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
33 TF/01 Failure to get a lock on a hints database can have serious
34 consequences so log it to the panic log.
36 TF/02 Log LMTP confirmation messages in the same way as SMTP,
37 controlled using the smtp_confirmation log selector.
39 TF/03 Include the error message when we fail to unlink a spool file.
41 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
42 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
43 for maintaining out-of-tree patches for some time.
45 PP/01 Bugzilla 139: Documentation and portability issues.
46 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
47 Handle per-OS dynamic-module compilation flags.
49 PP/02 Let /dev/null have normal permissions.
50 The 4.73 fixes were a little too stringent and complained about the
51 permissions on /dev/null. Exempt it from some checks.
52 Reported by Andreas M. Kirchwitz.
54 PP/03 Report version information for many libraries, including
55 Exim version information for dynamically loaded libraries. Created
56 version.h, now support a version extension string for distributors
57 who patch heavily. Dynamic module ABI change.
59 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
60 privilege escalation vulnerability whereby the Exim run-time user
61 can cause root to append content of the attacker's choosing to
64 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
67 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
68 If dropping privileges for untrusted macros, we disabled normal logging
69 on the basis that it would fail; for the Exim run-time user, this is not
70 the case, and it resulted in successful deliveries going unlogged.
71 Fixed. Reported by Andreas Metzler.
77 PP/01 Date: & Message-Id: revert to normally being appended to a message,
78 only prepend for the Resent-* case. Fixes regression introduced in
79 Exim 4.70 by NM/22 for Bugzilla 607.
81 PP/02 Include check_rfc2047_length in configure.default because we're seeing
82 increasing numbers of administrators be bitten by this.
84 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
86 PP/03 Bugzilla 994: added openssl_options main configuration option.
88 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
90 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
92 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
94 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
95 without a peer certificate, leading to a segfault because of an
96 assumption that peers always have certificates. Be a little more
97 paranoid. Problem reported by Martin Tscholak.
99 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
100 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
101 NB: ClamAV planning to remove STREAM in "middle of 2010".
102 CL also introduces -bmalware, various -d+acl logging additions and
103 more caution in buffer sizes.
105 PP/09 Implemented reverse_ip expansion operator.
107 PP/10 Bugzilla 937: provide a "debug" ACL control.
109 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
111 PP/12 Bugzilla 973: Implement --version.
113 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
115 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
117 PP/15 Bugzilla 816: support multiple condition rules on Routers.
119 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
120 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
121 ignore trailing whitespace.
123 JJ/02 prevent non-panic DKIM error from being sent to paniclog
125 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
128 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
129 Notification from Dr Andrew Aitchison.
131 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
132 ExtendedDetectionInfo response format.
133 Notification from John Horne.
135 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
138 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
139 XSL and documented dependency on system catalogs, with examples of how
142 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
145 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
146 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
147 configuration file which is writeable by the Exim user or group.
149 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
150 of configuration files to cover files specified with the -C option if
151 they are going to be used with root privileges, not just the default
154 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
155 option (effectively making it always true).
157 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
158 files to be used while preserving root privileges.
160 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
161 that rogue child processes cannot use them.
163 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
164 run-time user, instead of root.
166 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
167 Exim run-time user without dropping privileges.
169 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
170 result string, instead of calling string_vformat() twice with the same
173 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
174 for other users. Others should always drop root privileges if they use
175 -C on the command line, even for a whitelisted configure file.
177 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
179 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
185 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
186 $data_path, and $header_path variables; fixed documentation bugs and
189 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
190 exipick to access non-standard spools, including the "frozen" queue
193 NM/01 Bugzilla 965: Support mysql stored procedures.
194 Patch from Alain Williams
196 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
198 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
199 Patch from Andreas Metzler
201 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
202 Patch from Kirill Miazine
204 NM/05 Bugzilla 671: Added umask to procmail example.
206 JJ/03 installed exipick 20100323.0, fixing doc bug
208 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
209 directory. Notification and patch from Dan Rosenberg.
211 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
213 TK/02 Improve log output when DKIM signing operation fails.
215 MH/01 Treat the transport option dkim_domain as a colon separated
216 list, not as a single string, and sign the message with each element,
217 omitting multiple occurences of the same signer.
219 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
220 Bugzilla 985, 986. Patch by Simon Arlott
222 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
223 Patch by Simon Arlott
225 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
226 MBX locking. Notification from Dan Rosenberg.
232 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
234 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
236 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
238 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
240 NM/04 Bugzilla 915: Flag broken perl installation during build.
246 TK/01 Added patch by Johannes Berg that expands the main option
247 "spamd_address" if it starts with a dollar sign.
249 TK/02 Write list of recipients to X-Envelope-Sender header when building
250 the mbox-format spool file for content scanning (suggested by Jakob
253 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
254 (http://www.dcc-servers.net/) support via dccifd. Activated by
255 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
257 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
258 by Mark Daniel Reidel <mr@df.eu>.
260 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
261 When building exim an external PCRE library is now needed -
262 PCRE is a system library on the majority of modern systems.
263 See entry on PCRE_LIBS in EDITME file.
265 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
266 conversation. Added nologin parameter to request.
267 Patch contributed by Kirill Miazine.
269 TF/01 Do not log submission mode rewrites if they do not change the address.
271 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
273 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
274 log files in place. Contributed by Roberto Lima.
276 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
278 TF/03 Bugzilla 615: When checking the local_parts router precondition
279 after a local_part_suffix or local_part_prefix option, Exim now
280 does not use the address's named list lookup cache, since this
281 contains cached lookups for the whole local part.
283 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
284 Robert Millan. Documentation is in experimental-spec.txt.
286 TF/04 Bugzilla 668: Fix parallel build (make -j).
288 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
290 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
291 Patch provided by Jan Srzednicki.
293 TF/05 Leading white space used to be stripped from $spam_report which
294 wrecked the formatting. Now it is preserved.
296 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
297 that they are available at delivery time.
299 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
301 TF/08 TLS error reporting now respects the incoming_interface and
302 incoming_port log selectors.
304 TF/09 Produce a more useful error message if an SMTP transport's hosts
305 setting expands to an empty string.
307 NM/06 Bugzilla 744: EXPN did not work under TLS.
308 Patch provided by Phil Pennock.
310 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
311 Patch provided by Richard Godbee.
313 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
314 acl_smtp_notquit, added index entry.
316 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
317 Patch provided by Eugene Bujak.
319 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
320 accept(). Patch provided by Maxim Dounin.
322 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
323 Patch provided by Phil Pennock.
325 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
327 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
328 Patch provided by Brad "anomie" Jorsch.
330 NM/14 Bugzilla 622: Added timeout setting to transport filter.
331 Patch provided by Dean Brooks.
333 TK/05 Add native DKIM support (does not depend on external libraries).
335 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
336 Patch provided by Graeme Fowler.
338 NM/16 Bugzilla 851: Documentation example syntax fix.
340 NM/17 Changed NOTICE file to remove references to embedded PCRE.
342 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
345 NM/19 Bugzilla 745: TLS version reporting.
346 Patch provided by Phil Pennock.
348 NM/20 Bugzilla 167: bool: condition support.
349 Patch provided by Phil Pennock.
351 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
352 clients. Patch provided by Phil Pennock.
354 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
355 Patch provided by Brad "anomie" Jorsch.
357 NM/23 Bugzilla 687: Fix misparses in eximstats.
358 Patch provided by Heiko Schlittermann.
360 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
361 Patch provided by Heiko Schlittermann.
363 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
364 plus update to original patch.
366 NM/26 Bugzilla 799: Documentation correction for ratelimit.
368 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
369 Patch provided by David Brownlee.
371 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
373 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
375 NM/30 Bugzilla 888: TLS documentation bugfixes.
377 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
379 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
380 Unlike the original bugzilla I have changed all shell scripts in src tree.
382 NM/33 Bugzilla 898: Transport filter timeout fix.
383 Patch by Todd Rinaldo.
385 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
386 Patch by Serge Demonchaux.
388 NM/35 Bugzilla 39: Base64 decode bug fixes.
389 Patch by Jakob Hirsch.
391 NM/36 Bugzilla 909: Correct connect() call in dcc code.
393 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
395 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
397 NM/39 Bugzilla 911: Fixed MakeLinks build script.
403 TK/01 Add preliminary DKIM support. Currently requires a forked version of
404 ALT-N's libdkim that I have put here:
405 http://duncanthrax.net/exim-experimental/
407 Note to Michael Haardt: I had to rename some vars in sieve.c. They
408 were called 'true' and it seems that C99 defines that as a reserved
409 keyword to be used with 'bool' variable types. That means you could
410 not include C99-style headers which use bools without triggering
411 build errors in sieve.c.
413 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
414 as mailq or other aliases. Changed the --help handling significantly
415 to do whats expected. exim_usage() emits usage/help information.
417 SC/01 Added the -bylocaldomain option to eximstats.
419 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
421 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
423 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
429 PH/01 Another patch from the Sieve maintainer.
431 PH/02 When an IPv6 address is converted to a string for single-key lookup
432 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
433 dots are used instead of colons so that keys in lsearch files need not
434 contain colons. This was done some time before quoting was made available
435 in lsearch files. However, iplsearch files do require colons in IPv6 keys
436 (notated using the quote facility) so as to distinguish them from IPv4
437 keys. This meant that lookups for IP addresses in host lists did not work
438 for iplsearch lookups.
440 This has been fixed by arranging for IPv6 addresses to be expressed with
441 colons if the lookup type is iplsearch. This is not incompatible, because
442 previously such lookups could never work.
444 The situation is now rather anomolous, since one *can* have colons in
445 ordinary lsearch keys. However, making the change in all cases is
446 incompatible and would probably break a number of configurations.
448 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
451 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
452 conversion specification without a maximum field width, thereby enabling
453 a rogue spamd server to cause a buffer overflow. While nobody in their
454 right mind would setup Exim to query an untrusted spamd server, an
455 attacker that gains access to a server running spamd could potentially
456 exploit this vulnerability to run arbitrary code as the Exim user.
458 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
459 $primary_hostname instead of what libspf2 thinks the hosts name is.
461 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
462 a directory entry by the name of the lookup key. Previously, if a
463 symlink pointed to a non-existing file or a file in a directory that
464 Exim lacked permissions to read, a lookup for a key matching that
465 symlink would fail. Now it is enough that a matching directory entry
466 exists, symlink or not. (Bugzilla 503.)
468 PH/03 The body_linecount and body_zerocount variables are now exported in the
471 PH/04 Added the $dnslist_matched variable.
473 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
474 This means they are set thereafter only if the connection becomes
477 PH/06 Added the client_condition to authenticators so that some can be skipped
478 by clients under certain conditions.
480 PH/07 The error message for a badly-placed control=no_multiline_responses left
481 "_responses" off the end of the name.
483 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
485 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
486 (without spaces) instead of just copying the configuration text.
488 PH/10 Added the /noupdate option to the ratelimit ACL condition.
490 PH/11 Added $max_received_linelength.
492 PH/12 Added +ignore_defer and +include_defer to host lists.
494 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
495 way in which PCRE > 7.0 is built.
497 PH/14 Implemented queue_only_load_latch.
499 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
500 MAIL command. The effect was to mangle the value on 64-bit systems.
502 PH/16 Another patch from the Sieve maintainer.
504 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
506 PH/18 If a system quota error occurred while trying to create the file for
507 a maildir delivery, the message "Mailbox is full" was not appended to the
508 bounce if the delivery eventually timed out. Change 4.67/27 below applied
509 only to a quota excession during the actual writing of the file.
511 PH/19 It seems that peer DN values may contain newlines (and other non-printing
512 characters?) which causes problems in log lines. The DN values are now
513 passed through string_printing() before being added to log lines.
515 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
516 and InterBase are left for another time.)
518 PH/21 Added message_body_newlines option.
520 PH/22 Guard against possible overflow in moan_check_errorcopy().
522 PH/23 POSIX allows open() to be a macro; guard against that.
524 PH/24 If the recipient of an error message contained an @ in the local part
525 (suitably quoted, of course), incorrect values were put in $domain and
526 $local_part during the evaluation of errors_copy.
532 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
533 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
536 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
537 issue a MAIL command.
539 PH/02 In an ACL statement such as
541 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
543 if a client was not listed at all, or was listed with a value other than
544 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
545 the condition was not true (as it should be), so access was not denied.
546 The bug was that the ! inversion was incorrectly passed on to the second
547 item. This has been fixed.
549 PH/03 Added additional dnslists conditions == and =& which are different from
550 = and & when the dns lookup returns more than one IP address.
552 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
553 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
555 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
556 FSYNC, which compiles an option called disable_fsync that allows for
557 bypassing fsync(). The documentation is heavily laced with warnings.
559 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
561 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
562 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
563 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
564 including adding "make clean"; (3) Added -fPIC when compiling the test
565 dynamically loaded module, to get rid of a warning.
567 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
568 message fails, move_frozen_messages = true and ignore_bounce_errors_after
569 = 0s. The bug is otherwise harmless.
571 PH/07 There was a bug in the dovecot authenticator such that the value of
572 $auth1 could be overwritten, and so not correctly preserved, after a
573 successful authentication. This usually meant that the value preserved by
574 the server_setid option was incorrect.
576 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
578 PH/09 Installed PCRE release 7.0.
580 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
581 run for batched SMTP input. It is now run at the start of every message
582 in the batch. While fixing this I discovered that the process information
583 (output by running exiwhat) was not always getting set for -bs and -bS
584 input. This is fixed, and it now also says "batched" for BSMTP.
586 PH/11 Added control=no_pipelining.
588 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
589 patch, slightly modified), and move the expansion of helo_data till after
590 the connection is made in the smtp transport (so it can use these
593 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
595 PH/14 Added log_selector = +pid.
597 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
599 PH/16 Add ${if forany and ${if forall.
601 PH/17 Added dsn_from option to vary the From: line in DSNs.
603 PH/18 Flush SMTP output before performing a callout, unless control =
604 no_callout_flush is set.
606 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
607 was true (the default) a successful delivery failed to delete the retry
608 item, thus causing premature timeout of the address. The bug is now
611 PH/20 Added hosts_avoid_pipelining to the smtp transport.
613 PH/21 Long custom messages for fakedefer and fakereject are now split up
614 into multiline reponses in the same way that messages for "deny" and
615 other ACL rejections are.
617 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
618 with slight modification.
620 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
621 draft, changing the syntax and factoring some duplicate code".
623 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
624 for deliveries of the second and subsequent messages over the same SMTP
627 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
628 ${reduce, with only minor "tidies".
630 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
632 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
633 expansion side effects.
635 PH/27 When a message times out after an over-quota error from an Exim-imposed
636 quota, the bounce message says "mailbox is full". This message was not
637 being given when it was a system quota that was exceeded. It now should
640 MH/03 Made $recipients available in local_scan(). local_scan() already has
641 better access to the recipient list through recipients_list[], but
642 $recipients can be useful in postmaster-provided expansion strings.
644 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
645 in the case of a MAIL command with additional options following the
646 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
647 were accidentally chopped off.
649 PH/29 SMTP synchronization checks are implemented when a command is read -
650 there is a check that no more input is waiting when there shouldn't be
651 any. However, for some commands, a delay in an ACL can mean that it is
652 some time before the response is written. In this time, more input might
653 arrive, invalidly. So now there are extra checks after an ACL has run for
654 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
655 pipelining has not been advertised.
657 PH/30 MH's patch to allow iscntrl() characters to be list separators.
659 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
660 returned in the SMTP response when smtp_return_error_details was false.
663 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
664 instead of the C I/O that was originally supplied, because problems were
667 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
668 Exim which did not show up earlier: it was assuming that a call to
669 SSL_CTX_set_info_callback() might give an error value. In fact, there is
670 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
671 was a macro that became an assignment, so it seemed to work. This has
672 changed to a proper function call with a void return, hence the compile
673 error. Exim's code has been fixed.
675 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
678 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
680 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
682 PH/37 If a message is not accepted after it has had an id assigned (e.g.
683 because it turns out to be too big or there is a timeout) there is no
684 "Completed" line in the log. When some messages of this type were
685 selected by exigrep, they were listed as "not completed". Others were
686 picked up by some special patterns. I have improved the selection
687 criteria to be more general.
689 PH/38 The host_find_failed option in the manualroute router can now be set
690 to "ignore", to completely ignore a host whose IP address cannot be
691 found. If all hosts are ignored, the behaviour is controlled by the new
692 host_all_ignored option.
694 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
695 homing or because of multiple MX records with /mx) generated more than
696 one IP address, and the following item turned out to be the local host,
697 all the secondary addresses of the first item were incorrectly removed
698 from the list, along with the local host and any following hosts (which
699 is what is supposed to happen).
701 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
702 whoever called Exim into the -H file. In the case of the daemon it was
703 behaving confusingly. When first started, it used values for whoever
704 started the daemon, but after a SIGHUP it used the Exim user (because it
705 calls itself on a restart). I have changed the code so that it now always
708 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
709 message are rejected with the same error (e.g. no authentication or bad
710 sender address), and a DATA command is nevertheless sent (as can happen
711 with PIPELINING or a stupid MUA), the error message that was given to the
712 RCPT commands is included in the rejection of the DATA command. This is
713 intended to be helpful for MUAs that show only the final error to their
716 PH/42 Another patch from the Sieve maintainer.
718 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
719 Eximstats - Fixed some broken HTML links and added missing column headers
721 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
722 columns for Rejects, Temp Rejects, Ham, and Spam rows.
724 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
726 PH/43 Yet another patch from the Sieve maintainer.
728 PH/44 I found a way to check for a TCP/IP connection going away before sending
729 the response to the final '.' that terminates a message, but only in the
730 case where the client has not sent further data following the '.'
731 (unfortunately, this is allowed). However, in many cases there won't be
732 any further data because there won't be any more messages to send. A call
733 to select() can be used: if it shows that the input is "ready", there is
734 either input waiting, or the socket has been closed. An attempt to read
735 the next input character can distinguish the two cases. Previously, Exim
736 would have sent an OK response which the client would never have see.
737 This could lead to message repetition. This fix should cure that, at
738 least in a lot of common cases.
740 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
741 advertised in response to EHLO.
747 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
748 fixed by 4.65/MH/01 (is this a record?) are fixed:
750 (i) An empty string was always treated as zero by the numeric comparison
751 operators. This behaviour has been restored.
753 (ii) It is documented that the numeric comparison operators always treat
754 their arguments as decimal numbers. This was broken in that numbers
755 starting with 0 were being interpreted as octal.
757 While fixing these problems I realized that there was another issue that
758 hadn't been noticed. Values of message_size_limit (both the global option
759 and the transport option) were treated as octal if they started with 0.
760 The documentation was vague. These values are now always treated as
761 decimal, and I will make that clear in the documentation.
767 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
768 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
771 MH/01 Don't check that the operands of numeric comparison operators are
772 integers when their expansion is in "skipping" mode (fixes bug
773 introduced by 4.64-PH/07).
775 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
776 child addresses, Exim now panics and dies. Previously, because the count
777 is held in a short int, deliveries were likely to be lost. As such a
778 large number of recipients for a single message is ridiculous
779 (performance will be very, very poor), I have chosen to impose a limit
780 rather than extend the field.
786 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
787 leftover -K file (the existence of which was triggered by #402).
788 While we were at it, introduced process PID as part of the -K
789 filename. This should rule out race conditions when creating
792 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
793 processing considerably. Previous code took too long for large mails,
794 triggering a timeout which in turn triggers #401.
796 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
797 in the DK code in transports.c. sendfile() is not really portable,
798 hence the _LINUX specificness.
800 TF/01 In the add_headers option to the mail command in an Exim filter,
801 there was a bug that Exim would claim a syntax error in any
802 header after the first one which had an odd number of characters
805 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
806 callout verification, Exim cached a "reject" for the entire domain. This
807 is correct for most verifications, but it is not correct for a recipient
808 verification with use_sender or use_postmaster set, because in that case
809 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
810 case of MAIL FROM:<> rejection from other early rejections (e.g.
811 rejection of HELO). When verifying a recipient using a non-null MAIL
812 address, the cache is ignored if it shows MAIL FROM:<> rejection.
813 Whatever the result of the callout, the value of the domain cache is
814 left unchanged (for any other kind of callout, getting as far as trying
815 RCPT means that the domain itself is ok).
817 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
820 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
821 manifest itself as EPIPE rather than ECONNECT. When tidying away a
822 session, the daemon ignores ECONNECT errors and logs others; it now
823 ignores EPIPE as well.
825 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
826 (quoted-printable decoding).
828 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
829 later the small subsequent patch to fix an introduced bug.
831 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
833 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
835 PH/08 An error is now given if message_size_limit is specified negative.
837 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
838 to be given (somewhat) arbitrary names.
840 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
843 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
844 miscellaneous code fixes
846 PH/10 Added the log_reject_target ACL modifier to specify where to log
849 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
850 hostname. This is wrong, because it relates to the incoming message (and
851 probably the interface on which it is arriving) and not to the outgoing
852 callout (which could be using a different interface). This has been
853 changed to use the value of the helo_data option from the smtp transport
854 instead - this is what is used when a message is actually being sent. If
855 there is no remote transport (possible with a router that sets up host
856 addresses), $smtp_active_hostname is used.
858 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
859 tweaks were necessary in order to get it to work (see also 21 below):
860 (a) The code assumed that strncpy() returns a negative number on buffer
861 overflow, which isn't the case. Replaced with Exim's string_format()
863 (b) There were several signed/unsigned issues. I just did the minimum
864 hacking in of casts. There is scope for a larger refactoring.
865 (c) The code used strcasecmp() which is not a standard C function.
866 Replaced with Exim's strcmpic() function.
867 (d) The code set only $1; it now sets $auth1 as well.
868 (e) A simple test gave the error "authentication client didn't specify
869 service in request". It would seem that Dovecot has changed its
870 interface. Fortunately there's a specification; I followed it and
871 changed what the client sends and it appears to be working now.
873 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
876 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
877 address A is aliased to B and C, where B exists and C does not. Without
878 -v the output is "A verified" because verification stops after a
879 successful redirection if more than one address is generated. However,
880 with -v the child addresses are also verified. Exim was outputting "A
881 failed to verify" and then showing the successful verification for C,
882 with its parentage. It now outputs "B failed to verify", showing B's
883 parentage before showing the successful verification of C.
885 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
886 look up a TXT record in a specific list after matching in a combined
889 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
890 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
891 they consult the DNS. I had assumed they would set it the way they
892 wanted; and indeed my experiments on Linux seem to show that in some
893 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
894 To be on the safe side, however, I have now made the interface to
895 host_find_byname() similar to host_find_bydns(), with an argument
896 containing the DNS resolver options. The host_find_byname() function now
897 sets these options at its start, just as host_find_bydns() does. The smtp
898 transport options dns_qualify_single and dns_search_parents are passed to
899 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
900 of host_find_byname() use the default settings of RES_DEFNAMES
901 (qualify_single) but not RES_DNSRCH (search_parents).
903 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
904 spool_read_header() do less string testing, by means of a preliminary
905 switch on the second character of optional "-foo" lines. (This is
906 overdue, caused by the large number of possibilities that now exist.
907 Originally there were few.) While I was there, I also converted the
908 str(n)cmp tests so they don't re-test the leading "-" and the first
909 character, in the hope this might squeeze out yet more improvement.
911 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
912 flag allowing group syntax was set by the header_syntax check but not
913 turned off, possible causing trouble later; (2) The flag was not being
914 set at all for the header_verify test, causing "group"-style headers to
915 be rejected. I have now set it in this case, and also caused header_
916 verify to ignore an empty address taken from a group. While doing this, I
917 came across some other cases where the code for allowing group syntax
918 while scanning a header line wasn't quite right (mostly, not resetting
919 the flag correctly in the right place). These bugs could have caused
920 trouble for malformed header lines. I hope it is now all correct.
922 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
923 with the "reply" argument non-NULL. The code, however (which originally
924 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
925 but it didn't always do it. This confused somebody who was copying the
926 code for some other use. I have removed all the tests.
928 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
929 feature that was used to support insecure browsers during the U.S. crypto
930 embargo. It requires special client support, and Exim is probably the
931 only MTA that supported it -- and would never use it because real RSA is
932 always available. This code has been removed, because it had the bad
933 effect of slowing Exim down by computing (never used) parameters for the
934 RSA_EXPORT functionality.
936 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
937 authenticator to fail if there's a tab character in the incoming data
938 (there should never be unless someone is messing about, as it's supposed
939 to be base64-encoded). Also added, on Timo's advice, the "secured" option
940 if the connection is using TLS or if the remote IP is the same as the
941 local IP, and the "valid-client-cert option" if a client certificate has
944 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
945 authenticators. This can be used for authorization after authentication
946 succeeds. (In the case of plaintext, it servers for both authentication
949 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
950 if any retry times were supplied.
952 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
953 connection, where there is no client IP address to check. In this
954 situation, the verify now always succeeds.
956 PH/25 Applied John Jetmore's -Mset patch.
958 PH/26 Added -bem to be like -Mset, but loading a message from a file.
960 PH/27 In a string expansion for a processed (not raw) header when multiple
961 headers of the same name were present, leading whitespace was being
962 removed from all of them, but trailing whitespace was being removed only
963 from the last one. Now trailing whitespace is removed from each header
964 before concatenation. Completely empty headers in a concatenation (as
967 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
968 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
970 PH/29 [Removed. This was a change that I later backed out, and forgot to
971 correct the ChangeLog entry (that I had efficiently created) before
972 committing the later change.]
974 PH/30 Exim was sometimes attempting to deliver messages that had suffered
975 address errors (4xx response to RCPT) over the same connection as other
976 messages routed to the same hosts. Such deliveries are always "forced",
977 so retry times are not inspected. This resulted in far too many retries
978 for the affected addresses. The effect occurred only when there were more
979 hosts than the hosts_max_try setting in the smtp transport when it had
980 the 4xx errors. Those hosts that it had tried were not added to the list
981 of hosts for which the message was waiting, so if all were tried, there
982 was no problem. Two fixes have been applied:
984 (i) If there are any address or message errors in an SMTP delivery, none
985 of the hosts (tried or untried) are now added to the list of hosts
986 for which the message is waiting, so the message should not be a
987 candidate for sending over the same connection that was used for a
988 successful delivery of some other message. This seems entirely
989 reasonable: after all the message is NOT "waiting for some host".
990 This is so "obvious" that I'm not sure why it wasn't done
991 previously. Hope I haven't missed anything, but it can't do any
992 harm, as the worst effect is to miss an optimization.
994 (ii) If, despite (i), such a delivery is accidentally attempted, the
995 routing retry time is respected, so at least it doesn't keep
996 hammering the server.
998 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
999 in ${readsocket because some servers need this prod.
1001 PH/32 Added some extra debug output when updating a wait-xxx database.
1003 PH/33 The hint "could be header name not terminated by colon", which has been
1004 given for certain expansion errors for a long time, was not being given
1005 for the ${if def:h_colon_omitted{... case.
1007 PH/34 The spec says: "With one important exception, whenever a domain list is
1008 being scanned, $domain contains the subject domain." There was at least
1009 one case where this was not true.
1011 PH/35 The error "getsockname() failed: connection reset by peer" was being
1012 written to the panic log as well as the main log, but it isn't really
1013 panic-worthy as it just means the connection died rather early on. I have
1014 removed the panic log writing for the ECONNRESET error when getsockname()
1017 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1018 runs only) independently of the message's sender address. This meant
1019 that, if the 4xx error was in fact related to the sender, a different
1020 message to the same recipient with a different sender could confuse
1021 things. In particualar, this can happen when sending to a greylisting
1022 server, but other circumstances could also provoke similar problems.
1023 I have changed the default so that the retry time for these errors is now
1024 based a combination of the sender and recipient addresses. This change
1025 can be overridden by setting address_retry_include_sender=false in the
1028 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1029 remote server are returned as part of bounce messages. This was not
1030 happening for LMTP over a pipe (the lmtp transport), but now it is the
1031 same for both kinds of LMTP.
1033 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1034 in header lines that were in fact CNAMEs. This is no longer the case.
1036 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1037 and queue runs started by the daemon processed all messages. This has
1038 been fixed so that -R and -S can now usefully be given with -q<time>.
1040 PH/40 Import PCRE release 6.7 (fixes some bugs).
1042 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1044 PH/42 Give an error if -q is specified more than once.
1046 PH/43 Renamed the variables $interface_address and $interface_port as
1047 $received_ip_address and $received_port, to make it clear that these
1048 values apply to message reception, and not to the outgoing interface when
1049 a message is delivered. (The old names remain recognized, of course.)
1051 PH/44 There was no timeout on the connect() call when using a Unix domain
1052 socket in the ${readsocket expansion. There now is.
1054 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1055 be meaningful with "accept".
1057 SC/01 Eximstats V1.43
1058 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1060 SC/02 Eximstats V1.44
1061 Use a glob alias rather than an array ref in the generated
1062 parser. This improves both readability and performance.
1064 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1065 Collect SpamAssassin and rejection statistics.
1066 Don't display local sender or destination tables unless
1067 there is data to show.
1068 Added average volumes into the top table text output.
1070 SC/04 Eximstats V1.46
1071 Collect data on the number of addresses (recipients)
1072 as well as the number of messages.
1074 SC/05 Eximstats V1.47
1075 Added 'Message too big' to the list of mail rejection
1076 reasons (thanks to Marco Gaiarin).
1078 SC/06 Eximstats V1.48
1079 Mainlog lines which have GMT offsets and are too short to
1080 have a flag are now skipped.
1082 SC/07 Eximstats V1.49 (Alain Williams)
1083 Added the -emptyok flag.
1085 SC/08 Eximstats V1.50
1086 Fixes for obtaining the IP address from reject messages.
1088 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1089 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1090 whitesspace changes from 4.64-PH/27
1092 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1095 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1096 are found, allow negative numbers in numeric criteria)
1098 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1100 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1101 to match changes made in 4.64-PH/43
1103 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1105 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1106 "domain" is actually an IP address, and force a failure. This locks out
1107 those revolvers/nameservers that support "A-for-A" lookups, in
1108 contravention of the specifications.
1110 PH/48 When a host name was looked up from an IP address, and the subsequent
1111 forward lookup of the name timed out, the host name was left in
1112 $sender_host_name, contrary to the specification.
1114 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1115 restricted to single-key lookups, Exim was not diagnosing an error if
1116 * or *@ was used with a query-style lookup.
1118 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1120 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1121 long ago, but noone interested enough thought of it. Let's just say that
1122 the "1.1" means that there are some new functions that weren't there at
1123 some point in the past.
1125 PH/51 Error processing for expansion failure of helo_data from an smtp
1126 transport during callout processing was broken.
1128 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1129 tested/used via the -bh/-bhc/-bs options.
1131 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1132 bug, fixed in subsequent PCRE releases).
1134 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1135 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1137 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1143 SC/01 Use a glob alias rather than an array ref in eximstats generated
1144 parser. This improves both readability and performance.
1146 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1147 Don't display local sender or destination tables in eximstats unless
1148 there is data to show.
1149 Added average volumes into the eximstats top table text output.
1151 SC/03 Collect data on the number of addresses (recipients) as well
1152 as the number of messages in eximstats.
1154 TF/01 Correct an error in the documentation for the redirect router. Exim
1155 does (usually) call initgroups() when daemonizing.
1157 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1158 with consistent privilege compared to when running as a daemon.
1160 TF/03 Note in the spec that $authenticated_id is not set for local
1161 submissions from trusted users.
1163 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1164 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1166 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1167 by adding some example configuration directives to the default
1168 configuration file. A little bit of work is required to uncomment the
1169 directives and define how usernames and passwords are checked, but
1170 there is now a framework to start from.
1172 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1173 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1174 without this. I don't know how relevant this is to other LDAP libraries.
1176 PH/02 Add the verb name to the "unknown ACL verb" error.
1178 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1180 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1182 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1183 directory not expanded when it should be if an expanded home directory
1184 was set for the address (which is overridden by the transport).
1186 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1189 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1190 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1191 because it is too late at that time, and has no effect.
1193 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1194 security issue with \' (bugzilla #107). I could not use the
1195 PQescapeStringConn() function, because it needs a PGconn value as one of
1198 PH/08 When testing addresses using -bt, indicate those final addresses that
1199 are duplicates that would not cause an additional delivery. At least one
1200 person was confused, thinking that -bt output corresponded to deliveries.
1201 (Suppressing duplicates isn't a good idea as you lose the information
1202 about possibly different redirections that led to the duplicates.)
1204 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1205 systems where poll() doesn't work, in particular OS X.
1207 PH/10 Added more information to debugging output for retry time not reached.
1209 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1210 operations in malware.c.
1212 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1215 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1216 syslog_duplication set false, log lines that would normally be written
1217 both the the main log and to the reject log were not written to syslog at
1220 PH/14 In the default configuration, change the use of "message" in ACL warn
1221 statements to "add_header".
1223 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1224 not followed by a command (e.g. "seen endif").
1226 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1227 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1230 PH/17 Added extra conditions to the default value of delay_warning_condition
1234 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1235 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1236 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1239 The Auto-Submitted: and various List- headers are standardised, whereas I
1240 don't think Precedence: ever was.
1242 PH/18 Refactored debugging code in route_finduser() to show more information,
1243 in particular, the error code if getpwnam() issues one.
1245 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1246 This is apparently needed in addition to the PH/07 change above to avoid
1247 any possible encoding problems.
1249 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1250 but not after initializing Perl.
1252 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1253 output them only if debugging. By default they are written stderr,
1254 apparently, which is not desirable.
1256 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1259 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1262 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1264 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1265 authenticated or an ident call has been made. Suppress the default
1266 values for $authenticated_id and $authenticated_sender (but permit -oMai
1267 and -oMas) when testing with -bh.
1269 PH/24 Re-jigged the order of the tests in the default configuration so that the
1270 tests for valid domains and recipients precede the DNS black list and CSA
1271 tests, on the grounds that those ones are more expensive.
1273 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1274 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1275 This bug exists in every version of Exim that I still have, right back to
1278 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1279 However, an attempt to turn on case-sensitivity in a regex key by
1280 including (?-i) didn't work because the subject string was already
1281 lowercased, and the effects were non-intuitive. It turns out that a
1282 one-line patch can be used to allow (?-i) to work as expected.
1288 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1289 other effects) broke the use of negated acl sub-conditions.
1291 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1294 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1295 "Deny" causes Exim to reject the incoming connection with a 554 error.
1296 Unfortunately, if there is a major crisis, such as a disk failure,
1297 tcp-wrappers gives "deny", whereas what one would like would be some
1298 kind of temporary error. A kludge has been added to help with this.
1299 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1300 554 error is used if errno is still zero or contains ENOENT (which occurs
1301 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1304 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1306 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1307 errors. Otherwise a message that provokes a temporary error (when other
1308 messages do not) can cause a whole host to time out.
1310 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1311 addresses were routed directly to files or pipes from a redirect router.
1312 File deliveries just didn't batch; pipe deliveries might have suffered
1315 PH/06 A failure to get a lock for a hints database would erroneously always say
1316 "Failed to get write lock", even when it was really a read lock.
1318 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1319 of 0600. This has been changed to use the value of the lockfile_mode
1320 option (which defaults to 0600).
1322 PH/08 Applied small patch from the Sieve maintainer.
1324 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1325 folder from quota calculations, a direct delivery into this folder messed
1326 up the contents of the maildirsize file. This was because the regex was
1327 used only to exclude .Trash (or whatever) when the size of the mailbox
1328 was calculated. There was no check that a delivery was happening into an
1329 excluded directory. This bug has been fixed by ignoring all quota
1330 processing for deliveries into excluded directories.
1332 PH/10 Added the maildirfolder_create_regex option to appendfile.
1338 PH/01 The code for finding all the local interface addresses on a FreeBSD
1339 system running IPv6 was broken. This may well have applied to all BSD
1340 systems, as well as to others that have similar system calls. The broken
1341 code found IPv4 interfaces correctly, but gave incorrect values for the
1342 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1343 that it would not match correctly against @[] and not recognize the IPv6
1346 PH/02 The ipliteral router was not recognizing addresses of the form user@
1347 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1349 PH/03 Added disable_ipv6.
1351 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1352 decoded form, because it is most often used to construct To: headers
1353 lines in autoreplies, and the decoded form may well be syntactically
1354 invalid. However, $reply_address has leading white space removed, and all
1355 newlines turned into spaces so that the autoreply transport does not
1358 PH/05 If group was specified without a user on a router, and no group or user
1359 was specified on a transport, the group from the router was ignored.
1361 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1362 for visible compile-time settings that can be used to change these
1363 numbers, for those that want even more. Backwards compatibility with old
1364 spool files has been maintained. However, going back to a previous Exim
1365 release will lost any variables that are in spool files.
1367 PH/07 Two small changes when running in the test harness: increase delay when
1368 passing a TCP/IP connection to a new process, in case the original
1369 process has to generate a bounce, and remove special handling of
1370 127.0.0.2 (sic), which is no longer necessary.
1372 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1373 be the same on different OS.
1375 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1378 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1379 whether --show-vars was specified or not
1381 JJ/02 exipick: Added support for new ACL variable spool format introduced
1384 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1385 syntactically invalid From: or Reply-to: line, and a filter used this to
1386 generate an autoreply, and therefore failed to obtain an address for the
1387 autoreply, Exim could try to deliver to a non-existent relative file
1388 name, causing unrelated and misleading errors. What now happens is that
1389 it logs this as a hard delivery error, but does not attempt to create a
1392 PH/11 The exinext utility has a -C option for testing purposes, but although
1393 the given file was scanned by exinext itself; it wasn't being passed on
1394 when Exim was called.
1396 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1397 an end-of-file indication when reading a command response.
1399 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1400 compiled. In many other places in Exim, IPv6 addresses are always
1401 recognized, so I have changed this. It also means that IPv4 domain
1402 literals of the form [IPV4:n.n.n.n] are now always recognized.
1404 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1405 used if the router is not running as root, for example, when verifying at
1406 ACL time, or when using -bh. The debugging output from this situation was
1407 non-existent - all you got was a failure to exec. I have made two
1410 (a) Failures to set uid/gid, the current directory, or a process leader
1411 in a subprocess such as that created by queryprogram now generate
1412 suitable debugging ouput when -d is set.
1414 (b) The queryprogram router detects when it is not running as root,
1415 outputs suitable debugging information if -d is set, and then runs
1416 the subprocess without attempting to change uid/gid.
1418 PH/15 Minor change to Makefile for building test_host (undocumented testing
1421 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1422 additional section of a DNS packet that returns MX or SRV records.
1423 Instead, it always explicitly searches for A/AAAA records. This avoids
1424 major problems that occur when a DNS server includes only records of one
1425 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1426 fixed another bug: if SRV records were looked up and the corresponding
1427 address records were *not* found in the additional section, the port
1428 values from the SRV records were lost.
1430 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1431 using the correct key (the original address) when searching the retry
1432 rules in order to find which one to use for generating the retry hint.
1434 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1435 adding the default one. Similarly, if it contains a Reply-To: header, the
1436 errors_reply_to option, if set, is not used.
1438 PH/19 When calculating a retry time, Exim used to measure the "time since
1439 failure" by looking at the "first failed" field in the retry record. Now
1440 it does not use this if it is later than than the arrival time of the
1441 message. Instead it uses the arrival time. This makes for better
1442 behaviour in cases where some deliveries succeed, thus re-setting the
1443 "first failed" field. An example is a quota failure for a huge message
1444 when small messages continue to be delivered. Without this change, the
1445 "time since failure" will always be short, possible causing more frequent
1446 delivery attempts for the huge message than are intended.
1447 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1449 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1450 $1, $2, $3) because the numerical variables can be reset during some
1451 expansion items (e.g. "match"), thereby losing the authentication data.
1453 PH/21 Make -bV show the size of off_t variables so that the test suite can
1454 decide whether to run tests for quotas > 2G.
1456 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1457 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1458 filecount value is greater than 2G or if a quota value is greater than 2G
1459 on a system where the size of off_t is not greater than 4, a panic error
1462 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1463 never match. The debug and -bh output now contains an explicit error
1464 message indicating a malformed IPv4 address or mask.
1466 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1467 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1468 PH/23 above applies.
1470 PH/25 Do not write to syslog when running in the test harness. The only
1471 occasion when this arises is a failure to open the main or panic logs
1472 (for which there is an explicit test).
1474 PH/26 Added the /no_tell option to "control=freeze".
1476 PH/27 If a host name lookup failed very early in a connection, for example, if
1477 the IP address matched host_lookup and the reverse lookup yielded a name
1478 that did not have a forward lookup, an error message of the form "no IP
1479 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1480 could be logged. Now it outputs the IP address instead of "NULL".
1482 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1483 allows the sender and the authenticated sender to be set when
1484 submitting a message from within Exim. Since child_open_exim() is
1485 documented for local_scan(), the new function should be too.
1487 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1488 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1489 results in an empty string is now treated as unset.
1491 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1493 PH/31 Added sender_verify_fail logging option.
1495 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1496 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1497 tidied the source and removed it altogether.
1499 PH/33 When a queue run was abandoned because the load average was too high, a
1500 log line was always written; now it is written only if the queue_run log
1501 selector is set. In addition, the log line for abandonment now contains
1502 information about the queue run such as the pid. This is always present
1503 in "start" and "stop" lines but was omitted from the "abandon" line.
1505 PH/34 Omit spaces between a header name and the colon in the error message that
1506 is given when verify = headers_syntax fails (if there are lots of them,
1507 the message gets confusing).
1509 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1510 names, as there are now some PTR records that contain slashes. This check
1511 is only to protect against broken name servers that fall over on strange
1512 characters, so the fact that it applies to all lookups doesn't matter.
1514 PH/36 Now that the new test suite is complete, we can remove some of the
1515 special code in Exim that was needed for the old test suite. For example,
1516 sorting DNS records because real resolvers return them in an arbitrary
1517 order. The new test suite's fake resolver always returns records in the
1520 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1521 bounces) except when queue_only is set, to avoid logging races between
1522 the different processes.
1524 PH/38 Panic-die if .include specifies a non-absolute path.
1526 PH/39 A tweak to the "H" retry rule from its user.
1528 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1529 a label. They prevented compilation on older perls.
1531 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1532 a warning to be raised on newish perls.
1534 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1535 on queue. Changes to match documented behaviour of showing count of
1536 messages matching specified criteria.
1538 PH/40 Changed the default ident timeout from 30s to 5s.
1540 PH/41 Added support for the use of login_cap features, on those BSD systems
1541 that have them, for controlling the resources used by pipe deliveries.
1543 PH/42 The content-scanning code uses fopen() to create files in which to put
1544 message data. Previously it was not paying any attention to the mode of
1545 the files. Exim runs with umask(0) because the rest of the code creates
1546 files with open(), and sets the required mode explicitly. Thus, these
1547 files were ending up world-writeable. This was not a big issue, because,
1548 being within the spool directory, they were not world-accessible. I have
1549 created a function called modefopen, which takes an additional mode
1550 argument. It sets umask(777), creates the file, chmods it to the required
1551 mode, then resets the umask. All the relevant calls to fopen() in the
1552 content scanning code have been changed to use this function.
1554 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1555 to 24 hours. This avoids potential overflow problems when processing G
1556 and H retry rules. I suspect nobody ever tinkers with this value.
1558 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1560 PH/45 When the plaintext authenticator is running as a client, the server's
1561 challenges are checked to ensure they are valid base64 strings. By
1562 default, the authentication attempt is cancelled if an invalid string is
1563 received. Setting client_ignore_invalid_base64 true ignores these errors.
1564 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1565 they are received. Thus, the responses can be made to depend on the
1566 challenges. If an invalid string is ignored, an empty string is placed in
1569 PH/46 Messages that are created by the autoreply transport now contains a
1570 References: header, in accordance with RFCs 2822 and 3834.
1572 PH/47 Added authenticated_sender_force to the smtp transport.
1574 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1576 PH/49 Installed latest patch from the Sieve maintainer.
1578 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1579 was also set, the appendfile transport was unnecessarily scanning a
1580 directory of message files (e.g. for maildir delivery) to find the count
1581 of files (along with the size), even though it did not need this
1582 information. It now does the scan only if it needs to find either the
1583 size of the count of files.
1585 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1587 PH/52 Two bugs concerned with error handling when the smtp transport is
1590 (i) Exim was not creating retry information for temporary errors given
1591 for individual recipients after the DATA command when the smtp transport
1592 was used in LMTP mode. This meant that they could be retried too
1593 frequently, and not timed out correctly.
1595 (ii) Exim was setting the flag that allows error details to be returned
1596 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1597 recipients that were returned after the DATA command.
1599 PH/53 This is related to PH/52, but is more general: for any failing address,
1600 when detailed error information was permitted to be returned to the
1601 sender, but the error was temporary, then after the final timeout, only
1602 "retry timeout exceeded" was returned. Now it returns the full error as
1603 well as "retry timeout exceeded".
1605 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1606 do this, and (what is worse) MTAs that accept it.
1608 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1609 will now be deprecated.
1611 PH/56 New os.c-cygwin from the Cygwin maintainer.
1613 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1614 formats (previously only available in exim formats via -bpr, -bpru,
1615 and -bpra. Now also available in native and exiqgrep formats)
1617 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1618 with very large, slow to parse queues
1620 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1622 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1624 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1625 responses to authentication challenges, though it was showing the
1626 challenges; (ii) I've removed the CR characters from the debug output for
1629 PH/58 Allow for the insertion of a newline as well as a space when a string
1630 is turned into more than one encoded-word during RFC 2047 encoding. The
1631 Sieve code now uses this.
1633 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1634 data_4xx, lost_connection, tls_required.
1636 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1637 message was being sent as an SMTP response.
1639 PH/61 Add -l and -k options to exicyclog.
1641 PH/62 When verifying, if an address was redirected to one new address, so that
1642 verification continued, and the new address failed or deferred after
1643 having set something in $address_data, the value of $address_data was not
1644 passed back to the ACL. This was different to the case when no
1645 redirection occurred. The value is now passed back in both cases.
1647 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1648 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1649 use login_cap.h, so on its own it isn't the distinguishing feature. The
1650 new name refers directly to the setclassresources() function.
1652 PH/65 Added configuration files for NetBSD3.
1654 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1656 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1657 is preferred over IPv4.
1659 PH/68 The bounce_return_message and bounce_return_body options were not being
1660 honoured for bounces generated during the reception of non-SMTP messages.
1661 In particular, this applied to messages rejected by the ACL. This bug has
1662 been fixed. However, if bounce_return_message is true and bounce_return_
1663 body is false, the headers that are returned for a non-SMTP message
1664 include only those that have been read before the error was detected.
1665 (In the case of an ACL rejection, they have all been read.)
1667 PH/69 The HTML version of the specification is now built in a directory called
1668 spec_html instead of spec.html, because the latter looks like a path with
1669 a MIME-type, and this confuses some software.
1671 PH/70 Catch two compiler warnings in sieve.c.
1673 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1674 function verify_get_ident() calls ip_connect() to connect a socket, but
1675 if the "connect()" function timed out, ip_connect() used to close the
1676 socket. However, verify_get_ident() also closes the socket later, and in
1677 between Exim writes to the log, which may get opened at this point. When
1678 the socket was closed in ip_connect(), the log could get the same file
1679 descriptor number as the socket. This naturally causes chaos. The fix is
1680 not to close the socket in ip_connect(); the socket should be closed by
1681 the function that creates it. There was only one place in the code where
1682 this was missing, in the iplookup router, which I don't think anybody now
1683 uses, but I've fixed it anyway.
1685 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1686 well as to direct DNS lookups. Otherwise the handling of names in host
1687 lists is inconsistent and therefore confusing.
1693 PH/01 Two changes to the default runtime configuration:
1695 (1) Move the checks for relay_from_hosts and authenticated clients from
1696 after to before the (commented out) DNS black list checks.
1698 (2) Add control=submission to the relay_from_hosts and authenticated
1699 clients checks, on the grounds that messages accepted by these
1700 statements are most likely to be submissions.
1702 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1704 (1) Generate an error if the third argument for the ${prvs expansion is
1707 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1710 (3) Reset the variables that are obtained from the first argument of
1711 ${prvscheck and used in the second argument before leaving the code,
1712 because their memory is reclaimed, so using them afterwards may do
1715 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1716 one (it's much easier than Tom thought :-).
1718 (5) Because of (4), we can now allow for the use of $prvscheck_result
1719 inside the third argument.
1721 PH/03 For some reason, the default setting of PATH when running a command from
1722 a pipe transport was just "/usr/bin". I have changed it to
1725 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1726 anything to be listed in the output from -bV.
1728 PH/05 When a filter generated an autoreply, the entire To: header line was
1729 quoted in the delivery log line, like this:
1731 => >A.N.Other <ano@some.domain> <original@ddress> ...
1733 This has been changed so that it extracts the operative address. There
1734 may be more than one such address. If so, they are comma-separated, like
1737 => >ano@some.domain,ona@other.domain <original@ddress> ...
1739 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1740 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1741 not being looked up in the rDNS to get a host name, Exim was showing the
1742 IP address twice in Received: lines, even though the IP addresses were
1743 identical. For example:
1745 Received: from [1.2.3.4] (helo=[1.2.3.4])
1747 However, if the real host name was known, it was omitting the HELO data
1748 if it matched the actual IP address. This has been tidied up so that it
1749 doesn't show the same IP address twice.
1751 PH/07 When both +timestamp and +memory debugging was on, the value given by
1752 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1753 called by the debug printing, thereby overwriting the timestamp buffer.
1754 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1756 PH/08 When the original message was included in an autoreply transport, it
1757 always said "this is a copy of the message, including all the headers",
1758 even if body_only or headers_only was set. It now gives an appropriate
1761 PH/09 Applied a patch from the Sieve maintainer which:
1763 o fixes some comments
1764 o adds the (disabled) notify extension core
1765 o adds some debug output for the result of if/elsif tests
1766 o points to the current vacation draft in the documentation
1767 and documents the missing references header update
1771 o fixes a bug in processing the envelope test (when testing
1772 multiple envelope elements, the last element determinted the
1775 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1776 Electronic Mail") by including:
1778 Auto-submitted: auto-generated
1780 in the messages that it generates (bounce messages and others, such as
1781 warnings). In the case of bounce messages for non-SMTP mesages, there was
1782 also a typo: it was using "Auto_submitted" (underscore instead of
1783 hyphen). Since every message generated by Exim is necessarily in response
1784 to another message, thes have all been changed to:
1786 Auto-Submitted: auto-replied
1788 in accordance with these statements in the RFC:
1790 The auto-replied keyword:
1792 - SHOULD be used on messages sent in direct response to another
1793 message by an automatic process,
1795 - MUST NOT be used on manually-generated messages,
1797 - MAY be used on Delivery Status Notifications (DSNs) and Message
1798 Disposition Notifications (MDNs),
1800 - MUST NOT be used on messages generated by automatic or periodic
1801 processes, except for messages which are automatic responses to
1804 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1805 to the default Received: header definition.
1807 PH/12 Added log selector acl_warn_skipped (default on).
1809 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1810 variables because (a) they are in the wrong storage pool and (b) even if
1811 they were copied, it wouldn't work properly because of the caching.
1813 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1814 checking when decoding. Apparently there are clients that generate
1815 overlong encoded strings. Why am I not surprised?
1817 PH/15 If the first argument of "${if match_address" was not empty, but did not
1818 contain an "@" character, Exim crashed. Now it writes a panic log message
1819 and treats the condition as false.
1821 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1823 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1824 extension "envelope-auth". The code is finished and in agreement with
1825 other implementations, but there is no documentation so far and in fact,
1826 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1827 not changing the active code.
1829 Print executed "if" and "elsif" statements when debugging is used. This
1830 helps a great deal to understand what a filter does.
1832 Document more things not specified clearly in RFC3028. I had all this
1833 sorted out, when out of a sudden new issues came to my mind. Oops."
1835 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1838 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1839 canonical form (as documented). However, after a host name lookup from
1840 the IP address, check_host() was doing a simple string comparison with
1841 addresses acquired from the DNS when checking that the found name did
1842 have the original IP as one of its addresses. Since any found IPv6
1843 addresses are likely to be in abbreviated form, the comparison could
1844 fail. Luckily, there already exists a function for doing the comparison
1845 by converting both addresses to binary, so now that is used instead of
1846 the text comparison.
1848 PH/20 There was another similar case to PH/19, when a complete host name was
1849 given in a host list; looking up its IP address could give an abbreviated
1850 form, whereas the current host's name might or might not be abbreviated.
1851 The same fix has been applied.
1857 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1858 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1861 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1862 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1864 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1866 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1867 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1868 Also, for "Auto-Submitted", treat anything other than "no" as
1869 non-personal, in accordance with RFC 3834. (Previously it treated
1870 anything starting "auto-" as non-personal.)
1872 TF/01 The control=submission/name=... option had a problem with syntax
1873 errors if the name included a slash character. The /name= option
1874 now slurps the rest of the string, so it can include any characters
1875 but it must come last in the list of options (after /sender_retain
1878 PH/05 Some modifications to the interface to the fake nameserver for the new
1886 TK/01 Added the "success_on_redirect" address verification option. See
1887 NewStuff for rationale and an example.
1889 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1891 PH/02 Patch to exigrep to allow it to work on syslog lines.
1893 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1894 fread() to skip over the body file's header line, because in Cygwin the
1895 header line is locked and is inaccessible.
1897 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1898 co-exist for some time) to make it clear that it is the Exim ID that is
1899 referenced, not the Message-ID: header line.
1901 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1902 string_format() function, because snprintf() does not exist on all
1905 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1906 ${stat: expansion item.
1908 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1909 protocol synchronization error", to keep the pedants happy.
1911 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1912 well as for IRIX systems, when gcc is being used. See the host.c source
1915 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1917 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1920 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1921 transport and to the smtp transport in LMTP mode.
1923 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1925 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1926 run a filter in a subprocess. This could lead to confusion in subsequent
1927 lookups in the parent process. There should also be a search_tidyup() at
1928 the end of the subprocess.
1930 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1931 only if the host matched helo_try_verify_hosts, which caused the
1932 verification to occur when the EHLO/HELO command was issued. The ACL just
1933 tested the remembered result. Now, if a previous verification attempt has
1934 not happened, "verify = helo" does it there and then.
1936 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1938 TK/03 Fix log output including CR from clamd.
1940 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1941 exist provoked a memory error which could cause a segfault.
1943 PH/15 Installed PCRE 6.2
1945 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1947 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1948 of the problem. Specifically, suggested +O2 rather than +O1 for the
1951 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1953 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1954 record, the port was not correctly logged when the outgoing_port log
1955 selector was set (it logged the transort's default port).
1957 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1958 fallback_hosts, and "hosts" in the smtp transport.
1960 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1961 host errors such as "Connection refused".
1963 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1964 authentication with radiusclient 0.4.9:
1966 - Error returned from rc_read_config was caught wrongly
1967 - Username/password not passed on to radius server due to wrong length.
1969 The presumption is that some radiusclient API changes for 4.51/PH/17
1970 were not taken care of correctly. The code is still untested by me (my
1971 Linux distribution still has 0.3.2 of radiusclient), but it was
1972 contributed by a Radius user.
1974 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1975 expanding the "port" option of the smtp transport.
1977 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1978 while reading a MIME header. Thanks to Tom Hughes for a patch.
1980 PH/24 Include config.h inside local_scan.h so that configuration settings are
1983 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1984 that in an ACL for RCPT (for example), you can examine exactly what was
1987 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1988 commands, but it was not correctly comparing the address with the actual
1989 client host address. Thus, it would show the EHLO address in Received:
1990 header lines when this was not necessary.
1992 PH/27 Added the % operator to ${eval:}.
1994 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1995 it should be ignoring failures (because with -C, for example, it has lost
1996 privilege). It wasn't ignoring creation failures other than "already
1999 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2002 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2003 because an input line was too long, either on its own, or by virtue of
2004 too many continuations, the temporary file was not being removed, and the
2005 return code was incorrect.
2007 PH/31 Missing "BOOL" in function definition in filtertest.c.
2009 PH/32 Applied Sieve patches from the maintainer.
2011 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2013 PH/33 Added "verify = not_blind".
2015 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2016 Local/Makefile (with some defaults set). These are used in built scripts
2017 such as exicyclog, but they have never been used in the exim_install
2018 script (though there are many overriding facilities there). I have
2019 arranged that the exim_install script now takes note of these two
2022 PH/35 Installed configuration files for Dragonfly.
2024 PH/36 When a locally submitted message by a trusted user did not contain a
2025 From: header, and the sender address was obtained from -f or from an SMTP
2026 MAIL command, and the trusted user did not use -F to supply a sender
2027 name, $originator_name was incorrectly used when constructing a From:
2028 header. Furthermore, $originator_name was used for submission mode
2029 messages from external hosts without From: headers in a similar way,
2030 which is clearly wrong.
2032 PH/37 Added control=suppress_local_fixups.
2034 PH/38 When log_selector = +received_sender was set, and the addition of the
2035 sender made the log line's construction buffer exactly full, or one byte
2036 less than full, an overflow happened when the terminating "\n" was
2039 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2040 when the result of a list match is failure because a DNS lookup failed.
2042 PH/40 RM_COMMAND is now used in the building process.
2044 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2045 the "build-* directories that it finds.
2047 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2048 address in a domain literal was a prefix of an interface address.
2050 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2051 when verifying a sender address, unless rewrite_headers is false.
2053 PH/44 Wrote a long comment about why errors_to addresses are verified as
2054 recipients, not senders.
2056 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2057 the ratelimit ACL was added.
2059 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2061 PH/46 Added extra information about PostgreSQL errors to the error string.
2063 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2064 avoiding the need to install special zones in a real server. This is
2065 backwards compatible; if it can't find the fake resolver, it drops back.
2066 Thus, both old and new test suites can be run.
2068 TF/02 Added util/ratelimit.pl
2070 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2071 clock is set back in time.
2073 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2074 Candler <B.Candler@pobox.com>.
2076 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2077 OK for addresses that are the result of redirections.
2079 PH/48 A number of further additions for the benefit of the new test suite,
2080 including a fake gethostbyname() that interfaces to the fake DNS resolver
2083 TF/06 The fix for widen_domains has also been applied to qualify_single and
2084 search_parents which are the other dnslookup options that can cause
2087 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2090 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2092 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2093 service" if it encountered an MX record with an empty target hostname.
2094 The message is now "an MX or SRV record indicated no SMTP service".
2096 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2097 if the DNS of the sending site is misconfigured. This is quite a
2098 common situation. This change restores the behaviour of treating a
2099 helo verification defer as a failure.
2101 PH/51 If self=fail was set on a router, the bounce message did not include the
2102 actual error message.
2108 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2110 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2111 command itself ended in error, the underlying message about the transport
2112 filter timeout was being overwritten with the pipe command error. Now the
2113 underlying error message should be appended to the second error message.
2115 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2117 PH/02 Reduce the amount of output that "make" produces by default. Full output
2118 can still be requested.
2120 PH/03 The warning log line about a condition test deferring for a "warn" verb
2121 was being output only once per connection, rather than after each
2122 occurrence (because it was using the same function as for successful
2123 "warn" verbs). This seems wrong, so I have changed it.
2125 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2126 it should not have, which might have caused a crash in the right
2127 circumstances, but probably never did.
2129 PH/04 Installed a modified version of Tony Finch's patch to make submission
2130 mode fix the return path as well as the Sender: header line, and to
2131 add a /name= option so that you can make the user's friendly name appear
2134 TF/03 Added the control = fakedefer ACL modifier.
2136 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2137 Mark Lowes for thorough testing.
2139 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2141 TK/03 Merged latest SRS patch from Miles Wilton.
2143 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2144 with the definition in sysexits.h (which is #included earlier).
2145 Fortunately, Exim does not actually use EX_OK. The code used to try to
2146 preserve the sysexits.h value, by assumimg that macro definitions were
2147 scanned for macro replacements. I have been disabused of this notion,
2148 so now the code just undefines EX_OK before #including unistd.h.
2150 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2151 in the smtp transport. When a block could not be written in a single
2152 write() function, the timeout was being re-applied to each part-write.
2153 This seems wrong - if the receiver was accepting one byte at a time it
2154 would take for ever. The timeout is now adjusted when this happens. It
2155 doesn't have to be particularly precise.
2157 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2158 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2160 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2161 without a domain if the check to <postmaster@domain> fails.
2163 SC/01 Eximstats: added -xls and the ability to specify output files
2164 (patch written by Frank Heydlauf).
2166 SC/02 Eximstats: use FileHandles for outputing results.
2168 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2170 SC/04 Eximstats: fixed display of large numbers with -nvr option
2172 SC/05 Eximstats: fixed merging of reports with empty tables.
2174 SC/06 Eximstats: added the -include_original_destination flag
2176 SC/07 Eximstats: removed tabs and trailing whitespace.
2178 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2180 TK/06 MBOX spool code: Add real "From " MBOX separator line
2181 so the .eml file is really in mbox format (even though
2182 most programs do not really care). Patch from Alex Miller.
2184 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2185 The latter is generated from $received_to and is only set if the
2186 message has one envelope recipient. SA can use these headers,
2187 obviously out-of-the-box. Patch from Alex Miller.
2189 PH/08 The ${def test on a variable was returning false if the variable's
2190 value was "0", contrary to what the specification has always said!
2191 The result should be true unless the variable is empty.
2193 PH/09 The syntax error of a character other than { following "${if
2194 def:variable_name" (after optional whitespace) was not being diagnosed.
2195 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2196 accidental colon was present, for example, could give incorrect results.
2198 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2199 result is used (not including appendfile, where other changes are about
2202 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2203 This involved changing a lot of size variables from int to off_t. It
2204 should work with maildirs and everything.
2206 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2207 spamd dying while we are connected to it.
2209 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2212 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2213 function for BDB 4.3.
2215 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2217 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2218 the sources for PCRE have been re-arranged and more files are now
2221 PH/15 The code I had for printing potentially long long variables in PH/11
2222 above was not the best (it lost precision). The length of off_t variables
2223 is now inspected at build time, and an appropriate printing format (%ld
2224 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2225 to be "long long int" or "long int". This is needed for the internal
2226 formatting function string_vformat().
2228 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2229 the configuration file to be ":syslog", then the script "guesses" where
2230 the logs files are, rather than using the compiled in default. In our
2231 case the guess is not the same as the compiled default, so the script
2232 suddenly stopped working when I started to use syslog. The patch checks
2233 to see if log_file_path is "". If so, it attempts to read it from exim
2234 with no configuration file to get the compiled in version, before it
2235 falls back to the previous guessing code."
2237 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2238 implementing BATV in an Exim configuration. See NewStuff for the gory
2241 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2242 Makefile that are specific to HP-UX.
2244 PH/18 If the "use_postmaster" option was set for a recipient callout together
2245 with the "random" option, the postmaster address was used as the MAIL
2246 FROM address for the random test, but not for the subsequent recipient
2247 test. It is now used for both.
2249 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2250 patch removes a few documentation additions to RFC 3028, because the
2251 latest draft now contains them. It adds the new en;ascii-case comparator
2252 and a new error check for 8bit text in MIME parts. Comparator and
2253 require names are now matched exactly. I enabled the subaddress
2254 extension, but it is not well tested yet (read: it works for me)."
2256 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2257 rework some of the code of TK/09 above to avoid the hardwired use of
2258 "%lld" and "long long". Replaced the call to snprintf() with a call to
2261 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2262 records point to non-existent hosts", "retry timeout exceeded", and
2263 "retry time not reached for any host after a long failure period".
2265 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2266 experimental DomainKeys support:
2268 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2269 (2) On an error such as an illegally used "control", the wrong name for
2270 the control was given.
2272 These problems did NOT occur unless DomainKeys support was compiled.
2274 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2276 PH/24 Added ${if match_ip condition.
2278 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2279 databases so that it will be absolutely obvious if a crash occurs in the
2280 DB library. This is a regular occurrence (often caused by mis-matched
2283 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2284 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2285 picked up on a user's system that detects such things. There doesn't seem
2286 to be a gcc warning option for this - only an attribute that has to be
2287 put on the function's prototype. It seems that in Fedora Core 4 they have
2288 set this on a number of new functions. No doubt there will be more in due
2291 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2292 specify a transport. However, if an address that was verified by such a
2293 router was the subject of a callout, Exim crashed because it tried to
2294 read the rcpt_include_affixes from the non-existent transport. Now it
2295 just assumes that the setting of that option is false. This bug was
2296 introduced by 4.51/PH/31.
2298 PH/28 Changed -d+all to exclude +memory, because that information is very
2299 rarely of interest, but it makes the output a lot bigger. People tend to
2300 do -d+all out of habit.
2302 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2303 code in os-type was giving problems when libc.so lives in lib64, like on
2306 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2307 aren't the modern standard, and it seems that some systems' include files
2308 don't always have them. Exim was already checking for some of the newer
2309 ones like T_AAAA, and defining it itself. I've added checks for all the
2310 record types that Exim uses.
2312 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2313 not automatically generating a new one, as it is supposed to. This
2314 prevented TLS from working. If the file did exist, but contained invalid
2315 data, a new version was generated, as expected. It was only the case of a
2316 non-existent file that was broken.
2318 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2319 with a change in libdomainkeys > 0.64.
2321 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2322 from DNS. If the selector record carries the flag, it now has
2323 precedence over the domain-wide flag.
2325 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2327 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2328 the use of a port name that isn't defined in /etc/services) occurred, the
2329 message was deferred as in a normal delivery, and thus remained on the
2330 spool, instead of being failed because of the mua_wrapper setting. This
2331 is now fixed, and I tidied up some of the mua_wrapper messages at the
2334 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2335 the messages in a hash of arrays rather than using individual hashes.
2336 This is a bit cleaner and results in dramatic memory savings, albeit
2337 at a slight CPU cost.
2339 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2340 as requested by Marc Sherman.
2342 SC/10 Eximstats: added histograms for user specified patterns as requested
2345 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2347 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2348 fopen() in the content-scanning modules that did not already have it.
2354 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2355 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2357 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2359 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2361 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2362 to test Sieve filters that use "vacation".
2364 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2365 that changes the way the GnuTLS parameters are stored in the cache file.
2366 The new format can be generated externally. For backward compatibility,
2367 if the data in the cache doesn't make sense, Exim assumes it has read an
2368 old-format file, and it generates new data and writes a new file. This
2369 means that you can't go back to an older release without removing the
2372 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2373 work if there are any delivery delays because "one_time" forces the
2374 parent to be marked "delivered", so its unseen clone is never tried
2375 again. For this reason, Exim now forbids the simultaneous setting of
2378 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2379 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2380 entry if you want to know the details. The fix, however, neglected to
2381 consider the case when local delivery batching is involved. The test for
2382 "previously delivered" was not happening when checking to see if an
2383 address could be batched with a previous (undelivered) one; under
2384 certain circumstances this could lead to multiple deliveries to the same
2387 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2388 in its include files, and this causes problems building Exim.
2390 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2391 header_syntax) but Exim was just ignoring anything given after a slash.
2392 In particular, this caused confusion with an attempt to use "verify =
2393 reverse_host_lookup/defer_ok". An error is now given when options are
2394 supplied for verify items that do not have them. (Maybe reverse_host_
2395 lookup should have a defer_ok option, but that's a different point.)
2397 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2398 defined by RFC 821) to 2048, because there were problems with some AUTH
2399 commands, and RFC 1869 says the size should be increased for extended
2400 SMTP commands that take arguments.
2402 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2405 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2406 "unknown" error; now it says that the functionality isn't in the binary.
2408 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2409 an address' error message when a string expansion fails (syntax or
2410 whatever). Otherwise the password may appear in the log. Following change
2411 PH/42 below, there is no longer a chance of it appearing in a bounce
2414 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2416 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2417 of its addresses was ever tried. (Bugzilla bug #2.)
2419 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2420 the result incorrectly in the debug output. (It correctly added a newline
2421 to what was transported.)
2423 TF/01 Added $received_time.
2425 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2426 commented out examples of how to interface to a virus scanner and to
2427 SpamAssassin. Also added commented examples of av_scanner and
2428 spamd_address settings.
2430 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2431 and controls are allowed in which ACLs. There were a couple of minor
2432 errors. Some of the entries in the conditions table (which is a table of
2433 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2434 negation of where the condition IS allowed.
2436 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2438 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2439 header file does not have a version number, so I've had to invent a new
2440 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2441 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2442 radiusclient), but it was contributed by a Radius user.
2444 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2445 files or directories, for OpenSSL.
2447 PH/19 When an Exim process that is running as root has to create an Exim log
2448 file, it does so in a subprocess that runs as exim:exim so as to get the
2449 ownership right at creation (otherwise, other Exim processes might see
2450 the file with the wrong ownership). There was no test for failure of this
2451 fork() call, which would lead to the process getting stuck as it waited
2452 for a non-existent subprocess. Forks do occasionally fail when resources
2453 run out. I reviewed all the other calls to fork(); they all seem to check
2456 PH/20 When checking for unexpected SMTP input at connect time (before writing
2457 the banner), Exim was not dealing correctly with a non-positive return
2458 from the read() function. If the client had disconnected by this time,
2459 the result was a log entry for a synchronization error with an empty
2460 string after "input=" when read() returned zero. If read() returned -1
2461 (an event I could not check), uninitialized data bytes were printed.
2462 There were reports of junk text (parts of files, etc) appearing after
2465 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2467 PH/22 Added support for macro redefinition, and (re)definition in between
2468 driver and ACL definitions.
2470 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2471 forgetting to use the resulting value; it was using the unexpanded value.
2473 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2474 hadn't been configured. The fix is from Juergen Kreileder, who
2475 understands it better than I do:
2477 "Here's what I see happening with three configured cyrus_sasl
2478 authenticators configured (plain, login, cram-md5):
2480 On startup auth_cyrus_sasl_init() gets called for each of these.
2481 This means three calls to sasl_listmech() without a specified mech_list.
2482 => SASL tests which mechs of all available mechs actually work
2483 => three warnings about OTP not working
2484 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2486 With the patch, sasl_listmech() also gets called three times. But now
2487 SASL's mech_list option is set to the server_mech specified in the the
2488 authenticator. Or in other words, the answer from sasl_listmech()
2489 gets limited to just the mech you're testing for (which is different
2491 => the return list contains just 'plain' or 'login', 'cram-md5' or
2492 nothing depending on the value of ob->server_mech.
2494 I've just tested the patch: Authentication still works fine,
2495 unavailable mechs specified in the exim configuration are still
2496 caught, and the auth.log warnings about OTP are gone."
2498 PH/25 When debugging is enabled, the contents of the command line are added
2499 to the debugging output, even when log_selector=+arguments is not
2502 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2503 answer is "GNU", and only if the return is "GNU/something" is the answer
2506 PH/27 $acl_verify_message is now set immediately after the failure of a
2507 verification in an ACL, and so is available in subsequent modifiers. In
2508 particular, the message can be preserved by coding like this:
2510 warn !verify = sender
2511 set acl_m0 = $acl_verify_message
2513 Previously, $acl_verify_message was set only while expanding "message"
2514 and "log_message" when a very denied access.
2516 PH/28 Modified OS/os.c-Linux with
2518 -#ifndef OS_LOAD_AVERAGE
2519 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2521 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2522 nomenclature these days.)
2524 PH/29 Installed patch from the Sieve maintainer that adds the options
2525 sieve_useraddress and sieve_subaddress to the redirect router.
2527 PH/30 In these circumstances:
2528 . Two addresses routed to the same list of hosts;
2529 . First host does not offer TLS;
2530 . First host accepts first address;
2531 . First host gives temporary error to second address;
2532 . Second host offers TLS and a TLS session is established;
2533 . Second host accepts second address.
2534 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2535 and peerdn, if requested) that were in fact used only for the second
2538 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2539 attention to any local part prefix or suffix that was matched by the
2540 router that accepted the address. It now behaves in the same way as it
2541 does for delivery: the affixes are removed from the local part unless
2542 rcpt_include_affixes is set on the transport.
2544 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2545 timeout during the DATA phase of an incoming message.
2547 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2548 applied a patch sanctioned by the Sieve maintainer.
2550 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2551 the uid or gid is negative. A case of a negative gid caused this to be
2552 noticed. The fix allows for either to be negative.
2554 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2555 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2558 PH/36 The change PH/12 above was broken. Fixed it.
2560 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2561 the grounds that routing the same address twice would always produce the
2562 same answer. This might have been true once, but it is certainly no
2563 longer true now. Routing a child address may depend on the previous
2564 routing that produced that child. Some complicated redirection strategies
2565 went wrong when messages had multiple recipients, and made Exim's
2566 behaviour dependent on the order in which the addresses were given.
2568 I have moved the duplicate checking until after the routing is complete.
2569 Exim scans the addresses that are assigned to local and remote
2570 transports, and removes any duplicates. This means that more work will be
2571 done, as duplicates will always all be routed, but duplicates are
2572 presumably rare, so I don't expect this is of any significance.
2574 For deliveries to pipes, files, and autoreplies, the duplicate checking
2575 still happens during the routing process, since they are not going to be
2578 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2579 It corrects a timeout issue with spamd. This is Ian's comment: "The
2580 background is that sometimes spamd either never reads data from a
2581 connection it has accepted, or it never writes response data. The exiscan
2582 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2583 blindly assumes that writes won't block so it may never time out."
2585 PH/39 Allow G after quota size as well as K and M.
2587 PH/40 The value set for $authenticated_id in an authenticator may not contain
2588 binary zeroes or newlines because the value is written to log lines and
2589 to spool files. There was no check on this. Now the value is run through
2590 the string_printing() function so that such characters are converted to
2591 printable escape sequences.
2593 PH/41 $message_linecount is a new variable that contains the total number of
2594 lines in the message. Compare $body_linecount, which is the count for the
2597 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2598 bounce and delay warning messages, except in certain special cases, which
2601 (a) An SMTP error message from a remote host;
2602 (b) A message specified in a :fail: redirection;
2603 (c) A message specified in a "fail" command in a system filter;
2604 (d) A message specified in a FAIL return from the queryprogram router;
2605 (e) A message specified by the cannot_route_message router option.
2607 In these cases only, Exim does include the error details in bounce and
2608 warning messages. There are also a few cases where bland messages such
2609 as "unrouteable address" or "local delivery error" are given.
2611 PH/43 $value is now also set for the "else" part of a ${run expansion.
2613 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2614 being worked on, but at least Exim now implements the latest version to
2617 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2618 process to complete was treated as a delivery failure, a timeout while
2619 writing the message to the pipe was logged, but erroneously treated as a
2620 successful delivery. Such timeouts include transport filter timeouts. For
2621 consistency with the overall process timeout, these timeouts are now
2622 treated as errors, giving rise to delivery failures by default. However,
2623 there is now a new Boolean option for the pipe transport called
2624 timeout_defer, which, if set TRUE, converts the failures into defers for
2625 both kinds of timeout. A transport filter timeout is now identified in
2628 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2629 systems where "make" and "gmake" are different, calling "gmake" at top
2630 level broke things. I've arranged for the value of $(MAKE) to be passed
2631 from the Makefile to this script so that it can call the same version of
2635 A note about Exim versions 4.44 and 4.50
2636 ----------------------------------------
2638 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2639 changes of various kinds. As a consequence, a big documentation update was
2640 needed. This delayed the release for rather longer than seemed good, especially
2641 in the light of a couple of (minor) security issues. Therefore, the changes
2642 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2643 release. So 4.44 and 4.50 are in effect two different branches that both start
2646 I have left the 4.50 change log unchanged; it contains all the changes since
2647 4.43. The change log for 4.44 is below; many of its items are identical to
2648 those for 4.50. This seems to be the most sensible way to preserve the
2649 historical information.
2655 1. Minor wording change to the doc/README.SIEVE file.
2657 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2658 computation of the current number of files was incorrect.
2660 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2661 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2664 4. Give more explanation in the error message when the command for a transport
2665 filter fails to execute.
2667 5. There are several places where Exim runs a non-Exim command in a
2668 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2669 was being done only for the command run by the queryprogram router. It is
2670 now done for all such subprocesses. The other cases are: ${run, transport
2671 filters, and the commands run by the lmtp and pipe transports.
2673 6. Added CONFIGURE_GROUP build-time option.
2675 7. Some older OS have a limit of 256 on the maximum number of file
2676 descriptors. Exim was using setrlimit() to set 1000 as a large value
2677 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2678 systems. I've change it so that if it can't get 1000, it tries for 256.
2680 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2681 was an oversight, and furthermore, ever since the addition of extra
2682 controls (e.g. 4.43/32), the checks on when to allow different forms of
2683 "control" were broken. There should now be diagnostics for all cases when a
2684 control that does not make sense is encountered.
2686 9. Added the /retain_sender option to "control=submission".
2688 10. $recipients is now available in the predata ACL (oversight).
2690 11. Tidy the search cache before the fork to do a delivery from a message
2691 received from the command line. Otherwise the child will trigger a lookup
2692 failure and thereby defer the delivery if it tries to use (for example) a
2693 cached ldap connection that the parent has called unbind on.
2695 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2696 of $address_data from the recipient verification was clobbered by the
2697 sender verification.
2699 13. The value of address_data from a sender verification is now available in
2700 $sender_address_data in subsequent conditions in the ACL statement.
2702 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2704 15. Added a new option "connect=<time>" to callout options, to set a different
2707 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2708 was its contents. (It was OK if the option was not defined at all.)
2710 17. A "Completed" log line is now written for messages that are removed from
2711 the spool by the -Mrm option.
2713 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2714 information about exactly what failed.
2716 19. Added -dd to debug only the daemon process.
2718 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2719 handles timeouts, both on the server side and network timeouts. Renamed the
2720 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2722 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2723 It is now set to "smtps".
2725 22. $host_address is now set to the target address during the checking of
2726 ignore_target_hosts.
2728 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2729 being passed; this would have caused $sender_host_name to have been used if
2730 matching the list had actually called for a host name (not very likely,
2731 since this list is usually IP addresses). A host name is now passed as
2734 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2735 code that specifies a non-restarting handler (typically sigaction() in
2736 modern systems) in an attempt to fix a rare and obscure crash bug.
2738 25. Narrowed the window for a race in the daemon that could cause it to ignore
2739 SIGCHLD signals. This is not a major problem, because they are used only to
2740 wake it up if nothing else does.
2742 26. A malformed maildirsize file could cause Exim to calculate negative values
2743 for the mailbox size or file count. Odd effects could occur as a result.
2744 The maildirsize information is now recalculated if the size or filecount
2747 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2748 support for a long time. Removed HAVE_SYS_VFS_H.
2750 28. Installed the latest version of exipick from John Jetmore.
2752 29. In an address list, if the pattern was not a regular expression, an empty
2753 subject address (from a bounce message) matched only if the pattern was an
2754 empty string. Non-empty patterns were not even tested. This was the wrong
2755 because it is perfectly reasonable to use an empty address as part of a
2756 database query. An empty address is now tested by patterns that are
2757 lookups. However, all the other forms of pattern expect the subject to
2758 contain a local part and a domain, and therefore, for them, an empty
2759 address still always fails if the pattern is not itself empty.
2761 30. Exim went into a mad DNS loop when attempting to do a callout where the
2762 host was specified on an smtp transport, and looking it up yielded more
2763 than one IP address.
2765 31. Re-factored the code for checking spool and log partition space into a
2766 function that finds that data and another that does the check. The former
2767 is then used to implement four new variables: $spool_space, $log_space,
2768 $spool_inodes, and $log_inodes.
2770 32. The RFC2047 encoding function was originally intended for short strings
2771 such as real names; it was not keeping to the 75-character limit for
2772 encoded words that the RFC imposes. It now respects the limit, and
2773 generates multiple encoded words if necessary. To be on the safe side, I
2774 have increased the buffer size for the ${rfc2047: expansion operator from
2777 33. It is now permitted to omit both strings after an "if" condition; if the
2778 condition is true, the result is "true". As before, when the second string
2779 is omitted, a false condition yields an empty string. This makes it less
2780 cumbersome to write custom ACL and router conditions.
2782 34. Failure to deliver a bounce message always caused it to be frozen, even if
2783 there was an errors_to setting on the router. The errors_to setting is now
2786 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2787 canonical form (fully expanded) before being placed in
2788 $sender_host_address.
2790 36. The table in the code that translates DNS record types into text (T_A to
2791 "A" for instance) was missing entries for NS and CNAME. It is just possible
2792 that this could have caused confusion if both these types were looked up
2793 for the same domain, because the text type is used as part of Exim's
2794 per-process caching. But the chance of anyone hitting this buglet seems
2797 37. The dnsdb lookup has been extended in a number of ways.
2799 (1) There is a new type, "zns", which walks up the domain tree until it
2800 finds some nameserver records. It should be used with care.
2802 (2) There is a new type, "mxh", which is like "mx" except that it returns
2803 just the host names, not the priorities.
2805 (3) It is now possible to give a list of domains (or IP addresses) to be
2806 looked up. The behaviour when one of the lookups defers can be
2807 controlled by a keyword.
2809 (4) It is now possible to specify the separator character for use when
2810 multiple records are returned.
2812 38. The dnslists ACL condition has been extended: it is now possible to supply
2813 a list of IP addresses and/or domains to be looked up in a particular DNS
2816 39. Added log_selector=+queue_time_overall.
2818 40. When running the queue in the test harness, wait just a tad after forking a
2819 delivery process, to get repeatability of debugging output.
2821 41. Include certificate and key file names in error message when GnuTLS fails
2822 to set them up, because the GnuTLS error message doesn't include the name
2823 of the failing file when there is a problem reading it.
2825 42. Allow both -bf and -bF in the same test run.
2827 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2829 44. The "Exiscan patch" is now merged into the mainline Exim source.
2831 45. Sometimes the final signoff response after QUIT could fail to get
2832 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2833 < 0 before doing a fflush(). This bug looks as though it goes back to the
2834 introduction of TLS in release 3.20, but "sometimes" must have been rare
2835 because the tests only now provoked it.
2837 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2838 (this can affect the format of dates).
2840 47. exim_tidydb, when checking for the continued existence of a message for
2841 which it has found a message-specific retry record, was not finding
2842 messages that were in split spool directories. Consequently, it was
2843 deleting retry records that should have stayed in existence.
2845 48. Steve fixed some bugs in eximstats.
2847 49. The SPA authentication driver was not abandoning authentication and moving
2848 on to the next authenticator when an expansion was forced to fail,
2849 contradicting the general specification for all authenticators. Instead it
2850 was generating a temporary error. It now behaves as specified.
2852 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2853 (the order specifies the preference for clients). The order is now AES256,
2854 AES128, 3DES, ARCFOUR128.
2856 51. Small patch to Sieve code - explicitly set From: when generating an
2859 52. Exim crashed if a remote delivery caused a very long error message to be
2860 recorded - for instance if somebody sent an entire SpamAssassin report back
2861 as a large number of 550 error lines. This bug was coincidentally fixed by
2862 increasing the size of one of Exim's internal buffers (big_buffer) that
2863 happened as part of the Exiscan merge. However, to be on the safe side, I
2864 have made the code more robust (and fixed the comments that describe what
2867 53. Now that there can be additional text after "Completed" in log lines (if
2868 the queue_time_overall log selector is set), a one-byte patch to exigrep
2869 was needed to allow it to recognize "Completed" as not the last thing in
2872 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2873 patch that reportedly fixes this has been added. I am not expert enough to
2874 create a test for it. This is what the patch creator wrote:
2876 "I found a little strange behaviour of ldap code when working with
2877 Windows 2003 AD Domain, where users was placed in more than one
2878 Organization Units. When I tried to give exim partial DN, the exit code
2879 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2880 But simultaneously result of request was absolutely normal ldap result,
2881 so I produce this patch..."
2883 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2884 REFERENCE, so I have modified the code to exclude the patch when that macro
2887 55. Some experimental protocols are using DNS PTR records for new purposes. The
2888 keys for these records are domain names, not reversed IP addresses. The
2889 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2890 leaves it alone. Component reversal etc. now happens only for IP addresses.
2893 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2895 57. Double the size of the debug message buffer (to 2048) so that more of very
2896 long debug lines gets shown.
2898 58. The exicyclog utility now does better if the number of log files to keep
2899 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2901 59. Two changes related to the smtp_active_hostname option:
2903 (1) $smtp_active_hostname is now available as a variable.
2904 (2) The default for smtp_banner uses $smtp_active_hostname instead
2905 of $primary_hostname.
2907 60. The host_aton() function is supposed to be passed a string that is known
2908 to be a valid IP address. However, in the case of IPv6 addresses, it was
2909 not checking this. This is a hostage to fortune. Exim now panics and dies
2910 if the condition is not met. A case was found where this could be provoked
2911 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2912 components; fortuitously, this particular loophole had already been fixed
2913 by change 4.50/55 above.
2915 If there are any other similar loopholes, the new check in host_aton()
2916 itself should stop them being exploited. The report I received stated that
2917 data on the command line could provoke the exploit when Exim was running as
2918 exim, but did not say which command line option was involved. All I could
2919 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2920 running as the user.
2923 61. There was a buffer overflow vulnerability in the SPA authentication code
2924 (which came originally from the Samba project). I have added a test to the
2925 spa_base64_to_bits() function which I hope fixes it.
2928 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2929 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2931 63. The daemon start-up calls getloadavg() while still root for those OS that
2932 need the first call to be done as root, but it missed one case: when
2933 deliver_queue_load_max is set with deliver_drop_privilege. This is
2934 necessary for the benefit of the queue runner, because there is no re-exec
2935 when deliver_drop_privilege is set.
2937 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2938 This has been fixed.
2940 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2941 was in use, was not putting the data itself into the right store pool;
2942 consequently, it could be overwritten for a subsequent message in the same
2943 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2946 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2948 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2949 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2950 the function were treating the return as a boolean value, which happened to
2951 work because 0=false and not-0=true, but is not correct code.
2953 68. The host_aton() function was not handling scoped IPv6 addresses (those
2954 with, for example, "%eth0" on the end) correctly.
2956 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2957 negated items (that is, ~something) in unsigned ints. Some compilers
2958 apparently mutter when there is no cast.
2960 70. If an address verification called from an ACL failed, and did not produce a
2961 user-specific message (i.e. there was only a "system" message), nothing was
2962 put in $acl_verify_message. In this situation, it now puts the system
2965 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2966 session; change 4.31/43 added the unwanted input to the log line - except
2967 that it did not do this in the start of session case. It now does.
2969 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2970 This is wrong and can cause the other end to generate a synchronization
2971 error if it is another Exim or anything else that does the synchronization
2972 check. A QUIT command is no longer sent after a timeout.
2974 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2975 during host lookups.
2977 74. The defer_ok option of callout verification was not working if it was used
2978 when verifying addresses in header lines, that is, for this case:
2980 verify = header_sender/callout=defer_ok
2982 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2983 those file descriptors could be used for SMTP connections. If anything
2984 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2985 could be sent to the SMTP client, causing chaos. The daemon now opens
2986 stdin, stdout, and stderr to /dev/null when it puts itself into the
2989 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2990 log by default. The user can override this with suitable Perl magic.
2992 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2993 the log message when discard triggers, was not working for the DATA ACL or
2994 for the non-SMTP ACL.
2996 78. Error message wording change in sieve.c.
2998 79. If smtp_accept_max_per_host was set, the number of connections could be
2999 restricted to fewer than expected, because the daemon was trying to set up
3000 a new connection before checking whether the processes handling previous
3001 connections had finished. The check for completed processes is now done
3002 earlier. On busy systems, this bug wouldn't be noticed because something
3003 else would have woken the daemon, and it would have reaped the completed
3006 80. If a message was submitted locally by a user whose login name contained one
3007 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3008 It caused a spool format error. I have fixed the spool reading code. A
3009 related problem was that the "from" clause in the Received: line became
3010 illegal because of the space(s). It is now covered by ${quote_local_part.
3012 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3015 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3017 83. Patch from Sieve maintainer to fix unterminated string problem in
3018 "vacation" handling.
3020 84. Some minor changes to the Linux configuration files to help with other
3021 OS variants using glibc.
3023 85. One more patch for Sieve to update vacation handling to latest spec.
3026 ----------------------------------------------------
3027 See the note above about the 4.44 and 4.50 releases.
3028 ----------------------------------------------------
3034 1. Change 4.43/35 introduced a bug that caused file counts to be
3035 incorrectly computed when quota_filecount was set in an appendfile
3038 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3039 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3042 3. Give more explanation in the error message when the command for a transport
3043 filter fails to execute.
3045 4. There are several places where Exim runs a non-Exim command in a
3046 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3047 was being done only for the command run by the queryprogram router. It is
3048 now done for all such subprocesses. The other cases are: ${run, transport
3049 filters, and the commands run by the lmtp and pipe transports.
3051 5. Some older OS have a limit of 256 on the maximum number of file
3052 descriptors. Exim was using setrlimit() to set 1000 as a large value
3053 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3054 systems. I've change it so that if it can't get 1000, it tries for 256.
3056 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3057 was an oversight, and furthermore, ever since the addition of extra
3058 controls (e.g. 4.43/32), the checks on when to allow different forms of
3059 "control" were broken. There should now be diagnostics for all cases when a
3060 control that does not make sense is encountered.
3062 7. $recipients is now available in the predata ACL (oversight).
3064 8. Tidy the search cache before the fork to do a delivery from a message
3065 received from the command line. Otherwise the child will trigger a lookup
3066 failure and thereby defer the delivery if it tries to use (for example) a
3067 cached ldap connection that the parent has called unbind on.
3069 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3070 of $address_data from the recipient verification was clobbered by the
3071 sender verification.
3073 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3074 was its contents. (It was OK if the option was not defined at all.)
3076 11. A "Completed" log line is now written for messages that are removed from
3077 the spool by the -Mrm option.
3079 12. $host_address is now set to the target address during the checking of
3080 ignore_target_hosts.
3082 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3083 being passed; this would have caused $sender_host_name to have been used if
3084 matching the list had actually called for a host name (not very likely,
3085 since this list is usually IP addresses). A host name is now passed as
3088 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3089 code that specifies a non-restarting handler (typically sigaction() in
3090 modern systems) in an attempt to fix a rare and obscure crash bug.
3092 15. Narrowed the window for a race in the daemon that could cause it to ignore
3093 SIGCHLD signals. This is not a major problem, because they are used only to
3094 wake it up if nothing else does.
3096 16. A malformed maildirsize file could cause Exim to calculate negative values
3097 for the mailbox size or file count. Odd effects could occur as a result.
3098 The maildirsize information is now recalculated if the size or filecount
3101 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3102 support for a long time. Removed HAVE_SYS_VFS_H.
3104 18. Updated exipick to current release from John Jetmore.
3106 19. Allow an empty sender to be matched against a lookup in an address list.
3107 Previously the only cases considered were a regular expression, or an
3110 20. Exim went into a mad DNS lookup loop when doing a callout where the
3111 host was specified on the transport, if the DNS lookup yielded more than
3114 21. The RFC2047 encoding function was originally intended for short strings
3115 such as real names; it was not keeping to the 75-character limit for
3116 encoded words that the RFC imposes. It now respects the limit, and
3117 generates multiple encoded words if necessary. To be on the safe side, I
3118 have increased the buffer size for the ${rfc2047: expansion operator from
3121 22. Failure to deliver a bounce message always caused it to be frozen, even if
3122 there was an errors_to setting on the router. The errors_to setting is now
3125 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3126 canonical form (fully expanded) before being placed in
3127 $sender_host_address.
3129 24. Updated eximstats to version 1.33
3131 25. Include certificate and key file names in error message when GnuTLS fails
3132 to set them up, because the GnuTLS error message doesn't include the name
3133 of the failing file when there is a problem reading it.
3135 26. Expand error message when OpenSSL has problems setting up cert/key files.
3138 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3139 (this can affect the format of dates).
3141 28. exim_tidydb, when checking for the continued existence of a message for
3142 which it has found a message-specific retry record, was not finding
3143 messages that were in split spool directories. Consequently, it was
3144 deleting retry records that should have stayed in existence.
3146 29. eximstats updated to version 1.35
3147 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3148 1.35 - bugfix such that pie charts by volume are generated correctly
3150 30. The SPA authentication driver was not abandoning authentication and moving
3151 on to the next authenticator when an expansion was forced to fail,
3152 contradicting the general specification for all authenticators. Instead it
3153 was generating a temporary error. It now behaves as specified.
3155 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3156 (the order specifies the preference for clients). The order is now AES256,
3157 AES128, 3DES, ARCFOUR128.
3159 31. Small patch to Sieve code - explicitly set From: when generating an
3162 32. Exim crashed if a remote delivery caused a very long error message to be
3163 recorded - for instance if somebody sent an entire SpamAssassin report back
3164 as a large number of 550 error lines. This bug was coincidentally fixed by
3165 increasing the size of one of Exim's internal buffers (big_buffer) that
3166 happened as part of the Exiscan merge. However, to be on the safe side, I
3167 have made the code more robust (and fixed the comments that describe what
3170 33. Some experimental protocols are using DNS PTR records for new purposes. The
3171 keys for these records are domain names, not reversed IP addresses. The
3172 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3173 leaves it alone. Component reversal etc. now happens only for IP addresses.
3176 34. The host_aton() function is supposed to be passed a string that is known
3177 to be a valid IP address. However, in the case of IPv6 addresses, it was
3178 not checking this. This is a hostage to fortune. Exim now panics and dies
3179 if the condition is not met. A case was found where this could be provoked
3180 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3181 components; fortuitously, this particular loophole had already been fixed
3182 by change 4.50/55 or 4.44/33 above.
3184 If there are any other similar loopholes, the new check in host_aton()
3185 itself should stop them being exploited. The report I received stated that
3186 data on the command line could provoke the exploit when Exim was running as
3187 exim, but did not say which command line option was involved. All I could
3188 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3189 running as the user.
3192 35. There was a buffer overflow vulnerability in the SPA authentication code
3193 (which came originally from the Samba project). I have added a test to the
3194 spa_base64_to_bits() function which I hope fixes it.
3197 36. The daemon start-up calls getloadavg() while still root for those OS that
3198 need the first call to be done as root, but it missed one case: when
3199 deliver_queue_load_max is set with deliver_drop_privilege. This is
3200 necessary for the benefit of the queue runner, because there is no re-exec
3201 when deliver_drop_privilege is set.
3203 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3204 was in use, was not putting the data itself into the right store pool;
3205 consequently, it could be overwritten for a subsequent message in the same
3206 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3209 38. Sometimes the final signoff response after QUIT could fail to get
3210 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3211 < 0 before doing a fflush(). This bug looks as though it goes back to the
3212 introduction of TLS in release 3.20, but "sometimes" must have been rare
3213 because the tests only now provoked it.
3219 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3220 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3221 Now it returns an integer. A number of calls were still expecting a T/F
3222 return. Fortuitously, in all cases, the tests worked in OK situations,
3223 which is the norm. However, things would have gone wrong on any write
3224 failures on the smtp file descriptor. This function is used when sending
3225 messages over SMTP and also when doing verify callouts.
3227 2. When Exim is called to do synchronous delivery of a locally submitted
3228 message (the -odf or -odi options), it no longer closes stderr before doing
3231 3. Implemented the mua_wrapper option.
3233 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3235 5. Implemented the functions header_remove(), header_testname(),
3236 header_add_at_position(), and receive_remove_recipient(), and exported them
3239 6. If an ACL "warn" statement specified the addition of headers, Exim already
3240 inserted X-ACL-Warn: at the start if there was no header name. However, it
3241 was not making this test for the second and subsequent header lines if
3242 there were newlines in the string. This meant that an invalid header could
3243 be inserted if Exim was badly configured.
3245 7. Allow an ACL "warn" statement to add header lines at the start or after all
3246 the Received: headers, as well as at the end.
3248 8. Added the rcpt_4xx retry error code.
3250 9. Added postmaster_mailfrom=xxx to callout verification option.
3252 10. Added mailfrom=xxxx to the callout verification option, for verify=
3255 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3256 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3258 12. Inserted some casts to stop certain compilers warning when using pointer
3259 differences as field lengths or precisions in printf-type calls (mostly
3260 affecting debugging statements).
3262 13. Added optional readline() support for -be (dynamically loaded).
3264 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3265 same clock tick as a message's arrival, so that its received time was the
3266 same as the "first fail" time on the retry record, and that message
3267 remained on the queue past the ultimate address timeout, every queue runner
3268 would try a delivery (because it was past the ultimate address timeout) but
3269 after another failure, the ultimate address timeout, which should have then
3270 bounced the address, did not kick in. This was a "< instead of <=" error;
3271 in most cases the first failure would have been in the next clock tick
3272 after the received time, and all would be well.
3274 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3275 being recognized when the domain list was tested by the match_domain
3276 condition in an expansion string.
3278 16. Added the ${str2b64: operator.
3280 17. Exim was always calling setrlimit() to set a large limit for the number of
3281 processes, without checking whether the existing limit was already
3282 adequate. (It did check for the limit on file descriptors.) Furthermore,
3283 errors from getrlimit() and setrlimit() were being ignored. Now they are
3284 logged to the main and panic logs, but Exim does carry on, to try to do its
3285 job under whatever limits there are.
3287 18. Imported PCRE 5.0.
3289 19. Trivial typo in log message " temporarily refused connection" (the leading
3292 20. If the log selector return_path_on_delivery was set and an address was
3293 redirected to /dev/null, the delivery process crashed because it assumed
3294 that a return path would always be set for a "successful" delivery. In this
3295 case, the whole delivery is bypassed as an optimization, and therefore no
3298 21. Internal re-arrangement: the function for sending a challenge and reading
3299 a response while authentication was assuming a zero-terminated challenge
3300 string. It's now changed to take a pointer and a length, to allow for
3301 binary data in such strings.
3303 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3305 23. Exim was not respecting finduser_retries when seeking the login of the
3306 uid under which it was called; it was always trying 10 times. (The default
3307 setting of finduser_retries is zero.) Also, it was sleeping after the final
3308 failure, which is pointless.
3310 24. Implemented tls_on_connect_ports.
3312 25. Implemented acl_smtp_predata.
3314 26. If the domain in control=submission is set empty, Exim assumes that the
3315 authenticated id is a complete email address when it generates From: or
3316 Sender: header lines.
3318 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3319 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3320 chown and chgrp in /bin and hostname in /usr/bin.
3322 28. Exim was keeping the "process log" file open after each use, just as it
3323 does for the main log. This opens the possibility of it remaining open for
3324 long periods when the USR1 signal hits a daemon. Occasional processlog
3325 errors were reported, that could have been caused by this. Anyway, it seems
3326 much more sensible not to leave this file open at all, so that is what now
3329 29. The long-running daemon process does not normally write to the log once it
3330 has entered its main loop, and it closes the log before doing so. This is
3331 so that log files can straightforwardly be renamed and moved. However,
3332 there are a couple of unusual error situations where the daemon does write
3333 log entries, and I had neglected to close the log afterwards.
3335 30. The text of an SMTP error response that was received during a remote
3336 delivery was being truncated at 512 bytes. This is too short for some of
3337 the long messages that one sometimes sees. I've increased the limit to
3340 31. It is now possible to make retry rules that apply only when a message has a
3341 specific sender, in particular, an empty sender.
3343 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3344 it possible to be selective about when SMTP synchronization is enforced.
3346 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3348 32. Implemented hosts_connection_nolog.
3350 33. Added an ACL for QUIT.
3352 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3355 35. Added mailbox_size and mailbox_filecount to appendfile.
3357 36. Added control = no_multiline_responses to ACLs.
3359 37. There was a bug in the logic of the code that waits for the clock to tick
3360 in the case where the clock went backwards by a substantial amount such
3361 that the microsecond fraction of "now" was more than the microsecond
3362 fraction of "then" (but the whole seconds number was less).
3364 38. Added support for the libradius Radius client library this is found on
3365 FreeBSD (previously only the radiusclient library was supported).
3371 1. When certain lookups returned multiple values in the form name=value, the
3372 quoting of the values was not always being done properly. Specifically:
3373 (a) If the value started with a double quote, but contained no whitespace,
3375 (b) If the value contained whitespace other than a space character (i.e.
3376 tabs or newlines or carriage returns) it was not quoted.
3377 This fix has been applied to the mysql and pgsql lookups by writing a
3378 separate quoting function and calling it from the lookup code. The fix
3379 should probably also be applied to nisplus, ibase and oracle lookups, but
3380 since I cannot test any of those, I have not disturbed their existing code.
3382 2. A hit in the callout cache for a specific address caused a log line with no
3383 reason for rejecting RCPT. Now it says "Previous (cached) callout
3384 verification failure".
3386 3. There was an off-by-one bug in the queryprogram router. An over-long
3387 return line was truncated at 256 instead of 255 characters, thereby
3388 overflowing its buffer with the terminating zero. As well as fixing this, I
3389 have increased the buffer size to 1024 (and made a note to document this).
3391 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3392 when Exim is waiting for an SMTP response from a remote server, Exim
3393 restarts its select() call on the socket, thereby resetting its timeout.
3394 This is not a problem when such interrupts are rare. Somebody set up a cron
3395 job to run exiwhat every 2 minutes, which is less than the normal select()
3396 timeout (5 or 10 minutes). This meant that the select() timeout never
3397 kicked in because it was always reset. I have fixed this by comparing the
3398 time when an interrupt arrives with the time at the start of the first call
3399 to select(). If more time than the timeout has elapsed, the interrupt is
3400 treated as a timeout.
3402 5. Some internal re-factoring in preparation for the addition of Sieve
3403 extensions (by MH). In particular, the "personal" test is moved to a
3404 separate function, and given an option for scanning Cc: and Bcc: (which is
3405 not set for Exim filters).
3407 6. When Exim created an email address using the login of the caller as the
3408 local part (e.g. when creating a From: or Sender: header line), it was not
3409 quoting the local part when it contained special characters such as @.
3411 7. Installed new OpenBSD configuration files.
3413 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3414 try to make them clearer.
3416 9. Callout options, other than the timeout value, were being ignored when
3417 verifying sender addresses in header lines. For example, when using
3419 verify = header_sender/callout=no_cache
3421 the cache was (incorrectly) being used.
3423 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3424 only the Cygwin environment.
3426 11. When return_path_on_delivery was set as a log selector, if different remote
3427 addresses in the same message used different return paths and parallel
3428 remote delivery occurred, the wrong values would sometimes be logged.
3429 (Whenever a remote delivery process finished, the return path value from
3430 the most recently started remote delivery process was logged.)
3432 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3433 lines when AUTH and/or TLS are in use. This is the "received protocol"
3434 field. Exim used to use "asmtp" for authenticated SMTP, without any
3435 indication (in the protocol name) for TLS use. Now it follows the RFC and
3436 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3437 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3438 names appear in log lines as well as in Received: header lines.
3440 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3441 extensions, and comparison tests, and to fix some bugs.
3443 14. Changes to the "personal" filter test:
3445 (1) The test was buggy in that it was just doing the equivalent of
3446 "contains" tests on header lines. For example, if a user's address was
3447 anne@some.where, the "personal" test would incorrectly be true for
3449 To: susanne@some.where
3451 This test is now done by extracting each address from the header in turn,
3452 and checking the entire address. Other tests that are part of "personal"
3453 are now done using regular expressions (for example, to check local parts
3454 of addresses in From: header lines).
3456 (2) The list of non-personal local parts in From: addresses has been
3457 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3458 taken from the Sieve specification recommendations.
3460 (3) If the message contains any header line starting with "List-" it is
3461 treated as non-personal.
3463 (4) The test for "circular" in the Subject: header line has been removed
3464 because it now seems ill-conceived.
3466 15. Minor typos in src/EDITME comments corrected.
3468 16. Installed latest exipick from John Jetmore.
3470 17. If headers_add on a router specified a text string that was too long for
3471 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3472 of string_sprintf() is now avoided.
3474 18. $message_body_size was not set (it was always zero) when running the DATA
3475 ACL and the local_scan() function.
3477 19. For the "mail" command in an Exim filter, no default was being set for
3478 the once_repeat time, causing a random time value to be used if "once" was
3479 specified. (If the value happened to be <= 0, no repeat happened.) The
3480 default is now 0s, meaning "never repeat". The "vacation" command was OK
3481 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3482 (I found it when inspecting the code).
3484 20. There is now an overall timeout for performing a callout verification. It
3485 defaults to 4 times the callout timeout, which applies to individual SMTP
3486 commands during the callout. The overall timeout applies when there is more
3487 than one host that can be tried. The timeout is checked before trying the
3488 next host. This prevents very long delays if there are a large number of
3489 hosts and all are timing out (e.g. when the network connections are timing
3490 out). The value of the overall timeout can be changed by specifying an
3491 additional sub-option for "callout", called "maxwait". For example:
3493 verify = sender/callout=5s,maxwait=20s
3495 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3496 to the end before writing, but this should make it even safer).
3498 22. Exim was forgetting that it had advertised PIPELINING for the second and
3499 subsequent messages on an SMTP connection. It was also not resetting its
3500 memory on STARTTLS and an internal HELO.
3502 23. When Exim logs an SMTP synchronization error within a session, it now
3503 records whether PIPELINING has been advertised or not.
3505 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3506 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3507 rather than long int.
3509 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3511 26. Added the never_mail option to autoreply.
3517 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3518 crash if the getsockname() call failed; this can happen if a connection is
3519 closed very soon after it is established. The problem was simply in the
3520 order in which certain operations were done, causing Exim to try to write
3521 to the SMTP stream before it had set up the file descriptor. The bug has
3522 been fixed by making things happen in the correct order.
3528 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3529 before the connection was closed, thus losing the rejection response.
3531 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3532 some early Solaris releases, but causes trouble in current releases where
3533 socklen_t is defined.
3535 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3538 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3541 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3542 that releasing the top of it at the end releases what was used for sub-
3543 expansions (unless the block got too big). However, discard this block if
3544 the first thing is a variable or header, so that we can use its block when
3545 it is dynamic (useful for very large $message_headers, for example).
3547 6. Lookups now cache *every* query, not just the most recent. A new, separate
3548 store pool is used for this. It can be recovered when all lookup caches are
3549 flushed. Lookups now release memory at the end of their result strings.
3550 This has involved some general refactoring of the lookup sources.
3552 7. Some code has been added to the store_xxx() functions to reduce the amount
3553 of flapping under certain conditions.
3555 8. log_incoming_interface used to affect only the <= reception log lines. Now
3556 it causes the local interface and port to be added to several more SMTP log
3557 lines, for example "SMTP connection from", and rejection lines.
3559 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3561 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3563 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3564 could be overwritten at the end of the current message (or the start of a
3565 new message if it was set in a HELO ACL). The value is now preserved for
3566 the duration of the SMTP connection.
3568 12. If a transport had a headers_rewrite setting, and a matching header line
3569 contained an unqualified address, that address was qualified, even if it
3570 did not match any rewriting rules. The underlying bug was that the values
3571 of the flags that permit the existence of unqualified sender and recipient
3572 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3573 non-local messages, and by -bnq for local messages) were not being
3574 preserved with the message after it was received.
3576 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3577 "next input=" as part of the text comprising the host identity instead of
3578 the correct text. The code was using the same buffer for two different
3579 strings. However, depending on which order the printing function evaluated
3580 its arguments, the bug did not always show up. Under Linux, for example, my
3581 test suite worked just fine.
3583 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3584 doesn't work with some older versions of Perl. It has been changed to "my",
3585 which in any case is probably the better facility to use.
3587 15. A really picky compiler found some instances of statements for creating
3588 error messages that either had too many or two few arguments for the format
3591 16. The size of the buffer for calls to the DNS resolver has been increased
3592 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3593 for addresses that have a lot of PTR records. This alleviates a problem; it
3594 does not fully solve it.
3596 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3597 buffer now truncates the list and logs the incident, which is the same
3598 action as happens when Exim is looking up a host name and its aliases.
3599 Previously in this situation something unpredictable would happen;
3600 sometimes it was "internal error: store_reset failed".
3602 18. If a server dropped the connection unexpectedly when an Exim client was
3603 using GnuTLS and trying to read a response, the client delivery process
3604 crashed while trying to generate an error log message.
3606 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3607 string, for example:
3609 warn message = H1: something\nH2: something
3611 the text was added as a single header line from Exim's point of view
3612 though it ended up OK in the delivered message. However, searching for the
3613 second and subsequent header lines using $h_h2: did not work. This has been
3614 fixed. Similarly, if a system filter added multiple headers in this way,
3615 the routers could not see them.
3617 20. Expanded the error message when iplsearch is called with an invalid key to
3618 suggest using net-iplsearch in a host list.
3620 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3621 ACLs are no longer actually imposed (and a message to that effect is
3624 22. If a "gecos" field in a passwd entry contained escaped characters, in
3625 particular, if it contained a \" sequence, Exim got it wrong when building
3626 a From: or a Sender: header from that name. A second bug also caused
3627 incorrect handling when an unquoted " was present following a character
3628 that needed quoting.
3630 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3631 was not being matched caselessly.
3633 24. Arranged for all hyphens in the exim.8 source to be escaped with
3636 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3637 properly. Recipient callout cache records were still being keyed to include
3638 the sender, even when use_sender was set false. This led to far more
3639 callouts that were necessary. The sender is no longer included in the key
3640 when use_sender is false.
3642 26. Added "control = submission" modifier to ACLs.
3644 27. Added the ${base62d: operator to decode base 62 numbers.
3646 28. dnsdb lookups can now access SRV records.
3648 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3649 the configuration file.
3651 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3652 (-v) mode. This makes the output for a verbose queue run more intelligible.
3654 31. Added a use_postmaster feature to recipient callouts.
3656 32. Added the $body_zerocount variable, containing the number of binary zero
3657 bytes in the message body.
3659 33. The time of last modification of the "new" subdirectory is now used as the
3660 "mailbox time last read" when there is a quota error for a maildir
3663 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3665 35. Added +ignore_unknown as a special item in host lists.
3667 36. Code for decoding IPv6 addresses in host lists is now included, even if
3668 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3669 address was recognized as an IP address, but was then not correctly decoded
3670 into binary, causing unexpected and incorrect effects when compared with
3677 1. Very minor rewording of debugging text in manualroute to say "list of
3678 hosts" instead of "hostlist".
3680 2. If verify=header_syntax was set, and a header line with an unqualified
3681 address (no domain) and a large number of spaces between the end of the
3682 name and the colon was received, the reception process suffered a buffer
3683 overflow, and (when I tested it) crashed. This was caused by some obsolete
3684 code that should have been removed. The fix is to remove it!
3686 3. When running in the test harness, delay a bit after writing a bounce
3687 message to get a bit more predictability in the log output.
3689 4. Added a call to search_tidyup() just before forking a reception process. In
3690 theory, someone could use a lookup in the expansion of smtp_accept_max_
3691 per_host which, without the tidyup, could leave open a database connection.
3693 5. Added the variables $recipient_data and $sender_data which get set from a
3694 lookup success in an ACL "recipients" or "senders" condition, or a router
3695 "senders" option, similar to $domain_data and $local_part_data.
3697 6. Moved the writing of debug_print from before to after the "senders" test
3700 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3701 problems for message scanning, either using a data ACL, or using
3702 local_scan() because the Received: header was not generated till after they
3703 were called (in order to set the time as the time of reception completion).
3704 I have revised the way this works. The header is now generated after the
3705 body is received, but before the ACL or local_scan() are called. After they
3706 are run, the timestamp in the header is updated.
3712 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3713 before starting a queue runner without re-exec. This happened only when
3714 deliver_drop_privilege was set or when the Exim user was set to root. The
3715 effect of the bug was that timeouts during subsequent deliveries caused
3716 crashes instead of being properly handled. The handler is now left at its
3717 default (and expected) setting.
3719 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3720 message, again when deliver_drop_privilege is set or Exim is run as root.
3721 The bug described in (1) was not present in this case, but the tidying up
3722 of the other signals was missing. I have made the two cases consistent.
3724 3. The ignore_target_hosts setting on a manualroute router was being ignored
3725 for hosts that were looked up using the /MX notation.
3727 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3730 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3731 operated on the sender address. After changing the $sender_address to <>
3732 for the sender address verify, Exim was re-instated it as the original
3733 (before rewriting) address, but remembering that it had rewritten it, so it
3734 wasn't rewriting it again. This bug also had the effect of breaking the
3735 sender address verification caching when the sender address was rewritten.
3737 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3738 This has been changed so that if the ip literal address matches
3739 ignore_target_hosts, the router declines.
3741 7. Added expansion conditions match_domain, match_address, and match_local_
3742 part (NOT match_host).
3744 8. The placeholder for the Received: header didn't have a length field set.
3746 9. Added code to Exim itself and to exim_lock to test for a specific race
3747 condition that could lead to file corruption when using MBX delivery. The
3748 issue is with the lockfile that is created in /tmp. If this file is removed
3749 after a process has opened it but before that process has acquired a lock,
3750 there is the potential for a second process to recreate the file and also
3751 acquire a lock. This could lead to two Exim processes writing to the file
3752 at the same time. The added code performs the same test as UW imapd; it
3753 checks after acquiring the lock that its file descriptor still refers to
3754 the same named file.
3756 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3757 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3760 11. Added the smtp_active_hostname option. If used, this will typically be made
3761 to depend on the incoming interface address. Because $interface_address is
3762 not set up until the daemon has forked a reception process, error responses
3763 that can happen earlier (such as "too many connections") no longer contain
3766 12. If an expansion in a condition on a "warn" statement fails because a lookup
3767 defers, the "warn" statement is abandoned, and the next ACL statement is
3768 processed. Previously this caused the whole ACL to be aborted.
3770 13. Added the iplsearch lookup type.
3772 14. Added ident_timeout as a log selector.
3774 15. Added tls_certificate_verified as a log selector.
3776 16. Added a global option tls_require_ciphers (compare the smtp transport
3777 option of the same name). This controls incoming TLS connections.
3779 17. I finally figured out how to make tls_require_ciphers do a similar thing
3780 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3781 before starting the TLS session.
3783 18. Tabs are now shown as \t in -bP output.
3785 19. If the log selector return_path_on_delivery was set, Exim crashed when
3786 bouncing a message because it had too many Received: header lines.
3788 20. If two routers both had headers_remove settings, and the first one included
3789 a superfluous trailing colon, the final name in the first list and the
3790 first name in the second list were incorrectly joined into one item (with a
3791 colon in the middle).
3797 1. Added -C and -D options to the exinext utility, mainly to make it easier
3798 to include in the automated testing, but these could be helpful when
3799 multiple configurations are in use.
3801 2. The exinext utility was not formatting the output nicely when there was
3802 an alternate port involved in the retry record key, nor when there was a
3803 message id as well (for retries that were specific to a specific message
3804 and a specific host). It was also confused by IPv6 addresses, because of
3805 the additional colons they contain. I have fixed the IPv4 problem, and
3806 patched it up to do a reasonable job for IPv6.
3808 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3809 delivery, the log line now contains "pipelined" if PIPELINING was used.
3811 4. An SMTP transport process used to panic and die if the bind() call to set
3812 an explicit outgoing interface failed. This has been changed; it is now
3813 treated in the same way as a connect() failure.
3815 5. A reference to $sender_host_name in the part of a conditional expansion
3816 that was being skipped was still causing a DNS lookup. This no longer
3819 6. The def: expansion condition was not recognizing references to header lines
3820 that used bh_ and bheader_.
3822 7. Added the _cache feature to named lists.
3824 8. The code for checking quota_filecount in the appendfile transport was
3825 allowing one more file than it should have been.
3827 9. For compatibility with Sendmail, the command line option
3835 and sets the incoming protocol and host name (for trusted callers). The
3836 host name and its colon can be omitted when only the protocol is to be set.
3837 Note the Exim already has two private options, -pd and -ps, that refer to
3838 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3839 "s", but I don't think that's a major issue.
3841 10. A number of refactoring changes to the code, none of which should affect
3844 (a) The number of logging options was getting close to filling up the
3845 32-bit word that was used as a bit map. I have split them into two classes:
3846 those that are passed in the argument to log_write(), and those that are
3847 only ever tested independently outside of that function. These are now in
3848 separate 32-bit words, so there is plenty of room for expansion again.
3849 There is no change in the user interface or the logging behaviour.
3851 (b) When building, for example, log lines, the code previously used a
3852 macro that called string_cat() twice, in order to add two strings. This is
3853 not really sufficiently general. Furthermore, there was one instance where
3854 it was actually wrong because one of the argument was used twice, and in
3855 one call a function was used. (As it happened, calling the function twice
3856 did not affect the overall behaviour.) The macro has been replaced by a
3857 function that can join an arbitrary number of extra strings onto a growing
3860 (c) The code for expansion conditions now uses a table and a binary chop
3861 instead of a serial search (which was left over from when there were very
3862 few conditions). Also, it now recognizes conditions like "pam" even when
3863 the relevant support is not compiled in: a suitably worded error message is
3864 given if an attempt is made to use such a condition.
3866 11. Added ${time_interval:xxxxx}.
3868 12. A bug was causing one of the ddress fields not to be passed back correctly
3869 from remote delivery subprocesses. The field in question was not being
3870 subsequently used, so this caused to problems in practice.
3872 13. Added new log selectors queue_time and deliver_time.
3874 14. Might have fixed a bug in maildirsizefile handling that threw up
3875 "unexpected character" debug warnings, and recalculated the data
3876 unnecessarily. In any case, I expanded the warning message to give more
3879 15. Added the message "Restricted characters in address" to the statements in
3880 the default ACL that block characters like @ and % in local parts.
3882 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3883 Three changes have been made:
3885 (a) There was a serious bug; a negative response to MAIL caused the whole
3886 recipient domain to be cached as invalid, thereby blocking all messages
3887 to all local parts at the same domain, from all senders. This bug has
3888 been fixed. The domain is no longer cached after a negative response to
3889 MAIL if the sender used is not empty.
3891 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3894 (c) A new callout option, "use_sender" has been added for people who want
3895 the modified behaviour.
3901 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3904 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3905 indeed breaks things for older releases.
3907 3. Added additional logging to the case where there is a problem reading data
3908 from a filter that is running in a subprocess using a pipe, in order to
3909 try to track down a specific problem.
3911 4. Testing facility fudge: when running in the test harness and attempting
3912 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3913 getting "No route to host". Convert this to a timeout.
3915 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3918 6. Some OS don't have socklen_t but use size_t instead. This affects the
3919 fifth argument of getsockopt() amongst other things. This is now
3920 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3921 can be set for individual OS. I have set it for SunOS5, OSF1, and
3922 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3923 some earlier ones do not.
3925 7. Change 4.30/15 was not doing the test caselessly.
3927 8. The standard form for an IPv6 address literal was being rejected by address
3928 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3929 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3930 this, as well as the form without the "IPv6" on the front (but only when
3931 address literals are enabled, of course).
3933 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3935 10. Exim crashed if a message with an empty sender address specified by -f
3936 encountered a router with an errors_to setting. This could be provoked only
3937 by a command such as
3941 where an empty string was supplied; "<>" did not hit this bug.
3943 11. Installed PCRE release 4.5.
3945 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3946 remained set. It is now erased.
3948 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3949 times from message ids (which are base 36 rather than the normal 62).
3951 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3952 were being counted as actual protocol errors, and logged if the log
3953 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3954 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3955 and DATA following a set of rejected RCPTs do not count as protocol errors.
3956 In other words, Exim assumes they were pipelined, though this may not
3957 actually be the case. Of course, in all cases the client gets an
3958 appropriate error code.
3960 15. If a lookup fails in an ACL condition, a message about the failure may
3961 be available; it is used if testing the ACL cannot continue, because most
3962 such messages specify what the cause of the deferral is. However, some
3963 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3964 that caused an old message to be retained and used if a later statement
3965 caused a defer, replacing the real cause of the deferral.
3967 16. If an IP address had so many PTR records that the DNS lookup buffer
3968 was not large enough to hold them, Exim could crash while trying to process
3969 the truncated data. It now detects and logs this case.
3971 17. Further to 4.21/58, another change has been made: if (and only if) the
3972 first line of a message (the first header line) ends with CRLF, a bare LF
3973 in a subsequent header line has a space inserted after it, so as not to
3974 terminate the header.
3976 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3977 unnecessarily, used atoi() instead of strtol(), and didn't check the
3978 termination when getting file sizes from file names by regex.
3980 19. Completely re-implemented the support for maildirsize files, in the light
3981 of a number of problems with the previous contributed implementation
3982 (4.30/29). In particular:
3984 . If the quota is zero, the maildirsize file is maintained, but no quota is
3987 . If the maildir directory does not exist, it is created before any attempt
3988 to write a maildirsize file.
3990 . The quota value in the file is just a cache; if the quota is changed in
3991 the transport, the new value overrides.
3993 . A regular expression is available for excluding directories from the
3996 20. The autoreply transport checks the characters in options that define the
3997 message's headers; it allows continued headers, but it was checking with
3998 isspace() after an embedded newline instead of explicitly looking for a
4001 21. If all the "regular" hosts to which an address was routed had passed their
4002 expiry times, and had not reached their retry times, the address was
4003 bounced, even if fallback hosts were defined. Now Exim should go on to try
4006 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4007 equivalent code in the SMTP transport. Some hosts send humungous responses
4008 to HELO/EHLO, more than 1024 it seems.
4010 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4011 gives compiler warnings in some environments. I've now done it "properly",
4014 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4015 (because of problems with the built-in one) was declared to return uschar *
4016 instead of char *, causing compiler failure.
4018 25. Fixed a file descriptor leak when processing alias/forward files.
4020 26. Fixed a minor format string issue in dbfn.c.
4022 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4024 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4025 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4028 29. When a sender address is verified, it is cached, to save repeating the test
4029 when there is more than one recipient in a message. However, when the
4030 verification involves a callout, it is possible for different callout
4031 options to be set for different recipients. It is too complicated to keep
4032 track of this in the cache, so now Exim always runs a verification when a
4033 callout is required, relying on the callout cache for the optimization.
4034 The overhead is duplication of the address routing, but this should not be
4037 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4038 to be verified with callout=postmaster, and the main callout worked but the
4039 postmaster check failed, the verification correctly failed. However, if a
4040 subsequent RCPT command asked for sender verification *without* the
4041 postmaster check, incorrect caching caused this verification also to fail,
4044 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4045 it was not caching the DNS options (qualify_single, search_parents) that
4046 were used when the lookup failed. A subsequent lookup with different
4047 options therefore always gave the same answer, though there were cases
4048 where it should not have. (Example: a "domains = !$mx_any" option on a
4049 dnslookup router: the "domains" option is always processed without any
4050 widening, but the router might have qualify_single set.) Now Exim uses the
4051 cached value only when the same options are set.
4053 32. Added John Jetmore's "exipick" utility to the distribution.
4055 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4056 than a timeout (e.g. a certificate is required, and is not provided), an
4057 Exim server now closes the connection immediately. Previously it waited for
4058 the client to close - but if the client is SSL, it seems that they each
4059 wait for each other, leading to a delay before one of them times out.
4061 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4062 maintained 0.8.x compatibility because I don't think many are using it, and
4063 it is clearly obsolete.
4065 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4068 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4069 client certificate was expired. A simple patch fixes this, though I don't
4070 understand the full logic of why the verify callback is called multiple
4073 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4074 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4075 which causes problems with some clients (such as the Certicom SSL Plus
4076 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4077 disables the coutermeasure allowing Eudora to connect."
4079 38. Exim was not checking that a write() to a log file succeeded. This could
4080 lead to Bad Things if a log got too big, in particular if it hit a file
4081 size limit. Exim now panics and dies if it cannot write to a log file, just
4082 as it does if it cannot open a log file.
4084 39. Modified OS/Makefile-Linux so that it now contains
4086 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4088 The two -D definitions ensure that Exim is compiled with large file
4089 support, which makes it possible to handle log files that are bigger than
4092 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4093 instance) a domain was checked against a named list that involved a lookup,
4094 causing $domain_data to be set, then another domain was checked against the
4095 same list, then the first domain was re-checked, the value of $domain_data
4096 after the final check could be wrong. In particular, if the second check
4097 failed, it could be set empty. This bug probably also applied to
4100 41. The strip_trailing_dot option was not being applied to the address given
4101 with the -f command-line option.
4103 42. The code for reading a message's header from the spool was incrementing
4104 $received_count, but never initializing it. This meant that the value was
4105 incorrect (doubled) while delivering a message in the same process in which
4106 it was received. In the most common configuration of Exim, this never
4107 happens - a fresh exec is done - but it can happen when
4108 deliver_drop_privilege is set.
4110 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4111 it now includes up to 150 characters of the unexpected data in the log
4114 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4115 and building data strings. The size of both of these buffers was 10 000
4116 bytes - far larger than anybody would *ever* want, thought I. Needless to
4117 say, somebody hit the limit. I have increased the maximum line length to
4118 20 000 and the maximum data length of concatenated lines to 100 000. I have
4119 also fixed two bugs, because there was no checking on these buffers. Tsk,
4120 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4121 buffer is too small.
4123 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4124 lsearch lookups. Now it does.
4126 46. When parsing a route_list item in a manualroute router, a fixed-length
4127 buffer was used for the list of hosts. I made this 1024 bytes long,
4128 thinking that nobody would ever have a list of hosts that long. Wrong.
4129 Somebody had a whole pile of complicated expansion conditions, and the
4130 string was silently truncated, leading to an expansion error. It turns out
4131 that it is easier to change to an unlimited length (owing to other changes
4132 that have happened since this code was originally written) than to build
4133 structure for giving a limitation error. The length of the item that
4134 expands into the list of hosts is now unlimited.
4136 47. The lsearch lookup could not handle data where the length of text line was
4137 more than 4095 characters. Such lines were truncated, leading to shortened
4138 data being returned. It should now handle lines of any length.
4140 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4141 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4144 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4145 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4146 no longer happens in comment lines. A list of replacements is now placed
4147 at the head of all of the source files, except those whose only change is
4148 to replace PERL_COMMAND in the very first #! line.
4150 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4151 messages on the queue, with a bottom-up merge sort, using code contributed
4152 by Michael Haardt. This should make operations like -bp somewhat faster on
4153 large queues. It won't affect queue runners, except when queue_run_in_order
4156 51. Installed eximstats 1.31 in the distribution.
4158 52. Added support for SRV lookups to the dnslookup router.
4160 53. If an ACL referred to $message_body or $message_body_end, the value was not
4161 reset for any messages that followed in the same SMTP session.
4163 54. The store-handling optimization for building very long strings was not
4164 differentiating between the different store pools. I don't think this
4165 actually made any difference in practice, but I've tidied it.
4167 55. While running the routers to verify a sender address, $sender_address
4168 was still set to the sender address. This is wrong, because when routing to
4169 send a bounce to the sender, it would be empty. Therefore, I have changed
4170 it so that, while verifying a sender address, $sender_address is set to <>.
4171 (There is no change to what happens when verifying a recipient address.)
4173 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4174 target A or AAAA records (if not already returned) without resetting the
4175 qualify_single or search_parents options of the DNS resolver. These are
4176 inappropriate in this case because the targets of MX and SRV records must
4177 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4178 target that, when qualified, matched something in the local domain. These
4179 two options are now turned off when doing these lookups.
4181 57. It seems that at least some releases of Reiserfs (which does not have the
4182 concept of a fixed number of inodes) returns zero and not -1 for the
4183 number of available inodes. This interacted badly with check_spool_inodes,
4184 which assumed that -1 was the "no such thing" setting. What I have done is
4185 to check that the total number of inodes is greater than zero before doing
4186 the test of how many are available.
4188 58. When a "warn" ACL statement has a log_message modifier, the message is
4189 remembered, and not repeated. This is to avoid a lot of repetition when a
4190 message has many recipients that cause the same warning to be written.
4191 Howewer, Exim was preserving the list of already written lines for an
4192 entire SMTP session, which doesn't seem right. The memory is now reset if a
4193 new message is started.
4195 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4196 debugging output unless log_rewrite was also set.
4198 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4199 of (int)(handle) when we know that handle contains (void *)(-1).
4201 61. The Exim daemon panic-logs an error return when it closes the incoming
4202 connection. However "connection reset by peer" seems to be common, and
4203 isn't really an error worthy of noting specially, so that particular error
4206 62. When Exim is trying to find all the local interfaces, it used to panic and
4207 die if the ioctl to get the interface flags failed. However, it seems that
4208 on at least one OS (Solaris 9) it is possible to have an interface that is
4209 included in the list of interfaces, but for which you get a failure error
4210 for this call. This happens when the interface is not "plumbed" into a
4211 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4212 failure of the "get flags" call assumes that the interface is down.
4214 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4215 makes life easier for people who are doing arithmetic on fields extracted
4216 from dates, where you often get leading zeros that should not be
4217 interpreted as octal.
4219 64. Added qualify_domain to the redirect router, to override the global
4222 65. If a pathologically long header line contained very many addresses (the
4223 report of this problem mentioned 10 000) and each of them was rewritten,
4224 Exim could use up a very large amount of memory. (It kept on making new
4225 copies of the header line as it rewrote, and never released the old ones.)
4226 At the expense of a bit more processing, the header rewriting function has
4227 been changed so that it no longer eats memory in this way.
4229 66. The generation of the Received: header has been moved from the time that a
4230 message starts to be received, to the time that it finishes. The timestamp
4231 in the Received: header should now be very close to that of the <= log
4232 line. There are two side-effects of this change:
4234 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4235 logged header lines no longer include the local Received: line, because
4236 it has not yet been created. The same applies to a copy of the message
4237 that is returned to a non-SMTP sender when a message is rejected.
4239 (b) When a filter file is tested using -bf, no additional Received: header
4240 is added to the test message. After some thought, I decided that this
4243 This change does not affect the value of $received_for. It is still set
4244 after address rewriting, but before local_scan() is called.
4246 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4248 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4249 gave an unhelpful panic error message, and a defer error. I have managed to
4250 change this behaviour so that it now rejects any supplied certificate,
4251 which seems right, as the list of acceptable certificates is empty.
4253 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4254 gave an unhelpful defer error. I have not managed to make this reject any
4255 supplied certificates, but the error message it gives is "no certificate
4256 supplied", which is not helpful.
4258 70. exigrep's output now also includes lines that are not associated with any
4259 message, but which match the given pattern. Implemented by a patch from
4260 Martin Sluka, which also tidied up the Perl a bit.
4262 71. Recipient callout verification, like sender verification, was using <> in
4263 the MAIL FROM command. This isn't really the right thing, since the actual
4264 sender may affect whether the remote host accepts the recipient or not. I
4265 have changed it to use the actual sender in the callout; this means that
4266 the cache record is now keyed on a recipient/sender pair, not just the
4267 recipient address. There doesn't seem to be a real danger of callout loops,
4268 since a callout by the remote host to check the sender would use <>.
4269 [SEE ABOVE: changed after hitting problems.]
4271 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4272 temporary errors. However, in the case of such a code being given after
4273 the end of a data transmission (i.e. after ".") Exim was failing to write
4274 a retry record for the message. (Yes, there was some broken host that was
4275 actually sending 8xx at this point.)
4277 73. An unknown lookup type in a host list could cause Exim to panic-die when
4278 the list was checked. (An example that provoked this was putting <; in the
4279 middle of a list instead of at the start.) If this happened during a DATA
4280 ACL check, a -D file could be left lying around. This kind of configuration
4281 error no longer causes Exim to die; instead it causes a defer errror. The
4282 incident is still logged to the main and panic logs.
4284 74. Buglet left over from Exim 3 conversion. The message "too many messages
4285 in one connection" was written to the rejectlog but not the mainlog, except
4286 when address rewriting (yes!) was being logged.
4288 75. Added write_rejectlog option.
4290 76. When a system filter was run not as root (that is, when system_filter_user
4291 was set), the values of the $n variables were not being returned to the
4292 main process; thus, they were not subsequently available in the $sn
4295 77. Added +return_path_on_delivery log selector.
4297 78. A connection timeout was being treated differently from recipients deferred
4298 when testing hosts_max_try with a message that was older than the host's
4299 retry timeout. (The host should not be counted, thus allowing all hosts to
4300 be tried at least once before bouncing.) This may have been the cause of an
4301 occasionally reported bug whereby a message would remain on the queue
4302 longer than the retry timeout, but would be bounced if a delivery was
4303 forced. I say "may" because I never totally pinned down the problem;
4304 setting up timeout/retry tests is difficult. See also the next item.
4306 79. The ultimate address timeout was not being applied to errors that involved
4307 a combination of host plus message (for example, a timeout on a MAIL
4308 command). When an address resolved to a number of possible hosts, and they
4309 were not all tried for each delivery (e.g. because of hosts_max_try), a
4310 message could remain on the queue longer than the retry timeout.
4312 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4315 81. Fixed an obscure SMTP outgoing bug which required at least the following
4316 conditions: (a) there was another message waiting for the same server;
4317 (b) the server returned 5xx to all RCPT commands in the first message so
4318 that the message was not completed; (c) the server dropped the connection
4319 or gave a negative response to the RSET that Exim sends to abort the
4320 transaction. The observed case was a dropped connection after DATA that had
4321 been sent in pipelining mode. That is, the server had advertised PIPELINING
4322 but was not implementing it correctly. The effect of the bug was incorrect
4323 behaviour, such as trying another host, and this could lead to a crash.
4329 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4330 and daemon.c were passed as pointers to ints; they should have been
4331 pointers to socklen_t variables (which are typically unsigned ints).
4333 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4336 3. Fixed a really odd bug that affected only the testing scheme; patching a
4337 certain fixed string in the binary changed the value of another string that
4338 happened to be identical to the end of the original first string.
4340 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4341 name", it returns that address as the IP address. On some operating
4342 systems (e.g. Solaris), it also passes back the IP address string as the
4343 "host name". However, on others (e.g. Linux), it passes back an empty
4344 string. Exim wasn't checking for this, and was changing the host name to an
4345 empty string, assuming it had been canonicized.
4347 5. Although rare, it is permitted to have more than one PTR record for a given
4348 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4349 all the names associated with an address, because they do in Solaris.
4350 However, it seems that they do not in Linux for data that comes from the
4351 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4352 I found this out when I moved to a new Linux workstation and tried to run
4353 the Exim test suite.
4355 To get round this problem I have changed the code so that it now does its
4356 own call to the DNS to look up PTR records when searching for a host name.
4357 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4358 addresses that are only in /etc/hosts are still found.
4360 This behaviour is, however, controlled by an option called host_lookup_
4361 order, which defaults to "bydns:byaddr". If people want to use the other
4362 order, or indeed, just use one or the other means of lookup, they can
4363 specify it in this variable.
4365 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4366 some operating systems, this comes back from gethostbyaddr() as an empty
4367 string, and this is what Exim used to test for. However, it seems that in
4368 other systems, "." is yielded. Exim now tests for this case too.
4370 7. The values of check_spool_space and check_log_space are now held internally
4371 as a number of kilobytes instead of an absolute number of bytes. If a
4372 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4373 kilobyte. This means that much larger values can be stored.
4375 8. Exim monitor: an attempt to get the action menu when not actually pointing
4376 at a message produces an empty menu entitled "No message selected". This
4377 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4378 no entries in it ("Shell widget menu has zero width and/or height"). So I
4379 have added a single, blank menu entry in this case.
4381 9. Added ${quote_local_part.
4383 10. MIME decoding is now applied to the contents of Subject: header lines when
4386 11. Now that a reference to $sender_host_address automatically causes a reverse
4387 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4388 host lookup before query-style lookups in lists that might use this
4389 variable. This has therefore been abolished, and the "net-" prefix is no
4390 longer necessary for query-style lookups.
4392 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4393 have been a typo for LFLAGS, so it has been changed.
4395 13. The install script calls Exim with "-C /dev/null" in order to find the
4396 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4397 to be output. Howeve, since Exim outputs its version number before the
4398 error, it didn't break the script. It just looked ugly. I fixed this by
4399 always allowing "-C /dev/null" if the caller is root.
4401 14. Ignore overlarge ACL variable number when reading spool file - insurance
4402 against a later release with more variables having written the file.
4404 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4405 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4406 this, as well as the form without the "IPv6" on the front.
4408 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4409 OS/Makefile-Darwin file.
4411 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4412 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4414 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4415 variable. However, some OS use ints for inodes. I've added cast to long int
4416 to get rid of the compiler warning.
4418 19. I had forgotten to lock out "/../" in configuration file names when
4419 ALT_CONFIG_PREFIX was set.
4421 20. Routers used for verification do not need to specify transports. However,
4422 if such a router generated a host list, and callout was configured, Exim
4423 crashed, because it could not find a port number from the (non-existent)
4424 transport. It now assumes port 25 in this circumstance.
4426 21. Added the -t option to exigrep.
4428 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4429 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4430 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4431 them is set, LOOKUP_LSEARCH is forced.
4433 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4435 24. Added sender and host information to the "rejected by local_scan()" log
4436 line; previously there was no indication of these.
4438 25. Added .include_if_exists.
4440 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4441 Linux. It turns out that not all file systems support this. Apparently some
4442 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4443 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4444 not supported on the file descriptor, is now ignored when Exim is trying to
4445 sync a directory. This applies only to Linux.
4447 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4449 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4450 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4453 29. Added support for maildirsize files from supplied patch (modified a bit).
4455 30. The use of :fail: followed by an empty string could lead Exim to respond to
4456 sender verification failures with (e.g.):
4458 550 Verification failed for <xxx>
4459 550 Sender verify failed
4461 where the first response line was missing the '-' that indicates it is not
4462 the final line of the response.
4464 31. The loop for finding the name of the user that called Exim had a hardwired
4465 limit of 10; it now uses the value of finduser_retries, which is used for
4466 all other user lookups.
4468 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4471 33. Exim was neglecting to zero errno before one call of strtol() when
4472 expanding a string and expecting an integer value. On some systems this
4473 resulted in spurious "integer overflow" errors. Also, it was casting the
4474 result into an int without checking.
4476 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4477 did not work. The code looks as if it has *never* worked, though it appears
4478 to have been documented since at least releast 1.62. I have made it work.
4480 35. The "timeout_DNS" error in retry rules, also documented since at least
4481 1.62, also never worked. As it isn't clear exactly what this means, and
4482 clearly it isn't a major issue, I have abolished the feature by treating it
4483 as "timeout", and writing a warning to the main and panic logs.
4485 36. The display of retry rules for -brt wasn't always showing the error code
4488 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4489 timeout_connect_A, timeout_connect_MX.
4491 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4492 to the empty sender.
4494 39. The daemon was not analysing the content of -oX till after it had closed
4495 stderr and disconnected from the controlling terminal. This meant that any
4496 syntax errors were only noted on the panic log, and the return code from
4497 the command was 0. By re-arranging the code a little, I've made the
4498 decoding happen first, so such errors now appear on stderr, and the return
4499 code is 1. However, the actual setting up of the sockets still happens in
4500 the disconnected process, so errors there are still only recorded on the
4503 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4504 connections (as happens on some IP stacks) was logged at start up time as
4505 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4506 it from "IPv6 and IPv4", which means that two separate sockets are being
4509 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4510 says whether AF_INET or AF_INET6 was passed as an argument.
4512 42. Exiwhat output was messed up when time zones were included in log
4515 43. Exiwhat now gives more information about the daemon's listening ports,
4516 and whether -tls-on-connect was used.
4518 44. The "port" option of the smtp transport is now expanded.
4520 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4521 silently ignored. Now an error message is written to the main and panic
4524 46. There's a new ACL modifier called "logwrite" which writes to a log file
4525 as soon as it is encountered.
4527 47. Added $local_user_uid and $local_user_gid at routing time.
4529 48. Exim crashed when trying to verify a sender address that was being
4532 49. Exim was recognizing only a space character after ".include". It now also
4533 recognizes a tab character.
4535 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4536 extracting the relevant information from the specification. The man page no
4537 longer contains scrambled data for the -d option, and I've added a section
4538 at the front about calling Exim under different names.
4540 51. Added "extra_headers" argument to the "mail" command in filter files.
4542 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4545 53. Installed eximstats 1.29.
4547 54. Added transport_filter_timeout as a generic transport option.
4549 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4550 Cc: header lines. This was required by RFC 822, but it not required by RFC
4553 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4554 incoming messages that did not have them. Now it does so only if the
4555 message originates locally, that is, if there is no associated remote host
4556 address. When Resent- header lines are present, this applies to the Resent-
4557 lines rather than the non-Resent- lines.
4559 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4560 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4562 58. Messages for configuration errors now include the name of the main
4563 configuration file - useful now that there may be more than one file in a
4564 list (.included file names were always shown).
4566 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4567 for those rare installations that do not start the daemon as root or run it
4568 setuid root. I've cut out the call to initgroups() if the daemon is not
4571 60. The Exim user and group can now be bound into the binary as text strings
4572 that are looked up at the start of Exim's processing.
4574 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4576 62. Added $mailstore_basename variable.
4578 63. Installed patch to sieve.c from Michael Haardt.
4580 64. When Exim failed to open the panic log after failing to open the main log,
4581 the original message it was trying to log was written to stderr and debug
4582 output, but if they were not available (the usual case in production), it
4583 was lost. Now it is written to syslog before the two lines that record the
4584 failures to open the logs.
4586 65. Users' Exim filters run in subprocesses under the user's uid. It is
4587 possible for a "deliver" command or an alias in a "personal" command to
4588 provoke an address rewrite. If logging of address rewriting is configured,
4589 this fails because the process is not running as root or exim. There may be
4590 a better way of dealing with this, but for the moment (because 4.30 needs
4591 to be released), I have disabled address rewrite logging when running a
4592 filter in a non-root, non-exim process.
4598 1. The buildconfig auxiliary program wasn't quoting the value set for
4599 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4600 not defined. This bug was masked in 4.22 by the effect that was fixed in
4603 2. Some messages that were rejected after a message id was allocated were
4604 shown as "incomplete" by exigrep. It no longer does this for messages that
4605 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4607 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4608 have allow_domain_literals set, the ID did not get logged in the <= line.
4609 Domain literals are now always recognized in Message-ID: header lines.
4611 4. The first argument for a ${extract expansion item is the key name or field
4612 number. Leading and trailing spaces in this item were not being ignored,
4613 causing some misleading effects.
4615 5. When deliver_drop_privilege was set, single queue runner processes started
4616 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4617 same command in the process it spins off) were not dropping privilege.
4619 6. When the daemon running as "exim" started a queue runner, it always
4620 re-executed Exim in the spun-off process. This is a waste of effort when
4621 deliver_drop_privilege is set. The new process now just calls the
4622 queue-runner function directly.
4628 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4631 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4632 ignored. Though the use of -oP was forcing the writing of a pid file, it
4633 was always written to the default place.
4635 3. If the message "no IP address found for host xxxx" is generated during
4636 incoming verification, it is now followed by identification of the incoming
4637 connection (so you can more easily find what provoked it).
4639 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4641 5. Added some features to "harden" Exim a bit more against certain attacks:
4643 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4644 be put in Local/Makefile. This is like the never_users runtime option,
4645 but it cannot be overridden. The default setting is "root".
4647 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4648 prefix string with which any file named in a -C command line option
4651 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4652 is retained for -C and -D only if the caller of Exim is root. Without
4653 it, the exim user may also use -C and -D and retain privilege.
4655 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4656 command line option is disabled.
4658 6. Macro names set by the -D option must start with an upper case letter, just
4659 like macro names defined in the configuration file.
4661 7. Added "dereference=" facility to LDAP.
4663 8. Two instances of the typo "uknown" in the source files are fixed.
4665 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4666 the Configure-Makefile script screwed up while processing it.
4668 10. Incorporated PCRE 4.4.
4670 11. The SMTP synchronization check was not operating right at the start of an
4671 SMTP session. For example, it could not catch a HELO sent before the client
4672 waited for the greeting. There is now a check for outstanding input at the
4673 point when the greeting is written. Because of the duplex, asynchronous
4674 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4675 way, but not yet received, when the check is performed.
4677 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4678 on TCP/IP sockets, because this apparently causes some broken clients to
4681 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4682 unchanged) from the Cygwin maintainer.
4684 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4685 was supported instead of testing for mbx. Effectively a typo.
4687 15. The spa authenticator server code was not checking that the input it
4688 received was valid base64.
4690 16. The debug output line for the "set" modifier in ACLs was not showing the
4691 name of the variable that was being set.
4693 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4695 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4696 lookup on its own. Something else was needed to trigger the lookup. For
4697 example, a match in host_lookup or the need for a host name in a host list.
4698 Now, if $sender_host_name is referenced and the host name has not yet been
4699 looked up, a lookup is performed. If the lookup fails, the variable remains
4700 empty, and $host_lookup_failed is set to "1".
4702 19. Added "eqi" as a case-independent comparison operator.
4704 20. The saslauthd authentication condition could segfault if neither service
4705 nor realm was specified.
4707 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4708 error message that was logged was misleading, and incoming SMTP
4709 connections were dropped. The message is now more accurate, and temporary
4710 errors are given to SMTP connections.
4712 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4713 (except RSET) with a 421 error, until QUIT is received. However, it was
4714 failing to send a response to QUIT.
4716 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4717 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4718 ACL is now run after the helo_try_verify_hosts code.
4720 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4721 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4722 case-independent, so other case variants are also recognized). Apparently
4723 some systems use these upper case variants.
4725 25. If more than two messages were waiting for the same host, and a transport
4726 filter was specified for the transport, Exim sent two messages over the
4727 same TCP/IP connection, and then failed with "socket operation on non-
4728 socket" when it tried to send the third.
4730 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4732 27. The extern definition of crypt16() in expand.c was not being excluded when
4733 the OS had its own crypt16() function.
4735 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4736 as a preferred synonym for return_size_limit, both as an option and as an
4739 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4741 30. Changed the default configuration ACL to relax the local part checking rule
4742 for addresses that are not in any local domains. For these addresses,
4743 slashes and pipe symbols are allowed within local parts, but the sequence
4744 /../ is explicitly forbidden.
4746 31. SPA server authentication was not clearing the challenge buffer before
4749 32. log_message in a "warn" ACL statement was writing to the reject log as
4750 well as to the main log, which contradicts the documentation and doesn't
4751 seem right (because no rejection is happening). So I have stopped it.
4753 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4754 However, I am unable to do any testing of this.
4756 34. Fixed an infelicity in the appendfile transport. When checking directories
4757 for a mailbox, to see if any needed to be created, it was accidentally
4758 using path names with one or more superfluous leading slashes; tracing
4759 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4761 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4762 contents are added to the log line that is written for every discarded
4763 recipient. (Previously a log_message setting was ignored.)
4765 36. The ${quote: operator now quotes the string if it is empty.
4767 37. The install script runs exim in order to find its version number. If for
4768 some reason other than non-existence or emptiness, which it checks, it
4769 could not run './exim', it was installing it with an empty version number,
4770 i.e. as "exim-". This error state is now caught, and the installation is
4773 38. An argument was missing from the function that creates an error message
4774 when Exim fails to connect to the socket for saslauthd authentication.
4775 This could cause Exim to crash, or give a corrupted message.
4777 39. Added isip, isip4, and isip6 to ${if conditions.
4779 40. The ACL variables $acl_xx are now saved with the message, and can be
4780 accessed later in routers, transports, and filters.
4782 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4783 strings in the file are not string-expanded.
4785 42. If a MAIL command specified a SIZE value that was too large to fit into an
4786 int variable, the check against message_size_limit failed. Such values are
4787 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4788 day this will have to be increased, but I don't think I want to be around
4789 when emails are that large.
4796 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4797 iconv() is not standard in FreeBSD.
4799 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4800 IPv6 enabled. The observed symptom was a segmentation fault on return from
4801 the function os_common_find_running_interfaces() in src/os.c.
4803 3. In the check_special_case() function in daemon.c I had used "errno" as an
4804 argument name, which causes warnings on some systems. This was basically a
4805 typo, since it was named "eno" in the comments!
4807 4. The code that waits for the clock to tick (at a resolution of some fraction
4808 of a second) so as to ensure message-id uniqueness was always waiting for
4809 at least one whole tick, when it could have waited for less. [This is
4810 almost certainly not relevant at current processor speeds, where it is
4811 unlikely to ever wait at all. But we try to future-proof.]
4813 5. The function that sleeps for a time interval that includes fractions of a
4814 second contained a race. It did not block SIGALRM between setting the
4815 timer, and suspending (a couple of lines later). If the interval was short
4816 and the sigsuspend() was delayed until after it had expired, the suspension
4817 never ended. On busy systems this could lead to processes getting stuck for
4820 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4821 process, before it forks any delivery processes. The open lookup caching
4822 mechanism meant that the open file or database connection was passed into
4823 the delivery process. The problem was that delivery processes always tidy
4824 up cached lookup data. This could cause a problem for the next delivery
4825 process started by the queue runner, because the external queue runner
4826 process does not know about the closure. So the next delivery process
4827 still has data in the lookup cache. In the case of a file lookup, there was
4828 no problem because closing a file descriptor in a subprocess doesn't affect
4829 the parent. However, if the lookup was caching a connection to a database,
4830 the connection was closed, and the second delivery process was likely to
4831 see errors such as "PGSQL: query failed: server closed the connection
4832 unexpectedly". The problem has been fixed by closing all cached lookups
4833 in a queue runner before running a delivery process.
4835 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4836 seem to have the "const" qualifier which it has on other OS. I've
4839 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4840 *of the same type* (client or server) with the same public name that an
4841 error should be diagnosed.
4843 9. When Exim looked up a host name for an IP address, but failed to find the
4844 original IP address when looking up the host name (a safety check), it
4845 output the message "<ip address> does not match any IP for NULL", which was
4846 confusing, to say the least. The bug was that the host name should have
4847 appeared instead of "NULL".
4849 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4850 user that is built into the binary, and the -C or -D options is used, root
4851 privilege is dropped before the configuration file is read. In addition,
4852 logging is switched to stderr instead of the normal log files. If the
4853 configuration then re-defines the Exim user, the unprivileged environment
4854 is probably not what is expected, so Exim logs a panic warning message (but
4857 However, if deliver_drop_privilege is set, the unprivileged state may well
4858 be exactly what is intended, so the warning has been cut out in that case,
4859 and Exim is allowed to try to write to its normal log files.
4865 1. smtp_return_error_details was not giving details for temporary sender
4866 or receiver verification errors.
4868 2. Diagnose a configuration error if two authenticators have the same public
4871 3. Exim used not to create the message log file for a message until the first
4872 delivery attempt. This could be confusing when incoming messages were held
4873 for policy or load reasons. The message log file is now created at the time
4874 the message is received, and an initial "Received" line is written to it.
4876 4. The automatically generated man page for command line options had a minor
4877 bug that caused no ill effects; however, a more serious problem was that
4878 the procedure for building the man page automatically didn't always
4879 operate. Consequently, release 4.20 contains an out-of-date version. This
4880 shouldn't happen again.
4882 5. When building Exim with embedded Perl support, the script that builds the
4883 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4884 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4886 6. The freeze_tell option was not being used for messages that were frozen on
4887 arrival, either by an ACL or by local_scan().
4889 7. Added the smtp_incomplete_transaction log selector.
4891 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4892 was accepting AUTH without a new EHLO.
4894 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4895 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4898 10. Logging of TCP/IP connections (when configured) now happens in the main
4899 daemon process instead of the child process, so that the TCP/IP connection
4900 count is more accurate (but it can never be perfect).
4902 11. The use of "drop" in a nested ACL was not being handled correctly in the
4903 outer ACL. Now, if condition failure induced by the nested "drop" causes
4904 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4905 or "require"), the connection is dropped.
4907 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4908 that yield "discard" can now be used with an "accept" or a "discard" verb,
4909 but an error is generated for any others (because I can't see a useful way
4910 to define what should happen).
4912 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4913 are now processed in the same way as lines in the Exim configuration file.
4914 In particular, continuation lines are supported.
4916 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4918 15. Added -ti meaning -t -i.
4920 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4921 domains, and warn by logging if others are found.
4923 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4924 returned from GIFCONF when Exim is trying to find the list of interfaces on
4925 a host. The code in os.c has been modified to copy each ifreq to an aligned
4926 structure in all cases.
4928 Also, in some cases, the returned ifreq's were being copied to a 'struct
4929 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4930 means the last couple of bytes of an IPv6 address could be chopped if the
4931 ifreq contained only a normal sockaddr (14 bytes storage).
4933 18. Named domain lists were not supported in the hosts_treat_as_local option.
4934 An entry such as +xxxx was not recognized, and was treated as a literal
4937 19. Ensure that header lines added by a DATA ACL are included in the reject log
4938 if the ACL subsequently rejects the message.
4940 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4941 MD5 (which is deprecated).
4943 21. When testing a filter file using -bf, Exim was writing a message when it
4944 took the sender from a "From " line in the message, but it was not doing so
4945 when it took $return_path from a Return-Path: header line. It now does.
4947 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4948 with a valid header line field name (a series of printing characters
4949 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4951 23. Changed "disc" in the source to "disk" to conform to the documentation and
4952 the book and for uniformity.
4954 24. Ignore Sendmail's -Ooption=value command line item.
4956 25. When execve() failed while trying to run a command in a pipe transport,
4957 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4958 could be confused with a return value of 69 from the command itself. This
4959 has been changed to 127, the value the shell returns if it is asked to run
4960 a non-existent command. The wording for the related log line suggests a
4961 non-existent command as the problem.
4963 26. If received_header_text expands to an empty string, do not add a Received:
4964 header line to the message. (Well, it adds a token one on the spool, but
4965 marks it "old" so that it doesn't get used or transmitted.)
4967 27. Installed eximstats 1.28 (addition of -nt option).
4969 28. There was no check for failure on the call to getsockname() in the daemon
4970 code. This can fail if there is a shortage of resources on the system, with
4971 ENOMEM, for example. A temporary error is now given on failure.
4973 29. Contrary to the C standard, it seems that in some environments, the
4974 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4975 program. Exim now does this explicitly; it affects the formatting of
4976 timestamps using strftime().
4978 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4979 complaints. I've now initialized all the variables, to avoid this.
4981 32. Header lines added by a system filter were not being "seen" during
4982 transport-time rewrites.
4984 33. The info_callback() function passed to OpenSSL is set up with type void
4985 (*)(SSL *, int, int), as described somewhere. However, when calling the
4986 function (actually a macro) that sets it up, the type void(*)() is
4987 expected. I've put in a cast to prevent warnings from picky compilers.
4989 34. If a DNS black list lookup found a CNAME record, but there were no A
4990 records associated with the domain it pointed at, Exim crashed.
4992 35. If a DNS black list lookup returned more than one A record, Exim ignored
4993 all but the first. It now scans all returned addresses if a particular IP
4994 value is being sought. In this situation, the contents of the
4995 $dnslist_value variable are a list of all the addresses, separated by a
4998 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4999 to accept a host name and all its aliases if the forward lookup for any of
5000 them yielded the IP address of the incoming connection. Now it accepts only
5001 those names whose forward lookup yields the correct IP address. Any other
5002 names are discarded. This closes a loophole whereby a rogue DNS
5003 administrator could create reverse DNS records to break through a
5004 wildcarded host restriction in an ACL.
5006 37. If a user filter or a system filter that ran in a subprocess used any of
5007 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5008 the wrong values were passed to the pipe command ($thisaddress had the
5009 value of $0, $0 had the value of $1, etc). This bug was introduced by
5010 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5012 38. Improved the line breaking for long SMTP error messages from ACLs.
5013 Previously, if there was no break point between 40 and 75 characters, Exim
5014 left the rest of the message alone. Two changes have been made: (a) I've
5015 reduced the minimum length to 35 characters; (b) if it can't find a break
5016 point between 35 and 75 characters, it looks ahead and uses the first one
5017 that it finds. This may give the occasional overlong line, but at least the
5018 remaining text gets split now.
5020 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5021 file descriptors might be low, and that setting 1000 would always raise it.
5022 It turns out that in some environments, the limit is already over 1000 and
5023 that lowering it causes trouble. So now Exim takes care not to decrease it.
5025 40. When delivering a message, the value of $return_path is set to $sender_
5026 address at the start of routing (routers may change the value). By an
5027 oversight, this default was not being set up when an address was tested by
5028 -bt or -bv, which affected the outcome if any router or filter referred to
5031 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5032 log only when "message" or "log_message" are set. However, if one of the
5033 conditions was an address verification, or a call to a nested ACL, the
5034 messages generated by the underlying test were being passed through. This
5035 no longer happens. The underlying message is available in $acl_verify_
5036 message for both "message" and "log_message" expansions, so it can be
5037 passed through if needed.
5039 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5040 new expansion $bh_ to give the encoded byte string without charset
5041 translation. Translation happens only if iconv() is available; HAVE_ICONV
5042 indicates this at build time. HEADERS_CHARSET gives the charset to
5043 translate to; headers_charset can change it in the configuration, and
5044 "headers charset" can change it in an individual filter file.
5046 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5047 that creates RFC 2047 encoded "words" labels them as that charset instead
5048 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5049 expansion operator; (ii) when Exim creates a From: line for a local
5050 message; (iii) when a header line is rewritten to include a "phrase" part.
5052 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5053 buggy, causing it to skip the first lines of messages whose message ID
5054 ended in 'D'. This would not have bitten before Exim release 4.14, because
5055 message IDs were unlikely to end in 'D' before then. The effect was to have
5056 incorrect size information for certain domains.
5058 45. #include "config.h" was missing at the start of the crypt16.c module. This
5059 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5062 46. If there was a timeout during a "random" callout check, Exim treated it as
5063 a failure of the random address, and carried on sending RSET and the real
5064 address. If the delay was just some slowness somewhere, the response to the
5065 original RCPT would be taken as a response to RSET and so on, causing
5066 mayhem of various kinds.
5068 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5069 when I implemented it. It didn't allow for the fact that some option values
5070 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5071 the right test for positive values.
5073 48. Domain names in DNS records are case-independent. Exim always looks them up
5074 in lower case. Some resolvers return domain names in exactly the case they
5075 appear in the zone file, that is, they may contain uppercase letters. Not
5076 all resolvers do this - some return always lower case. Exim was treating a
5077 change of case by a resolver as a change of domain, similar to a widening
5078 of a domain abbreviation. This triggered its re-routing code and so it was
5079 trying to route what was effectively the same domain again. This normally
5080 caused routing to fail (because the router wouldn't handle the domain
5081 twice). Now Exim checks for this case specially, and just changes the
5082 casing of the domain that it ultimately uses when it transmits the message
5085 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5088 50. If a filter generated a file delivery with a non-absolute name (possible if
5089 no home directory exists for the router), the forbid_file option was not
5092 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5093 the existing equality matching.
5095 52. Exim was using ints instead of ino_t variables in some places where it was
5096 dealing with inode numbers.
5098 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5099 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5100 called TMPDIR, and if it finds it is different, it changes its value.
5102 54. The smtp_printf() function is now made available to local_scan() so
5103 additional output lines can be written before returning. There is also an
5104 smtp_fflush() function to enable the detection of a dropped connection.
5105 The variables smtp_input and smtp_batched_input are exported to
5108 55. Changed the default runtime configuration: the message "Unknown user"
5109 has been removed from the ACL, and instead placed on the localuser router,
5110 using the cannot_route_message feature. This means that any verification
5111 failures that generate their own messages won't get overridden. Similarly,
5112 the "Unrouteable address" message that was in the ACL for unverifiable
5113 relay addresses has also been removed.
5115 56. Added hosts_avoid_esmtp to the smtp transport.
5117 57. The exicyclog script was not checking for the esoteric option
5118 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5119 will work only if exicyclog is run under the appropriate euid.
5121 58. Following a discussion on the list, the rules by which Exim recognises line
5122 endings on incoming messages have been changed. The -dropcr and drop_cr
5123 options are now no-ops, retained only for backwards compatibility. The
5124 following line terminators are recognized: LF CRLF CR. However, special
5125 processing applies to CR:
5127 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5128 nor a local message in the state where . is a terminator.
5130 (ii) If a bare CR is encountered in a header line, an extra space is added
5131 after the line terminator so as not to end the header. The reasoning
5132 behind this is that bare CRs in header lines are most likely either
5133 to be mistakes, or people trying to play silly games.
5135 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5136 was being incorrectly given as 18 bytes larger than it should have been.
5137 This is a VOB (very old bug).
5139 60. This may never have affected anything current, but just in case it has:
5140 When the local host is found other than at the start of a list of hosts,
5141 the local host, those with the same MX, and any that follow, are discarded.
5142 When the list in question was part of a longer list of hosts, the following
5143 hosts (not currently being processed) were also being discarded. This no
5144 longer happens. I'm not sure if this situation could ever has previously
5147 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5150 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5151 header. This is something that is recommended in a new Internet Draft, and
5152 is something that is documented as being done by Sendmail. There are two
5153 possible values. For messages generated by the autoreply transport, Exim
5156 Auto-Submitted: auto-replied
5158 whereas for all other generated messages (e.g. bounces) it adds
5160 Auto-Submitted: auto-generated
5162 63. The "personal" condition in filters now includes a test for the
5163 Auto-Submitted: header. If it contains the string "auto-" the message it
5164 not considered personal.
5166 64. Added rcpt_include_affixes as a generic transport option.
5168 65. Added queue_only_override (default true).
5170 66. Added the syslog_duplication option.
5172 67. If what should have been the first header line of a message consisted of
5173 a space followed by a colon, Exim was mis-interpreting it as a header line.
5174 It isn't of course - it is syntactically invalid and should therefore be
5175 treated as the start of the message body. The misbehaviour could have
5176 caused a number of strange effects, including loss of data in subsequent
5177 header lines, and spool format errors.
5179 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5180 client host had authenticated. This control can now be exercised by an ACL
5181 for more flexibility.
5183 69. By default, callouts do not happen when testing with -bh. There is now a
5184 variant, -bhc, which does actually run the callout code, including
5185 consulting and updating the callout cache.
5187 70. Added support for saslauthd authentication, courtesy of Alexander
5190 71. If statvfs() failed on the spool or log directories while checking their
5191 size for availability, Exim confusingly gave the error "space shortage".
5192 Furthermore, in debugging mode it crashed with a floating point exception.
5193 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5194 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5195 serious problem, Exim now writes to the main and panic logs when this
5196 happens, with details of the failure. It then refuses to accept the
5197 incoming message, giving the message "spool directory problem" or "log
5198 directory problem" with a 421 code for SMTP messages.
5200 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5201 0, 1, and 2 exist, because some OS complain for execs without them (see
5202 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5203 descriptors. However, the code omitted to check that the open succeeded,
5204 causing mysterious errors if for some reason the permissions on /dev/null
5205 got screwed. Now Exim writes a message to the main and panic logs, and
5206 bombs out if it can't open /dev/null.
5208 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5209 interact so that it is all more flexible. It is supposed to remain
5210 backwards compatible. Also added extra_local_interfaces.
5212 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5213 to bomb out with an assertion failure - to the client this appears as a
5214 connection drop. This problem occurs in the part of the code that was taken
5215 from the Samba project. Fortunately, the assertion is in a very simple
5216 function, so I have fixed this by reproducing the function inline in the
5217 one place where it is called, and arranging for authentication to fail
5218 instead of killing the process with assert().
5220 75. The SPA client code was not working when the server requested OEM rather
5221 than Unicode encoding.
5223 76. Added code to make require_files with a specific uid setting more usable in
5224 the case where statting the file as root fails - usually a non-root-mounted
5225 NFS file system. When this happens and the failure is EACCES, Exim now
5226 forks a subprocess and does the per-uid checking as the relevant uid.
5228 77. Added process_log_path.
5230 78. If log_file_path was not explicitly set, a setting of check_log_space or
5231 check_log_inodes was ignored.
5233 79. If a space check for the spool or log partitions fails, the incident is now
5234 logged. Of course, in the latter case the data may get lost...
5236 80. Added the %p formatting code to string_format() so that it can be used to
5237 print addresses in debug_print(). Adjusted all the address printing in the
5238 debugging in store.c to use %p rather than %d.
5240 81. There was a concern that a line of code in smtp_in.c could overflow a
5241 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5242 initially expressed, the concern was not well-founded, because trailing
5243 spaces are removed early. However, if the trailing spaces were followed by
5244 a NULL, they did not get removed, so the overflow was possible. Two fixes
5247 (a) I re-wrote the offending code in a cleaner fashion.
5248 (b) If an incoming SMTP command contains a NULL character, it is rejected
5251 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5252 runs initgroups(), so that if the Exim user is in any additional groups,
5253 they will be used during message reception.
5259 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob