
3037 ----------------------------------------------------
3043 1. Change 4.43/35 introduced a bug that caused file counts to be
3044 incorrectly computed when quota_filecount was set in an appendfile
3047 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3048 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3051 3. Give more explanation in the error message when the command for a transport
3052 filter fails to execute.
3054 4. There are several places where Exim runs a non-Exim command in a
3055 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3056 was being done only for the command run by the queryprogram router. It is
3057 now done for all such subprocesses. The other cases are: ${run, transport
3058 filters, and the commands run by the lmtp and pipe transports.
3060 5. Some older OS have a limit of 256 on the maximum number of file
3061 descriptors. Exim was using setrlimit() to set 1000 as a large value
3062 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3063 systems. I've change it so that if it can't get 1000, it tries for 256.
3065 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3066 was an oversight, and furthermore, ever since the addition of extra
3067 controls (e.g. 4.43/32), the checks on when to allow different forms of
3068 "control" were broken. There should now be diagnostics for all cases when a
3069 control that does not make sense is encountered.
3071 7. $recipients is now available in the predata ACL (oversight).
3073 8. Tidy the search cache before the fork to do a delivery from a message
3074 received from the command line. Otherwise the child will trigger a lookup
3075 failure and thereby defer the delivery if it tries to use (for example) a
3076 cached ldap connection that the parent has called unbind on.
3078 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3079 of $address_data from the recipient verification was clobbered by the
3080 sender verification.
3082 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3083 was its contents. (It was OK if the option was not defined at all.)
3085 11. A "Completed" log line is now written for messages that are removed from
3086 the spool by the -Mrm option.
3088 12. $host_address is now set to the target address during the checking of
3089 ignore_target_hosts.
3091 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3092 being passed; this would have caused $sender_host_name to have been used if
3093 matching the list had actually called for a host name (not very likely,
3094 since this list is usually IP addresses). A host name is now passed as
3097 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3098 code that specifies a non-restarting handler (typically sigaction() in
3099 modern systems) in an attempt to fix a rare and obscure crash bug.
3101 15. Narrowed the window for a race in the daemon that could cause it to ignore
3102 SIGCHLD signals. This is not a major problem, because they are used only to
3103 wake it up if nothing else does.
3105 16. A malformed maildirsize file could cause Exim to calculate negative values
3106 for the mailbox size or file count. Odd effects could occur as a result.
3107 The maildirsize information is now recalculated if the size or filecount
3110 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3111 support for a long time. Removed HAVE_SYS_VFS_H.
3113 18. Updated exipick to current release from John Jetmore.
3115 19. Allow an empty sender to be matched against a lookup in an address list.
3116 Previously the only cases considered were a regular expression, or an
3119 20. Exim went into a mad DNS lookup loop when doing a callout where the
3120 host was specified on the transport, if the DNS lookup yielded more than
3123 21. The RFC2047 encoding function was originally intended for short strings
3124 such as real names; it was not keeping to the 75-character limit for
3125 encoded words that the RFC imposes. It now respects the limit, and
3126 generates multiple encoded words if necessary. To be on the safe side, I
3127 have increased the buffer size for the ${rfc2047: expansion operator from
3130 22. Failure to deliver a bounce message always caused it to be frozen, even if
3131 there was an errors_to setting on the router. The errors_to setting is now
3134 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3135 canonical form (fully expanded) before being placed in
3136 $sender_host_address.
3138 24. Updated eximstats to version 1.33
3140 25. Include certificate and key file names in error message when GnuTLS fails
3141 to set them up, because the GnuTLS error message doesn't include the name
3142 of the failing file when there is a problem reading it.
3144 26. Expand error message when OpenSSL has problems setting up cert/key files.
3147 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3148 (this can affect the format of dates).
3150 28. exim_tidydb, when checking for the continued existence of a message for
3151 which it has found a message-specific retry record, was not finding
3152 messages that were in split spool directories. Consequently, it was
3153 deleting retry records that should have stayed in existence.
3155 29. eximstats updated to version 1.35
3156 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3157 1.35 - bugfix such that pie charts by volume are generated correctly
3159 30. The SPA authentication driver was not abandoning authentication and moving
3160 on to the next authenticator when an expansion was forced to fail,
3161 contradicting the general specification for all authenticators. Instead it
3162 was generating a temporary error. It now behaves as specified.
3164 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3165 (the order specifies the preference for clients). The order is now AES256,
3166 AES128, 3DES, ARCFOUR128.
3168 31. Small patch to Sieve code - explicitly set From: when generating an
3171 32. Exim crashed if a remote delivery caused a very long error message to be
3172 recorded - for instance if somebody sent an entire SpamAssassin report back
3173 as a large number of 550 error lines. This bug was coincidentally fixed by
3174 increasing the size of one of Exim's internal buffers (big_buffer) that
3175 happened as part of the Exiscan merge. However, to be on the safe side, I
3176 have made the code more robust (and fixed the comments that describe what
3179 33. Some experimental protocols are using DNS PTR records for new purposes. The
3180 keys for these records are domain names, not reversed IP addresses. The
3181 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3182 leaves it alone. Component reversal etc. now happens only for IP addresses.
3185 34. The host_aton() function is supposed to be passed a string that is known
3186 to be a valid IP address. However, in the case of IPv6 addresses, it was
3187 not checking this. This is a hostage to fortune. Exim now panics and dies
3188 if the condition is not met. A case was found where this could be provoked
3189 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3190 components; fortuitously, this particular loophole had already been fixed
3191 by change 4.50/55 or 4.44/33 above.
3193 If there are any other similar loopholes, the new check in host_aton()
3194 itself should stop them being exploited. The report I received stated that
3195 data on the command line could provoke the exploit when Exim was running as
3196 exim, but did not say which command line option was involved. All I could
3197 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3198 running as the user.
3201 35. There was a buffer overflow vulnerability in the SPA authentication code
3202 (which came originally from the Samba project). I have added a test to the
3203 spa_base64_to_bits() function which I hope fixes it.
3206 36. The daemon start-up calls getloadavg() while still root for those OS that
3207 need the first call to be done as root, but it missed one case: when
3208 deliver_queue_load_max is set with deliver_drop_privilege. This is
3209 necessary for the benefit of the queue runner, because there is no re-exec
3210 when deliver_drop_privilege is set.
3212 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3213 was in use, was not putting the data itself into the right store pool;
3214 consequently, it could be overwritten for a subsequent message in the same
3215 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3218 38. Sometimes the final signoff response after QUIT could fail to get
3219 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3220 < 0 before doing a fflush(). This bug looks as though it goes back to the
3221 introduction of TLS in release 3.20, but "sometimes" must have been rare
3222 because the tests only now provoked it.
3228 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3229 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3230 Now it returns an integer. A number of calls were still expecting a T/F
3231 return. Fortuitously, in all cases, the tests worked in OK situations,
3232 which is the norm. However, things would have gone wrong on any write
3233 failures on the smtp file descriptor. This function is used when sending
3234 messages over SMTP and also when doing verify callouts.
3236 2. When Exim is called to do synchronous delivery of a locally submitted
3237 message (the -odf or -odi options), it no longer closes stderr before doing
3240 3. Implemented the mua_wrapper option.
3242 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3244 5. Implemented the functions header_remove(), header_testname(),
3245 header_add_at_position(), and receive_remove_recipient(), and exported them
3248 6. If an ACL "warn" statement specified the addition of headers, Exim already
3249 inserted X-ACL-Warn: at the start if there was no header name. However, it
3250 was not making this test for the second and subsequent header lines if
3251 there were newlines in the string. This meant that an invalid header could
3252 be inserted if Exim was badly configured.
3254 7. Allow an ACL "warn" statement to add header lines at the start or after all
3255 the Received: headers, as well as at the end.
3257 8. Added the rcpt_4xx retry error code.
3259 9. Added postmaster_mailfrom=xxx to callout verification option.
3261 10. Added mailfrom=xxxx to the callout verification option, for verify=
3264 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3265 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3267 12. Inserted some casts to stop certain compilers warning when using pointer
3268 differences as field lengths or precisions in printf-type calls (mostly
3269 affecting debugging statements).
3271 13. Added optional readline() support for -be (dynamically loaded).
3273 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3274 same clock tick as a message's arrival, so that its received time was the
3275 same as the "first fail" time on the retry record, and that message
3276 remained on the queue past the ultimate address timeout, every queue runner
3277 would try a delivery (because it was past the ultimate address timeout) but
3278 after another failure, the ultimate address timeout, which should have then
3279 bounced the address, did not kick in. This was a "< instead of <=" error;
3280 in most cases the first failure would have been in the next clock tick
3281 after the received time, and all would be well.
3283 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3284 being recognized when the domain list was tested by the match_domain
3285 condition in an expansion string.
3287 16. Added the ${str2b64: operator.
3289 17. Exim was always calling setrlimit() to set a large limit for the number of
3290 processes, without checking whether the existing limit was already
3291 adequate. (It did check for the limit on file descriptors.) Furthermore,
3292 errors from getrlimit() and setrlimit() were being ignored. Now they are
3293 logged to the main and panic logs, but Exim does carry on, to try to do its
3294 job under whatever limits there are.
3296 18. Imported PCRE 5.0.
3298 19. Trivial typo in log message " temporarily refused connection" (the leading
3301 20. If the log selector return_path_on_delivery was set and an address was
3302 redirected to /dev/null, the delivery process crashed because it assumed
3303 that a return path would always be set for a "successful" delivery. In this
3304 case, the whole delivery is bypassed as an optimization, and therefore no
3307 21. Internal re-arrangement: the function for sending a challenge and reading
3308 a response while authentication was assuming a zero-terminated challenge
3309 string. It's now changed to take a pointer and a length, to allow for
3310 binary data in such strings.
3312 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3314 23. Exim was not respecting finduser_retries when seeking the login of the
3315 uid under which it was called; it was always trying 10 times. (The default
3316 setting of finduser_retries is zero.) Also, it was sleeping after the final
3317 failure, which is pointless.
3319 24. Implemented tls_on_connect_ports.
3321 25. Implemented acl_smtp_predata.
3323 26. If the domain in control=submission is set empty, Exim assumes that the
3324 authenticated id is a complete email address when it generates From: or
3325 Sender: header lines.
3327 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3328 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3329 chown and chgrp in /bin and hostname in /usr/bin.
3331 28. Exim was keeping the "process log" file open after each use, just as it
3332 does for the main log. This opens the possibility of it remaining open for
3333 long periods when the USR1 signal hits a daemon. Occasional processlog
3334 errors were reported, that could have been caused by this. Anyway, it seems
3335 much more sensible not to leave this file open at all, so that is what now
3338 29. The long-running daemon process does not normally write to the log once it
3339 has entered its main loop, and it closes the log before doing so. This is
3340 so that log files can straightforwardly be renamed and moved. However,
3341 there are a couple of unusual error situations where the daemon does write
3342 log entries, and I had neglected to close the log afterwards.
3344 30. The text of an SMTP error response that was received during a remote
3345 delivery was being truncated at 512 bytes. This is too short for some of
3346 the long messages that one sometimes sees. I've increased the limit to
3349 31. It is now possible to make retry rules that apply only when a message has a
3350 specific sender, in particular, an empty sender.
3352 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3353 it possible to be selective about when SMTP synchronization is enforced.
3355 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3357 32. Implemented hosts_connection_nolog.
3359 33. Added an ACL for QUIT.
3361 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3364 35. Added mailbox_size and mailbox_filecount to appendfile.
3366 36. Added control = no_multiline_responses to ACLs.
3368 37. There was a bug in the logic of the code that waits for the clock to tick
3369 in the case where the clock went backwards by a substantial amount such
3370 that the microsecond fraction of "now" was more than the microsecond
3371 fraction of "then" (but the whole seconds number was less).
3373 38. Added support for the libradius Radius client library this is found on
3374 FreeBSD (previously only the radiusclient library was supported).
3380 1. When certain lookups returned multiple values in the form name=value, the
3381 quoting of the values was not always being done properly. Specifically:
3382 (a) If the value started with a double quote, but contained no whitespace,
3384 (b) If the value contained whitespace other than a space character (i.e.
3385 tabs or newlines or carriage returns) it was not quoted.
3386 This fix has been applied to the mysql and pgsql lookups by writing a
3387 separate quoting function and calling it from the lookup code. The fix
3388 should probably also be applied to nisplus, ibase and oracle lookups, but
3389 since I cannot test any of those, I have not disturbed their existing code.
3391 2. A hit in the callout cache for a specific address caused a log line with no
3392 reason for rejecting RCPT. Now it says "Previous (cached) callout
3393 verification failure".
3395 3. There was an off-by-one bug in the queryprogram router. An over-long
3396 return line was truncated at 256 instead of 255 characters, thereby
3397 overflowing its buffer with the terminating zero. As well as fixing this, I
3398 have increased the buffer size to 1024 (and made a note to document this).
3400 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3401 when Exim is waiting for an SMTP response from a remote server, Exim
3402 restarts its select() call on the socket, thereby resetting its timeout.
3403 This is not a problem when such interrupts are rare. Somebody set up a cron
3404 job to run exiwhat every 2 minutes, which is less than the normal select()
3405 timeout (5 or 10 minutes). This meant that the select() timeout never
3406 kicked in because it was always reset. I have fixed this by comparing the
3407 time when an interrupt arrives with the time at the start of the first call
3408 to select(). If more time than the timeout has elapsed, the interrupt is
3409 treated as a timeout.
3411 5. Some internal re-factoring in preparation for the addition of Sieve
3412 extensions (by MH). In particular, the "personal" test is moved to a
3413 separate function, and given an option for scanning Cc: and Bcc: (which is
3414 not set for Exim filters).
3416 6. When Exim created an email address using the login of the caller as the
3417 local part (e.g. when creating a From: or Sender: header line), it was not
3418 quoting the local part when it contained special characters such as @.
3420 7. Installed new OpenBSD configuration files.
3422 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3423 try to make them clearer.
3425 9. Callout options, other than the timeout value, were being ignored when
3426 verifying sender addresses in header lines. For example, when using
3428 verify = header_sender/callout=no_cache
3430 the cache was (incorrectly) being used.
3432 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3433 only the Cygwin environment.
3435 11. When return_path_on_delivery was set as a log selector, if different remote
3436 addresses in the same message used different return paths and parallel
3437 remote delivery occurred, the wrong values would sometimes be logged.
3438 (Whenever a remote delivery process finished, the return path value from
3439 the most recently started remote delivery process was logged.)
3441 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3442 lines when AUTH and/or TLS are in use. This is the "received protocol"
3443 field. Exim used to use "asmtp" for authenticated SMTP, without any
3444 indication (in the protocol name) for TLS use. Now it follows the RFC and
3445 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3446 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3447 names appear in log lines as well as in Received: header lines.
3449 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3450 extensions, and comparison tests, and to fix some bugs.
3452 14. Changes to the "personal" filter test:
3454 (1) The test was buggy in that it was just doing the equivalent of
3455 "contains" tests on header lines. For example, if a user's address was
3456 anne@some.where, the "personal" test would incorrectly be true for
3458 To: susanne@some.where
3460 This test is now done by extracting each address from the header in turn,
3461 and checking the entire address. Other tests that are part of "personal"
3462 are now done using regular expressions (for example, to check local parts
3463 of addresses in From: header lines).
3465 (2) The list of non-personal local parts in From: addresses has been
3466 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3467 taken from the Sieve specification recommendations.
3469 (3) If the message contains any header line starting with "List-" it is
3470 treated as non-personal.
3472 (4) The test for "circular" in the Subject: header line has been removed
3473 because it now seems ill-conceived.
3475 15. Minor typos in src/EDITME comments corrected.
3477 16. Installed latest exipick from John Jetmore.
3479 17. If headers_add on a router specified a text string that was too long for
3480 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3481 of string_sprintf() is now avoided.
3483 18. $message_body_size was not set (it was always zero) when running the DATA
3484 ACL and the local_scan() function.
3486 19. For the "mail" command in an Exim filter, no default was being set for
3487 the once_repeat time, causing a random time value to be used if "once" was
3488 specified. (If the value happened to be <= 0, no repeat happened.) The
3489 default is now 0s, meaning "never repeat". The "vacation" command was OK
3490 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3491 (I found it when inspecting the code).
3493 20. There is now an overall timeout for performing a callout verification. It
3494 defaults to 4 times the callout timeout, which applies to individual SMTP
3495 commands during the callout. The overall timeout applies when there is more
3496 than one host that can be tried. The timeout is checked before trying the
3497 next host. This prevents very long delays if there are a large number of
3498 hosts and all are timing out (e.g. when the network connections are timing
3499 out). The value of the overall timeout can be changed by specifying an
3500 additional sub-option for "callout", called "maxwait". For example:
3502 verify = sender/callout=5s,maxwait=20s
3504 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3505 to the end before writing, but this should make it even safer).
3507 22. Exim was forgetting that it had advertised PIPELINING for the second and
3508 subsequent messages on an SMTP connection. It was also not resetting its
3509 memory on STARTTLS and an internal HELO.
3511 23. When Exim logs an SMTP synchronization error within a session, it now
3512 records whether PIPELINING has been advertised or not.
3514 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3515 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3516 rather than long int.
3518 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3520 26. Added the never_mail option to autoreply.
3526 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3527 crash if the getsockname() call failed; this can happen if a connection is
3528 closed very soon after it is established. The problem was simply in the
3529 order in which certain operations were done, causing Exim to try to write
3530 to the SMTP stream before it had set up the file descriptor. The bug has
3531 been fixed by making things happen in the correct order.
3537 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3538 before the connection was closed, thus losing the rejection response.
3540 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3541 some early Solaris releases, but causes trouble in current releases where
3542 socklen_t is defined.
3544 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3547 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3550 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3551 that releasing the top of it at the end releases what was used for sub-
3552 expansions (unless the block got too big). However, discard this block if
3553 the first thing is a variable or header, so that we can use its block when
3554 it is dynamic (useful for very large $message_headers, for example).
3556 6. Lookups now cache *every* query, not just the most recent. A new, separate
3557 store pool is used for this. It can be recovered when all lookup caches are
3558 flushed. Lookups now release memory at the end of their result strings.
3559 This has involved some general refactoring of the lookup sources.
3561 7. Some code has been added to the store_xxx() functions to reduce the amount
3562 of flapping under certain conditions.
3564 8. log_incoming_interface used to affect only the <= reception log lines. Now
3565 it causes the local interface and port to be added to several more SMTP log
3566 lines, for example "SMTP connection from", and rejection lines.
3568 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3570 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3572 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3573 could be overwritten at the end of the current message (or the start of a
3574 new message if it was set in a HELO ACL). The value is now preserved for
3575 the duration of the SMTP connection.
3577 12. If a transport had a headers_rewrite setting, and a matching header line
3578 contained an unqualified address, that address was qualified, even if it
3579 did not match any rewriting rules. The underlying bug was that the values
3580 of the flags that permit the existence of unqualified sender and recipient
3581 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3582 non-local messages, and by -bnq for local messages) were not being
3583 preserved with the message after it was received.
3585 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3586 "next input=" as part of the text comprising the host identity instead of
3587 the correct text. The code was using the same buffer for two different
3588 strings. However, depending on which order the printing function evaluated
3589 its arguments, the bug did not always show up. Under Linux, for example, my
3590 test suite worked just fine.
3592 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3593 doesn't work with some older versions of Perl. It has been changed to "my",
3594 which in any case is probably the better facility to use.
3596 15. A really picky compiler found some instances of statements for creating
3597 error messages that either had too many or two few arguments for the format
3600 16. The size of the buffer for calls to the DNS resolver has been increased
3601 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3602 for addresses that have a lot of PTR records. This alleviates a problem; it
3603 does not fully solve it.
3605 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3606 buffer now truncates the list and logs the incident, which is the same
3607 action as happens when Exim is looking up a host name and its aliases.
3608 Previously in this situation something unpredictable would happen;
3609 sometimes it was "internal error: store_reset failed".
3611 18. If a server dropped the connection unexpectedly when an Exim client was
3612 using GnuTLS and trying to read a response, the client delivery process
3613 crashed while trying to generate an error log message.
3615 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3616 string, for example:
3618 warn message = H1: something\nH2: something
3620 the text was added as a single header line from Exim's point of view
3621 though it ended up OK in the delivered message. However, searching for the
3622 second and subsequent header lines using $h_h2: did not work. This has been
3623 fixed. Similarly, if a system filter added multiple headers in this way,
3624 the routers could not see them.
3626 20. Expanded the error message when iplsearch is called with an invalid key to
3627 suggest using net-iplsearch in a host list.
3629 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3630 ACLs are no longer actually imposed (and a message to that effect is
3633 22. If a "gecos" field in a passwd entry contained escaped characters, in
3634 particular, if it contained a \" sequence, Exim got it wrong when building
3635 a From: or a Sender: header from that name. A second bug also caused
3636 incorrect handling when an unquoted " was present following a character
3637 that needed quoting.
3639 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3640 was not being matched caselessly.
3642 24. Arranged for all hyphens in the exim.8 source to be escaped with
3645 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3646 properly. Recipient callout cache records were still being keyed to include
3647 the sender, even when use_sender was set false. This led to far more
3648 callouts that were necessary. The sender is no longer included in the key
3649 when use_sender is false.
3651 26. Added "control = submission" modifier to ACLs.
3653 27. Added the ${base62d: operator to decode base 62 numbers.
3655 28. dnsdb lookups can now access SRV records.
3657 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3658 the configuration file.
3660 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3661 (-v) mode. This makes the output for a verbose queue run more intelligible.
3663 31. Added a use_postmaster feature to recipient callouts.
3665 32. Added the $body_zerocount variable, containing the number of binary zero
3666 bytes in the message body.
3668 33. The time of last modification of the "new" subdirectory is now used as the
3669 "mailbox time last read" when there is a quota error for a maildir
3672 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3674 35. Added +ignore_unknown as a special item in host lists.
3676 36. Code for decoding IPv6 addresses in host lists is now included, even if
3677 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3678 address was recognized as an IP address, but was then not correctly decoded
3679 into binary, causing unexpected and incorrect effects when compared with
3686 1. Very minor rewording of debugging text in manualroute to say "list of
3687 hosts" instead of "hostlist".
3689 2. If verify=header_syntax was set, and a header line with an unqualified
3690 address (no domain) and a large number of spaces between the end of the
3691 name and the colon was received, the reception process suffered a buffer
3692 overflow, and (when I tested it) crashed. This was caused by some obsolete
3693 code that should have been removed. The fix is to remove it!
3695 3. When running in the test harness, delay a bit after writing a bounce
3696 message to get a bit more predictability in the log output.
3698 4. Added a call to search_tidyup() just before forking a reception process. In
3699 theory, someone could use a lookup in the expansion of smtp_accept_max_
3700 per_host which, without the tidyup, could leave open a database connection.
3702 5. Added the variables $recipient_data and $sender_data which get set from a
3703 lookup success in an ACL "recipients" or "senders" condition, or a router
3704 "senders" option, similar to $domain_data and $local_part_data.
3706 6. Moved the writing of debug_print from before to after the "senders" test
3709 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3710 problems for message scanning, either using a data ACL, or using
3711 local_scan() because the Received: header was not generated till after they
3712 were called (in order to set the time as the time of reception completion).
3713 I have revised the way this works. The header is now generated after the
3714 body is received, but before the ACL or local_scan() are called. After they
3715 are run, the timestamp in the header is updated.
3721 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3722 before starting a queue runner without re-exec. This happened only when
3723 deliver_drop_privilege was set or when the Exim user was set to root. The
3724 effect of the bug was that timeouts during subsequent deliveries caused
3725 crashes instead of being properly handled. The handler is now left at its
3726 default (and expected) setting.
3728 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3729 message, again when deliver_drop_privilege is set or Exim is run as root.
3730 The bug described in (1) was not present in this case, but the tidying up
3731 of the other signals was missing. I have made the two cases consistent.
3733 3. The ignore_target_hosts setting on a manualroute router was being ignored
3734 for hosts that were looked up using the /MX notation.
3736 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3739 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3740 operated on the sender address. After changing the $sender_address to <>
3741 for the sender address verify, Exim was re-instated it as the original
3742 (before rewriting) address, but remembering that it had rewritten it, so it
3743 wasn't rewriting it again. This bug also had the effect of breaking the
3744 sender address verification caching when the sender address was rewritten.
3746 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3747 This has been changed so that if the ip literal address matches
3748 ignore_target_hosts, the router declines.
3750 7. Added expansion conditions match_domain, match_address, and match_local_
3751 part (NOT match_host).
3753 8. The placeholder for the Received: header didn't have a length field set.
3755 9. Added code to Exim itself and to exim_lock to test for a specific race
3756 condition that could lead to file corruption when using MBX delivery. The
3757 issue is with the lockfile that is created in /tmp. If this file is removed
3758 after a process has opened it but before that process has acquired a lock,
3759 there is the potential for a second process to recreate the file and also
3760 acquire a lock. This could lead to two Exim processes writing to the file
3761 at the same time. The added code performs the same test as UW imapd; it
3762 checks after acquiring the lock that its file descriptor still refers to
3763 the same named file.
3765 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3766 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3769 11. Added the smtp_active_hostname option. If used, this will typically be made
3770 to depend on the incoming interface address. Because $interface_address is
3771 not set up until the daemon has forked a reception process, error responses
3772 that can happen earlier (such as "too many connections") no longer contain
3775 12. If an expansion in a condition on a "warn" statement fails because a lookup
3776 defers, the "warn" statement is abandoned, and the next ACL statement is
3777 processed. Previously this caused the whole ACL to be aborted.
3779 13. Added the iplsearch lookup type.
3781 14. Added ident_timeout as a log selector.
3783 15. Added tls_certificate_verified as a log selector.
3785 16. Added a global option tls_require_ciphers (compare the smtp transport
3786 option of the same name). This controls incoming TLS connections.
3788 17. I finally figured out how to make tls_require_ciphers do a similar thing
3789 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3790 before starting the TLS session.
3792 18. Tabs are now shown as \t in -bP output.
3794 19. If the log selector return_path_on_delivery was set, Exim crashed when
3795 bouncing a message because it had too many Received: header lines.
3797 20. If two routers both had headers_remove settings, and the first one included
3798 a superfluous trailing colon, the final name in the first list and the
3799 first name in the second list were incorrectly joined into one item (with a
3800 colon in the middle).
3806 1. Added -C and -D options to the exinext utility, mainly to make it easier
3807 to include in the automated testing, but these could be helpful when
3808 multiple configurations are in use.
3810 2. The exinext utility was not formatting the output nicely when there was
3811 an alternate port involved in the retry record key, nor when there was a
3812 message id as well (for retries that were specific to a specific message
3813 and a specific host). It was also confused by IPv6 addresses, because of
3814 the additional colons they contain. I have fixed the IPv4 problem, and
3815 patched it up to do a reasonable job for IPv6.
3817 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3818 delivery, the log line now contains "pipelined" if PIPELINING was used.
3820 4. An SMTP transport process used to panic and die if the bind() call to set
3821 an explicit outgoing interface failed. This has been changed; it is now
3822 treated in the same way as a connect() failure.
3824 5. A reference to $sender_host_name in the part of a conditional expansion
3825 that was being skipped was still causing a DNS lookup. This no longer
3828 6. The def: expansion condition was not recognizing references to header lines
3829 that used bh_ and bheader_.
3831 7. Added the _cache feature to named lists.
3833 8. The code for checking quota_filecount in the appendfile transport was
3834 allowing one more file than it should have been.
3836 9. For compatibility with Sendmail, the command line option
3844 and sets the incoming protocol and host name (for trusted callers). The
3845 host name and its colon can be omitted when only the protocol is to be set.
3846 Note the Exim already has two private options, -pd and -ps, that refer to
3847 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3848 "s", but I don't think that's a major issue.
3850 10. A number of refactoring changes to the code, none of which should affect
3853 (a) The number of logging options was getting close to filling up the
3854 32-bit word that was used as a bit map. I have split them into two classes:
3855 those that are passed in the argument to log_write(), and those that are
3856 only ever tested independently outside of that function. These are now in
3857 separate 32-bit words, so there is plenty of room for expansion again.
3858 There is no change in the user interface or the logging behaviour.
3860 (b) When building, for example, log lines, the code previously used a
3861 macro that called string_cat() twice, in order to add two strings. This is
3862 not really sufficiently general. Furthermore, there was one instance where
3863 it was actually wrong because one of the argument was used twice, and in
3864 one call a function was used. (As it happened, calling the function twice
3865 did not affect the overall behaviour.) The macro has been replaced by a
3866 function that can join an arbitrary number of extra strings onto a growing
3869 (c) The code for expansion conditions now uses a table and a binary chop
3870 instead of a serial search (which was left over from when there were very
3871 few conditions). Also, it now recognizes conditions like "pam" even when
3872 the relevant support is not compiled in: a suitably worded error message is
3873 given if an attempt is made to use such a condition.
3875 11. Added ${time_interval:xxxxx}.
3877 12. A bug was causing one of the ddress fields not to be passed back correctly
3878 from remote delivery subprocesses. The field in question was not being
3879 subsequently used, so this caused to problems in practice.
3881 13. Added new log selectors queue_time and deliver_time.
3883 14. Might have fixed a bug in maildirsizefile handling that threw up
3884 "unexpected character" debug warnings, and recalculated the data
3885 unnecessarily. In any case, I expanded the warning message to give more
3888 15. Added the message "Restricted characters in address" to the statements in
3889 the default ACL that block characters like @ and % in local parts.
3891 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3892 Three changes have been made:
3894 (a) There was a serious bug; a negative response to MAIL caused the whole
3895 recipient domain to be cached as invalid, thereby blocking all messages
3896 to all local parts at the same domain, from all senders. This bug has
3897 been fixed. The domain is no longer cached after a negative response to
3898 MAIL if the sender used is not empty.
3900 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3903 (c) A new callout option, "use_sender" has been added for people who want
3904 the modified behaviour.
3910 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3913 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3914 indeed breaks things for older releases.
3916 3. Added additional logging to the case where there is a problem reading data
3917 from a filter that is running in a subprocess using a pipe, in order to
3918 try to track down a specific problem.
3920 4. Testing facility fudge: when running in the test harness and attempting
3921 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3922 getting "No route to host". Convert this to a timeout.
3924 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3927 6. Some OS don't have socklen_t but use size_t instead. This affects the
3928 fifth argument of getsockopt() amongst other things. This is now
3929 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3930 can be set for individual OS. I have set it for SunOS5, OSF1, and
3931 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3932 some earlier ones do not.
3934 7. Change 4.30/15 was not doing the test caselessly.
3936 8. The standard form for an IPv6 address literal was being rejected by address
3937 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3938 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3939 this, as well as the form without the "IPv6" on the front (but only when
3940 address literals are enabled, of course).
3942 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3944 10. Exim crashed if a message with an empty sender address specified by -f
3945 encountered a router with an errors_to setting. This could be provoked only
3946 by a command such as
3950 where an empty string was supplied; "<>" did not hit this bug.
3952 11. Installed PCRE release 4.5.
3954 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3955 remained set. It is now erased.
3957 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3958 times from message ids (which are base 36 rather than the normal 62).
3960 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3961 were being counted as actual protocol errors, and logged if the log
3962 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3963 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3964 and DATA following a set of rejected RCPTs do not count as protocol errors.
3965 In other words, Exim assumes they were pipelined, though this may not
3966 actually be the case. Of course, in all cases the client gets an
3967 appropriate error code.
3969 15. If a lookup fails in an ACL condition, a message about the failure may
3970 be available; it is used if testing the ACL cannot continue, because most
3971 such messages specify what the cause of the deferral is. However, some
3972 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3973 that caused an old message to be retained and used if a later statement
3974 caused a defer, replacing the real cause of the deferral.
3976 16. If an IP address had so many PTR records that the DNS lookup buffer
3977 was not large enough to hold them, Exim could crash while trying to process
3978 the truncated data. It now detects and logs this case.
3980 17. Further to 4.21/58, another change has been made: if (and only if) the
3981 first line of a message (the first header line) ends with CRLF, a bare LF
3982 in a subsequent header line has a space inserted after it, so as not to
3983 terminate the header.
3985 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3986 unnecessarily, used atoi() instead of strtol(), and didn't check the
3987 termination when getting file sizes from file names by regex.
3989 19. Completely re-implemented the support for maildirsize files, in the light
3990 of a number of problems with the previous contributed implementation
3991 (4.30/29). In particular:
3993 . If the quota is zero, the maildirsize file is maintained, but no quota is
3996 . If the maildir directory does not exist, it is created before any attempt
3997 to write a maildirsize file.
3999 . The quota value in the file is just a cache; if the quota is changed in
4000 the transport, the new value overrides.
4002 . A regular expression is available for excluding directories from the
4005 20. The autoreply transport checks the characters in options that define the
4006 message's headers; it allows continued headers, but it was checking with
4007 isspace() after an embedded newline instead of explicitly looking for a
4010 21. If all the "regular" hosts to which an address was routed had passed their
4011 expiry times, and had not reached their retry times, the address was
4012 bounced, even if fallback hosts were defined. Now Exim should go on to try
4015 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4016 equivalent code in the SMTP transport. Some hosts send humungous responses
4017 to HELO/EHLO, more than 1024 it seems.
4019 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4020 gives compiler warnings in some environments. I've now done it "properly",
4023 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4024 (because of problems with the built-in one) was declared to return uschar *
4025 instead of char *, causing compiler failure.
4027 25. Fixed a file descriptor leak when processing alias/forward files.
4029 26. Fixed a minor format string issue in dbfn.c.
4031 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4033 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4034 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4037 29. When a sender address is verified, it is cached, to save repeating the test
4038 when there is more than one recipient in a message. However, when the
4039 verification involves a callout, it is possible for different callout
4040 options to be set for different recipients. It is too complicated to keep
4041 track of this in the cache, so now Exim always runs a verification when a
4042 callout is required, relying on the callout cache for the optimization.
4043 The overhead is duplication of the address routing, but this should not be
4046 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4047 to be verified with callout=postmaster, and the main callout worked but the
4048 postmaster check failed, the verification correctly failed. However, if a
4049 subsequent RCPT command asked for sender verification *without* the
4050 postmaster check, incorrect caching caused this verification also to fail,
4053 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4054 it was not caching the DNS options (qualify_single, search_parents) that
4055 were used when the lookup failed. A subsequent lookup with different
4056 options therefore always gave the same answer, though there were cases
4057 where it should not have. (Example: a "domains = !$mx_any" option on a
4058 dnslookup router: the "domains" option is always processed without any
4059 widening, but the router might have qualify_single set.) Now Exim uses the
4060 cached value only when the same options are set.
4062 32. Added John Jetmore's "exipick" utility to the distribution.
4064 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4065 than a timeout (e.g. a certificate is required, and is not provided), an
4066 Exim server now closes the connection immediately. Previously it waited for
4067 the client to close - but if the client is SSL, it seems that they each
4068 wait for each other, leading to a delay before one of them times out.
4070 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4071 maintained 0.8.x compatibility because I don't think many are using it, and
4072 it is clearly obsolete.
4074 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4077 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4078 client certificate was expired. A simple patch fixes this, though I don't
4079 understand the full logic of why the verify callback is called multiple
4082 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4083 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4084 which causes problems with some clients (such as the Certicom SSL Plus
4085 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4086 disables the coutermeasure allowing Eudora to connect."
4088 38. Exim was not checking that a write() to a log file succeeded. This could
4089 lead to Bad Things if a log got too big, in particular if it hit a file
4090 size limit. Exim now panics and dies if it cannot write to a log file, just
4091 as it does if it cannot open a log file.
4093 39. Modified OS/Makefile-Linux so that it now contains
4095 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4097 The two -D definitions ensure that Exim is compiled with large file
4098 support, which makes it possible to handle log files that are bigger than
4101 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4102 instance) a domain was checked against a named list that involved a lookup,
4103 causing $domain_data to be set, then another domain was checked against the
4104 same list, then the first domain was re-checked, the value of $domain_data
4105 after the final check could be wrong. In particular, if the second check
4106 failed, it could be set empty. This bug probably also applied to
4109 41. The strip_trailing_dot option was not being applied to the address given
4110 with the -f command-line option.
4112 42. The code for reading a message's header from the spool was incrementing
4113 $received_count, but never initializing it. This meant that the value was
4114 incorrect (doubled) while delivering a message in the same process in which
4115 it was received. In the most common configuration of Exim, this never
4116 happens - a fresh exec is done - but it can happen when
4117 deliver_drop_privilege is set.
4119 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4120 it now includes up to 150 characters of the unexpected data in the log
4123 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4124 and building data strings. The size of both of these buffers was 10 000
4125 bytes - far larger than anybody would *ever* want, thought I. Needless to
4126 say, somebody hit the limit. I have increased the maximum line length to
4127 20 000 and the maximum data length of concatenated lines to 100 000. I have
4128 also fixed two bugs, because there was no checking on these buffers. Tsk,
4129 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4130 buffer is too small.
4132 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4133 lsearch lookups. Now it does.
4135 46. When parsing a route_list item in a manualroute router, a fixed-length
4136 buffer was used for the list of hosts. I made this 1024 bytes long,
4137 thinking that nobody would ever have a list of hosts that long. Wrong.
4138 Somebody had a whole pile of complicated expansion conditions, and the
4139 string was silently truncated, leading to an expansion error. It turns out
4140 that it is easier to change to an unlimited length (owing to other changes
4141 that have happened since this code was originally written) than to build
4142 structure for giving a limitation error. The length of the item that
4143 expands into the list of hosts is now unlimited.
4145 47. The lsearch lookup could not handle data where the length of text line was
4146 more than 4095 characters. Such lines were truncated, leading to shortened
4147 data being returned. It should now handle lines of any length.
4149 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4150 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4153 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4154 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4155 no longer happens in comment lines. A list of replacements is now placed
4156 at the head of all of the source files, except those whose only change is
4157 to replace PERL_COMMAND in the very first #! line.
4159 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4160 messages on the queue, with a bottom-up merge sort, using code contributed
4161 by Michael Haardt. This should make operations like -bp somewhat faster on
4162 large queues. It won't affect queue runners, except when queue_run_in_order
4165 51. Installed eximstats 1.31 in the distribution.
4167 52. Added support for SRV lookups to the dnslookup router.
4169 53. If an ACL referred to $message_body or $message_body_end, the value was not
4170 reset for any messages that followed in the same SMTP session.
4172 54. The store-handling optimization for building very long strings was not
4173 differentiating between the different store pools. I don't think this
4174 actually made any difference in practice, but I've tidied it.
4176 55. While running the routers to verify a sender address, $sender_address
4177 was still set to the sender address. This is wrong, because when routing to
4178 send a bounce to the sender, it would be empty. Therefore, I have changed
4179 it so that, while verifying a sender address, $sender_address is set to <>.
4180 (There is no change to what happens when verifying a recipient address.)
4182 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4183 target A or AAAA records (if not already returned) without resetting the
4184 qualify_single or search_parents options of the DNS resolver. These are
4185 inappropriate in this case because the targets of MX and SRV records must
4186 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4187 target that, when qualified, matched something in the local domain. These
4188 two options are now turned off when doing these lookups.
4190 57. It seems that at least some releases of Reiserfs (which does not have the
4191 concept of a fixed number of inodes) returns zero and not -1 for the
4192 number of available inodes. This interacted badly with check_spool_inodes,
4193 which assumed that -1 was the "no such thing" setting. What I have done is
4194 to check that the total number of inodes is greater than zero before doing
4195 the test of how many are available.
4197 58. When a "warn" ACL statement has a log_message modifier, the message is
4198 remembered, and not repeated. This is to avoid a lot of repetition when a
4199 message has many recipients that cause the same warning to be written.
4200 Howewer, Exim was preserving the list of already written lines for an
4201 entire SMTP session, which doesn't seem right. The memory is now reset if a
4202 new message is started.
4204 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4205 debugging output unless log_rewrite was also set.
4207 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4208 of (int)(handle) when we know that handle contains (void *)(-1).
4210 61. The Exim daemon panic-logs an error return when it closes the incoming
4211 connection. However "connection reset by peer" seems to be common, and
4212 isn't really an error worthy of noting specially, so that particular error
4215 62. When Exim is trying to find all the local interfaces, it used to panic and
4216 die if the ioctl to get the interface flags failed. However, it seems that
4217 on at least one OS (Solaris 9) it is possible to have an interface that is
4218 included in the list of interfaces, but for which you get a failure error
4219 for this call. This happens when the interface is not "plumbed" into a
4220 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4221 failure of the "get flags" call assumes that the interface is down.
4223 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4224 makes life easier for people who are doing arithmetic on fields extracted
4225 from dates, where you often get leading zeros that should not be
4226 interpreted as octal.
4228 64. Added qualify_domain to the redirect router, to override the global
4231 65. If a pathologically long header line contained very many addresses (the
4232 report of this problem mentioned 10 000) and each of them was rewritten,
4233 Exim could use up a very large amount of memory. (It kept on making new
4234 copies of the header line as it rewrote, and never released the old ones.)
4235 At the expense of a bit more processing, the header rewriting function has
4236 been changed so that it no longer eats memory in this way.
4238 66. The generation of the Received: header has been moved from the time that a
4239 message starts to be received, to the time that it finishes. The timestamp
4240 in the Received: header should now be very close to that of the <= log
4241 line. There are two side-effects of this change:
4243 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4244 logged header lines no longer include the local Received: line, because
4245 it has not yet been created. The same applies to a copy of the message
4246 that is returned to a non-SMTP sender when a message is rejected.
4248 (b) When a filter file is tested using -bf, no additional Received: header
4249 is added to the test message. After some thought, I decided that this
4252 This change does not affect the value of $received_for. It is still set
4253 after address rewriting, but before local_scan() is called.
4255 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4257 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4258 gave an unhelpful panic error message, and a defer error. I have managed to
4259 change this behaviour so that it now rejects any supplied certificate,
4260 which seems right, as the list of acceptable certificates is empty.
4262 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4263 gave an unhelpful defer error. I have not managed to make this reject any
4264 supplied certificates, but the error message it gives is "no certificate
4265 supplied", which is not helpful.
4267 70. exigrep's output now also includes lines that are not associated with any
4268 message, but which match the given pattern. Implemented by a patch from
4269 Martin Sluka, which also tidied up the Perl a bit.
4271 71. Recipient callout verification, like sender verification, was using <> in
4272 the MAIL FROM command. This isn't really the right thing, since the actual
4273 sender may affect whether the remote host accepts the recipient or not. I
4274 have changed it to use the actual sender in the callout; this means that
4275 the cache record is now keyed on a recipient/sender pair, not just the
4276 recipient address. There doesn't seem to be a real danger of callout loops,
4277 since a callout by the remote host to check the sender would use <>.
4278 [SEE ABOVE: changed after hitting problems.]
4280 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4281 temporary errors. However, in the case of such a code being given after
4282 the end of a data transmission (i.e. after ".") Exim was failing to write
4283 a retry record for the message. (Yes, there was some broken host that was
4284 actually sending 8xx at this point.)
4286 73. An unknown lookup type in a host list could cause Exim to panic-die when
4287 the list was checked. (An example that provoked this was putting <; in the
4288 middle of a list instead of at the start.) If this happened during a DATA
4289 ACL check, a -D file could be left lying around. This kind of configuration
4290 error no longer causes Exim to die; instead it causes a defer errror. The
4291 incident is still logged to the main and panic logs.
4293 74. Buglet left over from Exim 3 conversion. The message "too many messages
4294 in one connection" was written to the rejectlog but not the mainlog, except
4295 when address rewriting (yes!) was being logged.
4297 75. Added write_rejectlog option.
4299 76. When a system filter was run not as root (that is, when system_filter_user
4300 was set), the values of the $n variables were not being returned to the
4301 main process; thus, they were not subsequently available in the $sn
4304 77. Added +return_path_on_delivery log selector.
4306 78. A connection timeout was being treated differently from recipients deferred
4307 when testing hosts_max_try with a message that was older than the host's
4308 retry timeout. (The host should not be counted, thus allowing all hosts to
4309 be tried at least once before bouncing.) This may have been the cause of an
4310 occasionally reported bug whereby a message would remain on the queue
4311 longer than the retry timeout, but would be bounced if a delivery was
4312 forced. I say "may" because I never totally pinned down the problem;
4313 setting up timeout/retry tests is difficult. See also the next item.
4315 79. The ultimate address timeout was not being applied to errors that involved
4316 a combination of host plus message (for example, a timeout on a MAIL
4317 command). When an address resolved to a number of possible hosts, and they
4318 were not all tried for each delivery (e.g. because of hosts_max_try), a
4319 message could remain on the queue longer than the retry timeout.
4321 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4324 81. Fixed an obscure SMTP outgoing bug which required at least the following
4325 conditions: (a) there was another message waiting for the same server;
4326 (b) the server returned 5xx to all RCPT commands in the first message so
4327 that the message was not completed; (c) the server dropped the connection
4328 or gave a negative response to the RSET that Exim sends to abort the
4329 transaction. The observed case was a dropped connection after DATA that had
4330 been sent in pipelining mode. That is, the server had advertised PIPELINING
4331 but was not implementing it correctly. The effect of the bug was incorrect
4332 behaviour, such as trying another host, and this could lead to a crash.
4338 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4339 and daemon.c were passed as pointers to ints; they should have been
4340 pointers to socklen_t variables (which are typically unsigned ints).
4342 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4345 3. Fixed a really odd bug that affected only the testing scheme; patching a
4346 certain fixed string in the binary changed the value of another string that
4347 happened to be identical to the end of the original first string.
4349 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4350 name", it returns that address as the IP address. On some operating
4351 systems (e.g. Solaris), it also passes back the IP address string as the
4352 "host name". However, on others (e.g. Linux), it passes back an empty
4353 string. Exim wasn't checking for this, and was changing the host name to an
4354 empty string, assuming it had been canonicized.
4356 5. Although rare, it is permitted to have more than one PTR record for a given
4357 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4358 all the names associated with an address, because they do in Solaris.
4359 However, it seems that they do not in Linux for data that comes from the
4360 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4361 I found this out when I moved to a new Linux workstation and tried to run
4362 the Exim test suite.
4364 To get round this problem I have changed the code so that it now does its
4365 own call to the DNS to look up PTR records when searching for a host name.
4366 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4367 addresses that are only in /etc/hosts are still found.
4369 This behaviour is, however, controlled by an option called host_lookup_
4370 order, which defaults to "bydns:byaddr". If people want to use the other
4371 order, or indeed, just use one or the other means of lookup, they can
4372 specify it in this variable.
4374 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4375 some operating systems, this comes back from gethostbyaddr() as an empty
4376 string, and this is what Exim used to test for. However, it seems that in
4377 other systems, "." is yielded. Exim now tests for this case too.
4379 7. The values of check_spool_space and check_log_space are now held internally
4380 as a number of kilobytes instead of an absolute number of bytes. If a
4381 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4382 kilobyte. This means that much larger values can be stored.
4384 8. Exim monitor: an attempt to get the action menu when not actually pointing
4385 at a message produces an empty menu entitled "No message selected". This
4386 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4387 no entries in it ("Shell widget menu has zero width and/or height"). So I
4388 have added a single, blank menu entry in this case.
4390 9. Added ${quote_local_part.
4392 10. MIME decoding is now applied to the contents of Subject: header lines when
4395 11. Now that a reference to $sender_host_address automatically causes a reverse
4396 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4397 host lookup before query-style lookups in lists that might use this
4398 variable. This has therefore been abolished, and the "net-" prefix is no
4399 longer necessary for query-style lookups.
4401 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4402 have been a typo for LFLAGS, so it has been changed.
4404 13. The install script calls Exim with "-C /dev/null" in order to find the
4405 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4406 to be output. Howeve, since Exim outputs its version number before the
4407 error, it didn't break the script. It just looked ugly. I fixed this by
4408 always allowing "-C /dev/null" if the caller is root.
4410 14. Ignore overlarge ACL variable number when reading spool file - insurance
4411 against a later release with more variables having written the file.
4413 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4414 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4415 this, as well as the form without the "IPv6" on the front.
4417 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4418 OS/Makefile-Darwin file.
4420 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4421 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4423 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4424 variable. However, some OS use ints for inodes. I've added cast to long int
4425 to get rid of the compiler warning.
4427 19. I had forgotten to lock out "/../" in configuration file names when
4428 ALT_CONFIG_PREFIX was set.
4430 20. Routers used for verification do not need to specify transports. However,
4431 if such a router generated a host list, and callout was configured, Exim
4432 crashed, because it could not find a port number from the (non-existent)
4433 transport. It now assumes port 25 in this circumstance.
4435 21. Added the -t option to exigrep.
4437 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4438 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4439 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4440 them is set, LOOKUP_LSEARCH is forced.
4442 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4444 24. Added sender and host information to the "rejected by local_scan()" log
4445 line; previously there was no indication of these.
4447 25. Added .include_if_exists.
4449 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4450 Linux. It turns out that not all file systems support this. Apparently some
4451 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4452 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4453 not supported on the file descriptor, is now ignored when Exim is trying to
4454 sync a directory. This applies only to Linux.
4456 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4458 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4459 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4462 29. Added support for maildirsize files from supplied patch (modified a bit).
4464 30. The use of :fail: followed by an empty string could lead Exim to respond to
4465 sender verification failures with (e.g.):
4467 550 Verification failed for <xxx>
4468 550 Sender verify failed
4470 where the first response line was missing the '-' that indicates it is not
4471 the final line of the response.
4473 31. The loop for finding the name of the user that called Exim had a hardwired
4474 limit of 10; it now uses the value of finduser_retries, which is used for
4475 all other user lookups.
4477 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4480 33. Exim was neglecting to zero errno before one call of strtol() when
4481 expanding a string and expecting an integer value. On some systems this
4482 resulted in spurious "integer overflow" errors. Also, it was casting the
4483 result into an int without checking.
4485 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4486 did not work. The code looks as if it has *never* worked, though it appears
4487 to have been documented since at least releast 1.62. I have made it work.
4489 35. The "timeout_DNS" error in retry rules, also documented since at least
4490 1.62, also never worked. As it isn't clear exactly what this means, and
4491 clearly it isn't a major issue, I have abolished the feature by treating it
4492 as "timeout", and writing a warning to the main and panic logs.
4494 36. The display of retry rules for -brt wasn't always showing the error code
4497 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4498 timeout_connect_A, timeout_connect_MX.
4500 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4501 to the empty sender.
4503 39. The daemon was not analysing the content of -oX till after it had closed
4504 stderr and disconnected from the controlling terminal. This meant that any
4505 syntax errors were only noted on the panic log, and the return code from
4506 the command was 0. By re-arranging the code a little, I've made the
4507 decoding happen first, so such errors now appear on stderr, and the return
4508 code is 1. However, the actual setting up of the sockets still happens in
4509 the disconnected process, so errors there are still only recorded on the
4512 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4513 connections (as happens on some IP stacks) was logged at start up time as
4514 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4515 it from "IPv6 and IPv4", which means that two separate sockets are being
4518 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4519 says whether AF_INET or AF_INET6 was passed as an argument.
4521 42. Exiwhat output was messed up when time zones were included in log
4524 43. Exiwhat now gives more information about the daemon's listening ports,
4525 and whether -tls-on-connect was used.
4527 44. The "port" option of the smtp transport is now expanded.
4529 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4530 silently ignored. Now an error message is written to the main and panic
4533 46. There's a new ACL modifier called "logwrite" which writes to a log file
4534 as soon as it is encountered.
4536 47. Added $local_user_uid and $local_user_gid at routing time.
4538 48. Exim crashed when trying to verify a sender address that was being
4541 49. Exim was recognizing only a space character after ".include". It now also
4542 recognizes a tab character.
4544 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4545 extracting the relevant information from the specification. The man page no
4546 longer contains scrambled data for the -d option, and I've added a section
4547 at the front about calling Exim under different names.
4549 51. Added "extra_headers" argument to the "mail" command in filter files.
4551 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4554 53. Installed eximstats 1.29.
4556 54. Added transport_filter_timeout as a generic transport option.
4558 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4559 Cc: header lines. This was required by RFC 822, but it not required by RFC
4562 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4563 incoming messages that did not have them. Now it does so only if the
4564 message originates locally, that is, if there is no associated remote host
4565 address. When Resent- header lines are present, this applies to the Resent-
4566 lines rather than the non-Resent- lines.
4568 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4569 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4571 58. Messages for configuration errors now include the name of the main
4572 configuration file - useful now that there may be more than one file in a
4573 list (.included file names were always shown).
4575 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4576 for those rare installations that do not start the daemon as root or run it
4577 setuid root. I've cut out the call to initgroups() if the daemon is not
4580 60. The Exim user and group can now be bound into the binary as text strings
4581 that are looked up at the start of Exim's processing.
4583 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4585 62. Added $mailstore_basename variable.
4587 63. Installed patch to sieve.c from Michael Haardt.
4589 64. When Exim failed to open the panic log after failing to open the main log,
4590 the original message it was trying to log was written to stderr and debug
4591 output, but if they were not available (the usual case in production), it
4592 was lost. Now it is written to syslog before the two lines that record the
4593 failures to open the logs.
4595 65. Users' Exim filters run in subprocesses under the user's uid. It is
4596 possible for a "deliver" command or an alias in a "personal" command to
4597 provoke an address rewrite. If logging of address rewriting is configured,
4598 this fails because the process is not running as root or exim. There may be
4599 a better way of dealing with this, but for the moment (because 4.30 needs
4600 to be released), I have disabled address rewrite logging when running a
4601 filter in a non-root, non-exim process.
4607 1. The buildconfig auxiliary program wasn't quoting the value set for
4608 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4609 not defined. This bug was masked in 4.22 by the effect that was fixed in
4612 2. Some messages that were rejected after a message id was allocated were
4613 shown as "incomplete" by exigrep. It no longer does this for messages that
4614 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4616 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4617 have allow_domain_literals set, the ID did not get logged in the <= line.
4618 Domain literals are now always recognized in Message-ID: header lines.
4620 4. The first argument for a ${extract expansion item is the key name or field
4621 number. Leading and trailing spaces in this item were not being ignored,
4622 causing some misleading effects.
4624 5. When deliver_drop_privilege was set, single queue runner processes started
4625 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4626 same command in the process it spins off) were not dropping privilege.
4628 6. When the daemon running as "exim" started a queue runner, it always
4629 re-executed Exim in the spun-off process. This is a waste of effort when
4630 deliver_drop_privilege is set. The new process now just calls the
4631 queue-runner function directly.
4637 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4640 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4641 ignored. Though the use of -oP was forcing the writing of a pid file, it
4642 was always written to the default place.
4644 3. If the message "no IP address found for host xxxx" is generated during
4645 incoming verification, it is now followed by identification of the incoming
4646 connection (so you can more easily find what provoked it).
4648 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4650 5. Added some features to "harden" Exim a bit more against certain attacks:
4652 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4653 be put in Local/Makefile. This is like the never_users runtime option,
4654 but it cannot be overridden. The default setting is "root".
4656 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4657 prefix string with which any file named in a -C command line option
4660 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4661 is retained for -C and -D only if the caller of Exim is root. Without
4662 it, the exim user may also use -C and -D and retain privilege.
4664 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4665 command line option is disabled.
4667 6. Macro names set by the -D option must start with an upper case letter, just
4668 like macro names defined in the configuration file.
4670 7. Added "dereference=" facility to LDAP.
4672 8. Two instances of the typo "uknown" in the source files are fixed.
4674 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4675 the Configure-Makefile script screwed up while processing it.
4677 10. Incorporated PCRE 4.4.
4679 11. The SMTP synchronization check was not operating right at the start of an
4680 SMTP session. For example, it could not catch a HELO sent before the client
4681 waited for the greeting. There is now a check for outstanding input at the
4682 point when the greeting is written. Because of the duplex, asynchronous
4683 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4684 way, but not yet received, when the check is performed.
4686 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4687 on TCP/IP sockets, because this apparently causes some broken clients to
4690 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4691 unchanged) from the Cygwin maintainer.
4693 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4694 was supported instead of testing for mbx. Effectively a typo.
4696 15. The spa authenticator server code was not checking that the input it
4697 received was valid base64.
4699 16. The debug output line for the "set" modifier in ACLs was not showing the
4700 name of the variable that was being set.
4702 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4704 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4705 lookup on its own. Something else was needed to trigger the lookup. For
4706 example, a match in host_lookup or the need for a host name in a host list.
4707 Now, if $sender_host_name is referenced and the host name has not yet been
4708 looked up, a lookup is performed. If the lookup fails, the variable remains
4709 empty, and $host_lookup_failed is set to "1".
4711 19. Added "eqi" as a case-independent comparison operator.
4713 20. The saslauthd authentication condition could segfault if neither service
4714 nor realm was specified.
4716 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4717 error message that was logged was misleading, and incoming SMTP
4718 connections were dropped. The message is now more accurate, and temporary
4719 errors are given to SMTP connections.
4721 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4722 (except RSET) with a 421 error, until QUIT is received. However, it was
4723 failing to send a response to QUIT.
4725 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4726 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4727 ACL is now run after the helo_try_verify_hosts code.
4729 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4730 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4731 case-independent, so other case variants are also recognized). Apparently
4732 some systems use these upper case variants.
4734 25. If more than two messages were waiting for the same host, and a transport
4735 filter was specified for the transport, Exim sent two messages over the
4736 same TCP/IP connection, and then failed with "socket operation on non-
4737 socket" when it tried to send the third.
4739 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4741 27. The extern definition of crypt16() in expand.c was not being excluded when
4742 the OS had its own crypt16() function.
4744 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4745 as a preferred synonym for return_size_limit, both as an option and as an
4748 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4750 30. Changed the default configuration ACL to relax the local part checking rule
4751 for addresses that are not in any local domains. For these addresses,
4752 slashes and pipe symbols are allowed within local parts, but the sequence
4753 /../ is explicitly forbidden.
4755 31. SPA server authentication was not clearing the challenge buffer before
4758 32. log_message in a "warn" ACL statement was writing to the reject log as
4759 well as to the main log, which contradicts the documentation and doesn't
4760 seem right (because no rejection is happening). So I have stopped it.
4762 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4763 However, I am unable to do any testing of this.
4765 34. Fixed an infelicity in the appendfile transport. When checking directories
4766 for a mailbox, to see if any needed to be created, it was accidentally
4767 using path names with one or more superfluous leading slashes; tracing
4768 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4770 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4771 contents are added to the log line that is written for every discarded
4772 recipient. (Previously a log_message setting was ignored.)
4774 36. The ${quote: operator now quotes the string if it is empty.
4776 37. The install script runs exim in order to find its version number. If for
4777 some reason other than non-existence or emptiness, which it checks, it
4778 could not run './exim', it was installing it with an empty version number,
4779 i.e. as "exim-". This error state is now caught, and the installation is
4782 38. An argument was missing from the function that creates an error message
4783 when Exim fails to connect to the socket for saslauthd authentication.
4784 This could cause Exim to crash, or give a corrupted message.
4786 39. Added isip, isip4, and isip6 to ${if conditions.
4788 40. The ACL variables $acl_xx are now saved with the message, and can be
4789 accessed later in routers, transports, and filters.
4791 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4792 strings in the file are not string-expanded.
4794 42. If a MAIL command specified a SIZE value that was too large to fit into an
4795 int variable, the check against message_size_limit failed. Such values are
4796 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4797 day this will have to be increased, but I don't think I want to be around
4798 when emails are that large.
4805 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4806 iconv() is not standard in FreeBSD.
4808 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4809 IPv6 enabled. The observed symptom was a segmentation fault on return from
4810 the function os_common_find_running_interfaces() in src/os.c.
4812 3. In the check_special_case() function in daemon.c I had used "errno" as an
4813 argument name, which causes warnings on some systems. This was basically a
4814 typo, since it was named "eno" in the comments!
4816 4. The code that waits for the clock to tick (at a resolution of some fraction
4817 of a second) so as to ensure message-id uniqueness was always waiting for
4818 at least one whole tick, when it could have waited for less. [This is
4819 almost certainly not relevant at current processor speeds, where it is
4820 unlikely to ever wait at all. But we try to future-proof.]
4822 5. The function that sleeps for a time interval that includes fractions of a
4823 second contained a race. It did not block SIGALRM between setting the
4824 timer, and suspending (a couple of lines later). If the interval was short
4825 and the sigsuspend() was delayed until after it had expired, the suspension
4826 never ended. On busy systems this could lead to processes getting stuck for
4829 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4830 process, before it forks any delivery processes. The open lookup caching
4831 mechanism meant that the open file or database connection was passed into
4832 the delivery process. The problem was that delivery processes always tidy
4833 up cached lookup data. This could cause a problem for the next delivery
4834 process started by the queue runner, because the external queue runner
4835 process does not know about the closure. So the next delivery process
4836 still has data in the lookup cache. In the case of a file lookup, there was
4837 no problem because closing a file descriptor in a subprocess doesn't affect
4838 the parent. However, if the lookup was caching a connection to a database,
4839 the connection was closed, and the second delivery process was likely to
4840 see errors such as "PGSQL: query failed: server closed the connection
4841 unexpectedly". The problem has been fixed by closing all cached lookups
4842 in a queue runner before running a delivery process.
4844 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4845 seem to have the "const" qualifier which it has on other OS. I've
4848 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4849 *of the same type* (client or server) with the same public name that an
4850 error should be diagnosed.
4852 9. When Exim looked up a host name for an IP address, but failed to find the
4853 original IP address when looking up the host name (a safety check), it
4854 output the message "<ip address> does not match any IP for NULL", which was
4855 confusing, to say the least. The bug was that the host name should have
4856 appeared instead of "NULL".
4858 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4859 user that is built into the binary, and the -C or -D options is used, root
4860 privilege is dropped before the configuration file is read. In addition,
4861 logging is switched to stderr instead of the normal log files. If the
4862 configuration then re-defines the Exim user, the unprivileged environment
4863 is probably not what is expected, so Exim logs a panic warning message (but
4866 However, if deliver_drop_privilege is set, the unprivileged state may well
4867 be exactly what is intended, so the warning has been cut out in that case,
4868 and Exim is allowed to try to write to its normal log files.
4874 1. smtp_return_error_details was not giving details for temporary sender
4875 or receiver verification errors.
4877 2. Diagnose a configuration error if two authenticators have the same public
4880 3. Exim used not to create the message log file for a message until the first
4881 delivery attempt. This could be confusing when incoming messages were held
4882 for policy or load reasons. The message log file is now created at the time
4883 the message is received, and an initial "Received" line is written to it.
4885 4. The automatically generated man page for command line options had a minor
4886 bug that caused no ill effects; however, a more serious problem was that
4887 the procedure for building the man page automatically didn't always
4888 operate. Consequently, release 4.20 contains an out-of-date version. This
4889 shouldn't happen again.
4891 5. When building Exim with embedded Perl support, the script that builds the
4892 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4893 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4895 6. The freeze_tell option was not being used for messages that were frozen on
4896 arrival, either by an ACL or by local_scan().
4898 7. Added the smtp_incomplete_transaction log selector.
4900 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4901 was accepting AUTH without a new EHLO.
4903 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4904 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4907 10. Logging of TCP/IP connections (when configured) now happens in the main
4908 daemon process instead of the child process, so that the TCP/IP connection
4909 count is more accurate (but it can never be perfect).
4911 11. The use of "drop" in a nested ACL was not being handled correctly in the
4912 outer ACL. Now, if condition failure induced by the nested "drop" causes
4913 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4914 or "require"), the connection is dropped.
4916 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4917 that yield "discard" can now be used with an "accept" or a "discard" verb,
4918 but an error is generated for any others (because I can't see a useful way
4919 to define what should happen).
4921 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4922 are now processed in the same way as lines in the Exim configuration file.
4923 In particular, continuation lines are supported.
4925 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4927 15. Added -ti meaning -t -i.
4929 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4930 domains, and warn by logging if others are found.
4932 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4933 returned from GIFCONF when Exim is trying to find the list of interfaces on
4934 a host. The code in os.c has been modified to copy each ifreq to an aligned
4935 structure in all cases.
4937 Also, in some cases, the returned ifreq's were being copied to a 'struct
4938 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4939 means the last couple of bytes of an IPv6 address could be chopped if the
4940 ifreq contained only a normal sockaddr (14 bytes storage).
4942 18. Named domain lists were not supported in the hosts_treat_as_local option.
4943 An entry such as +xxxx was not recognized, and was treated as a literal
4946 19. Ensure that header lines added by a DATA ACL are included in the reject log
4947 if the ACL subsequently rejects the message.
4949 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4950 MD5 (which is deprecated).
4952 21. When testing a filter file using -bf, Exim was writing a message when it
4953 took the sender from a "From " line in the message, but it was not doing so
4954 when it took $return_path from a Return-Path: header line. It now does.
4956 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4957 with a valid header line field name (a series of printing characters
4958 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4960 23. Changed "disc" in the source to "disk" to conform to the documentation and
4961 the book and for uniformity.
4963 24. Ignore Sendmail's -Ooption=value command line item.
4965 25. When execve() failed while trying to run a command in a pipe transport,
4966 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4967 could be confused with a return value of 69 from the command itself. This
4968 has been changed to 127, the value the shell returns if it is asked to run
4969 a non-existent command. The wording for the related log line suggests a
4970 non-existent command as the problem.
4972 26. If received_header_text expands to an empty string, do not add a Received:
4973 header line to the message. (Well, it adds a token one on the spool, but
4974 marks it "old" so that it doesn't get used or transmitted.)
4976 27. Installed eximstats 1.28 (addition of -nt option).
4978 28. There was no check for failure on the call to getsockname() in the daemon
4979 code. This can fail if there is a shortage of resources on the system, with
4980 ENOMEM, for example. A temporary error is now given on failure.
4982 29. Contrary to the C standard, it seems that in some environments, the
4983 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4984 program. Exim now does this explicitly; it affects the formatting of
4985 timestamps using strftime().
4987 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4988 complaints. I've now initialized all the variables, to avoid this.
4990 32. Header lines added by a system filter were not being "seen" during
4991 transport-time rewrites.
4993 33. The info_callback() function passed to OpenSSL is set up with type void
4994 (*)(SSL *, int, int), as described somewhere. However, when calling the
4995 function (actually a macro) that sets it up, the type void(*)() is
4996 expected. I've put in a cast to prevent warnings from picky compilers.
4998 34. If a DNS black list lookup found a CNAME record, but there were no A
4999 records associated with the domain it pointed at, Exim crashed.
5001 35. If a DNS black list lookup returned more than one A record, Exim ignored
5002 all but the first. It now scans all returned addresses if a particular IP
5003 value is being sought. In this situation, the contents of the
5004 $dnslist_value variable are a list of all the addresses, separated by a
5007 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5008 to accept a host name and all its aliases if the forward lookup for any of
5009 them yielded the IP address of the incoming connection. Now it accepts only
5010 those names whose forward lookup yields the correct IP address. Any other
5011 names are discarded. This closes a loophole whereby a rogue DNS
5012 administrator could create reverse DNS records to break through a
5013 wildcarded host restriction in an ACL.
5015 37. If a user filter or a system filter that ran in a subprocess used any of
5016 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5017 the wrong values were passed to the pipe command ($thisaddress had the
5018 value of $0, $0 had the value of $1, etc). This bug was introduced by
5019 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5021 38. Improved the line breaking for long SMTP error messages from ACLs.
5022 Previously, if there was no break point between 40 and 75 characters, Exim
5023 left the rest of the message alone. Two changes have been made: (a) I've
5024 reduced the minimum length to 35 characters; (b) if it can't find a break
5025 point between 35 and 75 characters, it looks ahead and uses the first one
5026 that it finds. This may give the occasional overlong line, but at least the
5027 remaining text gets split now.
5029 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5030 file descriptors might be low, and that setting 1000 would always raise it.
5031 It turns out that in some environments, the limit is already over 1000 and
5032 that lowering it causes trouble. So now Exim takes care not to decrease it.
5034 40. When delivering a message, the value of $return_path is set to $sender_
5035 address at the start of routing (routers may change the value). By an
5036 oversight, this default was not being set up when an address was tested by
5037 -bt or -bv, which affected the outcome if any router or filter referred to
5040 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5041 log only when "message" or "log_message" are set. However, if one of the
5042 conditions was an address verification, or a call to a nested ACL, the
5043 messages generated by the underlying test were being passed through. This
5044 no longer happens. The underlying message is available in $acl_verify_
5045 message for both "message" and "log_message" expansions, so it can be
5046 passed through if needed.
5048 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5049 new expansion $bh_ to give the encoded byte string without charset
5050 translation. Translation happens only if iconv() is available; HAVE_ICONV
5051 indicates this at build time. HEADERS_CHARSET gives the charset to
5052 translate to; headers_charset can change it in the configuration, and
5053 "headers charset" can change it in an individual filter file.
5055 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5056 that creates RFC 2047 encoded "words" labels them as that charset instead
5057 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5058 expansion operator; (ii) when Exim creates a From: line for a local
5059 message; (iii) when a header line is rewritten to include a "phrase" part.
5061 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5062 buggy, causing it to skip the first lines of messages whose message ID
5063 ended in 'D'. This would not have bitten before Exim release 4.14, because
5064 message IDs were unlikely to end in 'D' before then. The effect was to have
5065 incorrect size information for certain domains.
5067 45. #include "config.h" was missing at the start of the crypt16.c module. This
5068 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5071 46. If there was a timeout during a "random" callout check, Exim treated it as
5072 a failure of the random address, and carried on sending RSET and the real
5073 address. If the delay was just some slowness somewhere, the response to the
5074 original RCPT would be taken as a response to RSET and so on, causing
5075 mayhem of various kinds.
5077 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5078 when I implemented it. It didn't allow for the fact that some option values
5079 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5080 the right test for positive values.
5082 48. Domain names in DNS records are case-independent. Exim always looks them up
5083 in lower case. Some resolvers return domain names in exactly the case they
5084 appear in the zone file, that is, they may contain uppercase letters. Not
5085 all resolvers do this - some return always lower case. Exim was treating a
5086 change of case by a resolver as a change of domain, similar to a widening
5087 of a domain abbreviation. This triggered its re-routing code and so it was
5088 trying to route what was effectively the same domain again. This normally
5089 caused routing to fail (because the router wouldn't handle the domain
5090 twice). Now Exim checks for this case specially, and just changes the
5091 casing of the domain that it ultimately uses when it transmits the message
5094 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5097 50. If a filter generated a file delivery with a non-absolute name (possible if
5098 no home directory exists for the router), the forbid_file option was not
5101 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5102 the existing equality matching.
5104 52. Exim was using ints instead of ino_t variables in some places where it was
5105 dealing with inode numbers.
5107 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5108 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5109 called TMPDIR, and if it finds it is different, it changes its value.
5111 54. The smtp_printf() function is now made available to local_scan() so
5112 additional output lines can be written before returning. There is also an
5113 smtp_fflush() function to enable the detection of a dropped connection.
5114 The variables smtp_input and smtp_batched_input are exported to
5117 55. Changed the default runtime configuration: the message "Unknown user"
5118 has been removed from the ACL, and instead placed on the localuser router,
5119 using the cannot_route_message feature. This means that any verification
5120 failures that generate their own messages won't get overridden. Similarly,
5121 the "Unrouteable address" message that was in the ACL for unverifiable
5122 relay addresses has also been removed.
5124 56. Added hosts_avoid_esmtp to the smtp transport.
5126 57. The exicyclog script was not checking for the esoteric option
5127 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5128 will work only if exicyclog is run under the appropriate euid.
5130 58. Following a discussion on the list, the rules by which Exim recognises line
5131 endings on incoming messages have been changed. The -dropcr and drop_cr
5132 options are now no-ops, retained only for backwards compatibility. The
5133 following line terminators are recognized: LF CRLF CR. However, special
5134 processing applies to CR:
5136 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5137 nor a local message in the state where . is a terminator.
5139 (ii) If a bare CR is encountered in a header line, an extra space is added
5140 after the line terminator so as not to end the header. The reasoning
5141 behind this is that bare CRs in header lines are most likely either
5142 to be mistakes, or people trying to play silly games.
5144 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5145 was being incorrectly given as 18 bytes larger than it should have been.
5146 This is a VOB (very old bug).
5148 60. This may never have affected anything current, but just in case it has:
5149 When the local host is found other than at the start of a list of hosts,
5150 the local host, those with the same MX, and any that follow, are discarded.
5151 When the list in question was part of a longer list of hosts, the following
5152 hosts (not currently being processed) were also being discarded. This no
5153 longer happens. I'm not sure if this situation could ever has previously
5156 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5159 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5160 header. This is something that is recommended in a new Internet Draft, and
5161 is something that is documented as being done by Sendmail. There are two
5162 possible values. For messages generated by the autoreply transport, Exim
5165 Auto-Submitted: auto-replied
5167 whereas for all other generated messages (e.g. bounces) it adds
5169 Auto-Submitted: auto-generated
5171 63. The "personal" condition in filters now includes a test for the
5172 Auto-Submitted: header. If it contains the string "auto-" the message it
5173 not considered personal.
5175 64. Added rcpt_include_affixes as a generic transport option.
5177 65. Added queue_only_override (default true).
5179 66. Added the syslog_duplication option.
5181 67. If what should have been the first header line of a message consisted of
5182 a space followed by a colon, Exim was mis-interpreting it as a header line.
5183 It isn't of course - it is syntactically invalid and should therefore be
5184 treated as the start of the message body. The misbehaviour could have
5185 caused a number of strange effects, including loss of data in subsequent
5186 header lines, and spool format errors.
5188 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5189 client host had authenticated. This control can now be exercised by an ACL
5190 for more flexibility.
5192 69. By default, callouts do not happen when testing with -bh. There is now a
5193 variant, -bhc, which does actually run the callout code, including
5194 consulting and updating the callout cache.
5196 70. Added support for saslauthd authentication, courtesy of Alexander
5199 71. If statvfs() failed on the spool or log directories while checking their
5200 size for availability, Exim confusingly gave the error "space shortage".
5201 Furthermore, in debugging mode it crashed with a floating point exception.
5202 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5203 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5204 serious problem, Exim now writes to the main and panic logs when this
5205 happens, with details of the failure. It then refuses to accept the
5206 incoming message, giving the message "spool directory problem" or "log
5207 directory problem" with a 421 code for SMTP messages.
5209 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5210 0, 1, and 2 exist, because some OS complain for execs without them (see
5211 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5212 descriptors. However, the code omitted to check that the open succeeded,
5213 causing mysterious errors if for some reason the permissions on /dev/null
5214 got screwed. Now Exim writes a message to the main and panic logs, and
5215 bombs out if it can't open /dev/null.
5217 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5218 interact so that it is all more flexible. It is supposed to remain
5219 backwards compatible. Also added extra_local_interfaces.
5221 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5222 to bomb out with an assertion failure - to the client this appears as a
5223 connection drop. This problem occurs in the part of the code that was taken
5224 from the Samba project. Fortunately, the assertion is in a very simple
5225 function, so I have fixed this by reproducing the function inline in the
5226 one place where it is called, and arranging for authentication to fail
5227 instead of killing the process with assert().
5229 75. The SPA client code was not working when the server requested OEM rather
5230 than Unicode encoding.
5232 76. Added code to make require_files with a specific uid setting more usable in
5233 the case where statting the file as root fails - usually a non-root-mounted
5234 NFS file system. When this happens and the failure is EACCES, Exim now
5235 forks a subprocess and does the per-uid checking as the relevant uid.
5237 77. Added process_log_path.
5239 78. If log_file_path was not explicitly set, a setting of check_log_space or
5240 check_log_inodes was ignored.
5242 79. If a space check for the spool or log partitions fails, the incident is now
5243 logged. Of course, in the latter case the data may get lost...
5245 80. Added the %p formatting code to string_format() so that it can be used to
5246 print addresses in debug_print(). Adjusted all the address printing in the
5247 debugging in store.c to use %p rather than %d.
5249 81. There was a concern that a line of code in smtp_in.c could overflow a
5250 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5251 initially expressed, the concern was not well-founded, because trailing
5252 spaces are removed early. However, if the trailing spaces were followed by
5253 a NULL, they did not get removed, so the overflow was possible. Two fixes
5256 (a) I re-wrote the offending code in a cleaner fashion.
5257 (b) If an incoming SMTP command contains a NULL character, it is rejected
5260 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5261 runs initgroups(), so that if the Exim user is in any additional groups,
5262 they will be used during message reception.
5268 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob