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).
27 TF/01 Failure to get a lock on a hints database can have serious
28 consequences so log it to the panic log.
30 TF/02 Log LMTP confirmation messages in the same way as SMTP,
31 controlled using the smtp_confirmation log selector.
33 TF/03 Include the error message when we fail to unlink a spool file.
35 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
36 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
37 for maintaining out-of-tree patches for some time.
39 PP/01 Bugzilla 139: Documentation and portability issues.
40 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
41 Handle per-OS dynamic-module compilation flags.
43 PP/02 Let /dev/null have normal permissions.
44 The 4.73 fixes were a little too stringent and complained about the
45 permissions on /dev/null. Exempt it from some checks.
46 Reported by Andreas M. Kirchwitz.
48 PP/03 Report version information for many libraries, including
49 Exim version information for dynamically loaded libraries. Created
50 version.h, now support a version extension string for distributors
51 who patch heavily. Dynamic module ABI change.
53 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
54 privilege escalation vulnerability whereby the Exim run-time user
55 can cause root to append content of the attacker's choosing to
58 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
61 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
62 If dropping privileges for untrusted macros, we disabled normal logging
63 on the basis that it would fail; for the Exim run-time user, this is not
64 the case, and it resulted in successful deliveries going unlogged.
65 Fixed. Reported by Andreas Metzler.
71 PP/01 Date: & Message-Id: revert to normally being appended to a message,
72 only prepend for the Resent-* case. Fixes regression introduced in
73 Exim 4.70 by NM/22 for Bugzilla 607.
75 PP/02 Include check_rfc2047_length in configure.default because we're seeing
76 increasing numbers of administrators be bitten by this.
78 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
80 PP/03 Bugzilla 994: added openssl_options main configuration option.
82 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
84 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
86 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
88 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
89 without a peer certificate, leading to a segfault because of an
90 assumption that peers always have certificates. Be a little more
91 paranoid. Problem reported by Martin Tscholak.
93 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
94 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
95 NB: ClamAV planning to remove STREAM in "middle of 2010".
96 CL also introduces -bmalware, various -d+acl logging additions and
97 more caution in buffer sizes.
99 PP/09 Implemented reverse_ip expansion operator.
101 PP/10 Bugzilla 937: provide a "debug" ACL control.
103 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
105 PP/12 Bugzilla 973: Implement --version.
107 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
109 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
111 PP/15 Bugzilla 816: support multiple condition rules on Routers.
113 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
114 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
115 ignore trailing whitespace.
117 JJ/02 prevent non-panic DKIM error from being sent to paniclog
119 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
122 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
123 Notification from Dr Andrew Aitchison.
125 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
126 ExtendedDetectionInfo response format.
127 Notification from John Horne.
129 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
132 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
133 XSL and documented dependency on system catalogs, with examples of how
136 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
139 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
140 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
141 configuration file which is writeable by the Exim user or group.
143 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
144 of configuration files to cover files specified with the -C option if
145 they are going to be used with root privileges, not just the default
148 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
149 option (effectively making it always true).
151 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
152 files to be used while preserving root privileges.
154 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
155 that rogue child processes cannot use them.
157 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
158 run-time user, instead of root.
160 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
161 Exim run-time user without dropping privileges.
163 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
164 result string, instead of calling string_vformat() twice with the same
167 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
168 for other users. Others should always drop root privileges if they use
169 -C on the command line, even for a whitelisted configure file.
171 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
173 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
179 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
180 $data_path, and $header_path variables; fixed documentation bugs and
183 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
184 exipick to access non-standard spools, including the "frozen" queue
187 NM/01 Bugzilla 965: Support mysql stored procedures.
188 Patch from Alain Williams
190 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
192 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
193 Patch from Andreas Metzler
195 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
196 Patch from Kirill Miazine
198 NM/05 Bugzilla 671: Added umask to procmail example.
200 JJ/03 installed exipick 20100323.0, fixing doc bug
202 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
203 directory. Notification and patch from Dan Rosenberg.
205 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
207 TK/02 Improve log output when DKIM signing operation fails.
209 MH/01 Treat the transport option dkim_domain as a colon separated
210 list, not as a single string, and sign the message with each element,
211 omitting multiple occurences of the same signer.
213 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
214 Bugzilla 985, 986. Patch by Simon Arlott
216 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
217 Patch by Simon Arlott
219 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
220 MBX locking. Notification from Dan Rosenberg.
226 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
228 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
230 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
232 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
234 NM/04 Bugzilla 915: Flag broken perl installation during build.
240 TK/01 Added patch by Johannes Berg that expands the main option
241 "spamd_address" if it starts with a dollar sign.
243 TK/02 Write list of recipients to X-Envelope-Sender header when building
244 the mbox-format spool file for content scanning (suggested by Jakob
247 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
248 (http://www.dcc-servers.net/) support via dccifd. Activated by
249 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
251 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
252 by Mark Daniel Reidel <mr@df.eu>.
254 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
255 When building exim an external PCRE library is now needed -
256 PCRE is a system library on the majority of modern systems.
257 See entry on PCRE_LIBS in EDITME file.
259 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
260 conversation. Added nologin parameter to request.
261 Patch contributed by Kirill Miazine.
263 TF/01 Do not log submission mode rewrites if they do not change the address.
265 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
267 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
268 log files in place. Contributed by Roberto Lima.
270 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
272 TF/03 Bugzilla 615: When checking the local_parts router precondition
273 after a local_part_suffix or local_part_prefix option, Exim now
274 does not use the address's named list lookup cache, since this
275 contains cached lookups for the whole local part.
277 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
278 Robert Millan. Documentation is in experimental-spec.txt.
280 TF/04 Bugzilla 668: Fix parallel build (make -j).
282 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
284 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
285 Patch provided by Jan Srzednicki.
287 TF/05 Leading white space used to be stripped from $spam_report which
288 wrecked the formatting. Now it is preserved.
290 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
291 that they are available at delivery time.
293 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
295 TF/08 TLS error reporting now respects the incoming_interface and
296 incoming_port log selectors.
298 TF/09 Produce a more useful error message if an SMTP transport's hosts
299 setting expands to an empty string.
301 NM/06 Bugzilla 744: EXPN did not work under TLS.
302 Patch provided by Phil Pennock.
304 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
305 Patch provided by Richard Godbee.
307 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
308 acl_smtp_notquit, added index entry.
310 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
311 Patch provided by Eugene Bujak.
313 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
314 accept(). Patch provided by Maxim Dounin.
316 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
317 Patch provided by Phil Pennock.
319 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
321 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
322 Patch provided by Brad "anomie" Jorsch.
324 NM/14 Bugzilla 622: Added timeout setting to transport filter.
325 Patch provided by Dean Brooks.
327 TK/05 Add native DKIM support (does not depend on external libraries).
329 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
330 Patch provided by Graeme Fowler.
332 NM/16 Bugzilla 851: Documentation example syntax fix.
334 NM/17 Changed NOTICE file to remove references to embedded PCRE.
336 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
339 NM/19 Bugzilla 745: TLS version reporting.
340 Patch provided by Phil Pennock.
342 NM/20 Bugzilla 167: bool: condition support.
343 Patch provided by Phil Pennock.
345 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
346 clients. Patch provided by Phil Pennock.
348 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
349 Patch provided by Brad "anomie" Jorsch.
351 NM/23 Bugzilla 687: Fix misparses in eximstats.
352 Patch provided by Heiko Schlittermann.
354 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
355 Patch provided by Heiko Schlittermann.
357 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
358 plus update to original patch.
360 NM/26 Bugzilla 799: Documentation correction for ratelimit.
362 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
363 Patch provided by David Brownlee.
365 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
367 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
369 NM/30 Bugzilla 888: TLS documentation bugfixes.
371 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
373 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
374 Unlike the original bugzilla I have changed all shell scripts in src tree.
376 NM/33 Bugzilla 898: Transport filter timeout fix.
377 Patch by Todd Rinaldo.
379 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
380 Patch by Serge Demonchaux.
382 NM/35 Bugzilla 39: Base64 decode bug fixes.
383 Patch by Jakob Hirsch.
385 NM/36 Bugzilla 909: Correct connect() call in dcc code.
387 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
389 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
391 NM/39 Bugzilla 911: Fixed MakeLinks build script.
397 TK/01 Add preliminary DKIM support. Currently requires a forked version of
398 ALT-N's libdkim that I have put here:
399 http://duncanthrax.net/exim-experimental/
401 Note to Michael Haardt: I had to rename some vars in sieve.c. They
402 were called 'true' and it seems that C99 defines that as a reserved
403 keyword to be used with 'bool' variable types. That means you could
404 not include C99-style headers which use bools without triggering
405 build errors in sieve.c.
407 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
408 as mailq or other aliases. Changed the --help handling significantly
409 to do whats expected. exim_usage() emits usage/help information.
411 SC/01 Added the -bylocaldomain option to eximstats.
413 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
415 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
417 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
423 PH/01 Another patch from the Sieve maintainer.
425 PH/02 When an IPv6 address is converted to a string for single-key lookup
426 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
427 dots are used instead of colons so that keys in lsearch files need not
428 contain colons. This was done some time before quoting was made available
429 in lsearch files. However, iplsearch files do require colons in IPv6 keys
430 (notated using the quote facility) so as to distinguish them from IPv4
431 keys. This meant that lookups for IP addresses in host lists did not work
432 for iplsearch lookups.
434 This has been fixed by arranging for IPv6 addresses to be expressed with
435 colons if the lookup type is iplsearch. This is not incompatible, because
436 previously such lookups could never work.
438 The situation is now rather anomolous, since one *can* have colons in
439 ordinary lsearch keys. However, making the change in all cases is
440 incompatible and would probably break a number of configurations.
442 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
445 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
446 conversion specification without a maximum field width, thereby enabling
447 a rogue spamd server to cause a buffer overflow. While nobody in their
448 right mind would setup Exim to query an untrusted spamd server, an
449 attacker that gains access to a server running spamd could potentially
450 exploit this vulnerability to run arbitrary code as the Exim user.
452 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
453 $primary_hostname instead of what libspf2 thinks the hosts name is.
455 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
456 a directory entry by the name of the lookup key. Previously, if a
457 symlink pointed to a non-existing file or a file in a directory that
458 Exim lacked permissions to read, a lookup for a key matching that
459 symlink would fail. Now it is enough that a matching directory entry
460 exists, symlink or not. (Bugzilla 503.)
462 PH/03 The body_linecount and body_zerocount variables are now exported in the
465 PH/04 Added the $dnslist_matched variable.
467 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
468 This means they are set thereafter only if the connection becomes
471 PH/06 Added the client_condition to authenticators so that some can be skipped
472 by clients under certain conditions.
474 PH/07 The error message for a badly-placed control=no_multiline_responses left
475 "_responses" off the end of the name.
477 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
479 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
480 (without spaces) instead of just copying the configuration text.
482 PH/10 Added the /noupdate option to the ratelimit ACL condition.
484 PH/11 Added $max_received_linelength.
486 PH/12 Added +ignore_defer and +include_defer to host lists.
488 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
489 way in which PCRE > 7.0 is built.
491 PH/14 Implemented queue_only_load_latch.
493 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
494 MAIL command. The effect was to mangle the value on 64-bit systems.
496 PH/16 Another patch from the Sieve maintainer.
498 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
500 PH/18 If a system quota error occurred while trying to create the file for
501 a maildir delivery, the message "Mailbox is full" was not appended to the
502 bounce if the delivery eventually timed out. Change 4.67/27 below applied
503 only to a quota excession during the actual writing of the file.
505 PH/19 It seems that peer DN values may contain newlines (and other non-printing
506 characters?) which causes problems in log lines. The DN values are now
507 passed through string_printing() before being added to log lines.
509 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
510 and InterBase are left for another time.)
512 PH/21 Added message_body_newlines option.
514 PH/22 Guard against possible overflow in moan_check_errorcopy().
516 PH/23 POSIX allows open() to be a macro; guard against that.
518 PH/24 If the recipient of an error message contained an @ in the local part
519 (suitably quoted, of course), incorrect values were put in $domain and
520 $local_part during the evaluation of errors_copy.
526 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
527 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
530 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
531 issue a MAIL command.
533 PH/02 In an ACL statement such as
535 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
537 if a client was not listed at all, or was listed with a value other than
538 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
539 the condition was not true (as it should be), so access was not denied.
540 The bug was that the ! inversion was incorrectly passed on to the second
541 item. This has been fixed.
543 PH/03 Added additional dnslists conditions == and =& which are different from
544 = and & when the dns lookup returns more than one IP address.
546 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
547 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
549 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
550 FSYNC, which compiles an option called disable_fsync that allows for
551 bypassing fsync(). The documentation is heavily laced with warnings.
553 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
555 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
556 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
557 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
558 including adding "make clean"; (3) Added -fPIC when compiling the test
559 dynamically loaded module, to get rid of a warning.
561 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
562 message fails, move_frozen_messages = true and ignore_bounce_errors_after
563 = 0s. The bug is otherwise harmless.
565 PH/07 There was a bug in the dovecot authenticator such that the value of
566 $auth1 could be overwritten, and so not correctly preserved, after a
567 successful authentication. This usually meant that the value preserved by
568 the server_setid option was incorrect.
570 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
572 PH/09 Installed PCRE release 7.0.
574 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
575 run for batched SMTP input. It is now run at the start of every message
576 in the batch. While fixing this I discovered that the process information
577 (output by running exiwhat) was not always getting set for -bs and -bS
578 input. This is fixed, and it now also says "batched" for BSMTP.
580 PH/11 Added control=no_pipelining.
582 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
583 patch, slightly modified), and move the expansion of helo_data till after
584 the connection is made in the smtp transport (so it can use these
587 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
589 PH/14 Added log_selector = +pid.
591 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
593 PH/16 Add ${if forany and ${if forall.
595 PH/17 Added dsn_from option to vary the From: line in DSNs.
597 PH/18 Flush SMTP output before performing a callout, unless control =
598 no_callout_flush is set.
600 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
601 was true (the default) a successful delivery failed to delete the retry
602 item, thus causing premature timeout of the address. The bug is now
605 PH/20 Added hosts_avoid_pipelining to the smtp transport.
607 PH/21 Long custom messages for fakedefer and fakereject are now split up
608 into multiline reponses in the same way that messages for "deny" and
609 other ACL rejections are.
611 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
612 with slight modification.
614 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
615 draft, changing the syntax and factoring some duplicate code".
617 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
618 for deliveries of the second and subsequent messages over the same SMTP
621 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
622 ${reduce, with only minor "tidies".
624 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
626 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
627 expansion side effects.
629 PH/27 When a message times out after an over-quota error from an Exim-imposed
630 quota, the bounce message says "mailbox is full". This message was not
631 being given when it was a system quota that was exceeded. It now should
634 MH/03 Made $recipients available in local_scan(). local_scan() already has
635 better access to the recipient list through recipients_list[], but
636 $recipients can be useful in postmaster-provided expansion strings.
638 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
639 in the case of a MAIL command with additional options following the
640 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
641 were accidentally chopped off.
643 PH/29 SMTP synchronization checks are implemented when a command is read -
644 there is a check that no more input is waiting when there shouldn't be
645 any. However, for some commands, a delay in an ACL can mean that it is
646 some time before the response is written. In this time, more input might
647 arrive, invalidly. So now there are extra checks after an ACL has run for
648 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
649 pipelining has not been advertised.
651 PH/30 MH's patch to allow iscntrl() characters to be list separators.
653 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
654 returned in the SMTP response when smtp_return_error_details was false.
657 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
658 instead of the C I/O that was originally supplied, because problems were
661 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
662 Exim which did not show up earlier: it was assuming that a call to
663 SSL_CTX_set_info_callback() might give an error value. In fact, there is
664 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
665 was a macro that became an assignment, so it seemed to work. This has
666 changed to a proper function call with a void return, hence the compile
667 error. Exim's code has been fixed.
669 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
672 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
674 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
676 PH/37 If a message is not accepted after it has had an id assigned (e.g.
677 because it turns out to be too big or there is a timeout) there is no
678 "Completed" line in the log. When some messages of this type were
679 selected by exigrep, they were listed as "not completed". Others were
680 picked up by some special patterns. I have improved the selection
681 criteria to be more general.
683 PH/38 The host_find_failed option in the manualroute router can now be set
684 to "ignore", to completely ignore a host whose IP address cannot be
685 found. If all hosts are ignored, the behaviour is controlled by the new
686 host_all_ignored option.
688 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
689 homing or because of multiple MX records with /mx) generated more than
690 one IP address, and the following item turned out to be the local host,
691 all the secondary addresses of the first item were incorrectly removed
692 from the list, along with the local host and any following hosts (which
693 is what is supposed to happen).
695 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
696 whoever called Exim into the -H file. In the case of the daemon it was
697 behaving confusingly. When first started, it used values for whoever
698 started the daemon, but after a SIGHUP it used the Exim user (because it
699 calls itself on a restart). I have changed the code so that it now always
702 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
703 message are rejected with the same error (e.g. no authentication or bad
704 sender address), and a DATA command is nevertheless sent (as can happen
705 with PIPELINING or a stupid MUA), the error message that was given to the
706 RCPT commands is included in the rejection of the DATA command. This is
707 intended to be helpful for MUAs that show only the final error to their
710 PH/42 Another patch from the Sieve maintainer.
712 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
713 Eximstats - Fixed some broken HTML links and added missing column headers
715 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
716 columns for Rejects, Temp Rejects, Ham, and Spam rows.
718 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
720 PH/43 Yet another patch from the Sieve maintainer.
722 PH/44 I found a way to check for a TCP/IP connection going away before sending
723 the response to the final '.' that terminates a message, but only in the
724 case where the client has not sent further data following the '.'
725 (unfortunately, this is allowed). However, in many cases there won't be
726 any further data because there won't be any more messages to send. A call
727 to select() can be used: if it shows that the input is "ready", there is
728 either input waiting, or the socket has been closed. An attempt to read
729 the next input character can distinguish the two cases. Previously, Exim
730 would have sent an OK response which the client would never have see.
731 This could lead to message repetition. This fix should cure that, at
732 least in a lot of common cases.
734 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
735 advertised in response to EHLO.
741 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
742 fixed by 4.65/MH/01 (is this a record?) are fixed:
744 (i) An empty string was always treated as zero by the numeric comparison
745 operators. This behaviour has been restored.
747 (ii) It is documented that the numeric comparison operators always treat
748 their arguments as decimal numbers. This was broken in that numbers
749 starting with 0 were being interpreted as octal.
751 While fixing these problems I realized that there was another issue that
752 hadn't been noticed. Values of message_size_limit (both the global option
753 and the transport option) were treated as octal if they started with 0.
754 The documentation was vague. These values are now always treated as
755 decimal, and I will make that clear in the documentation.
761 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
762 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
765 MH/01 Don't check that the operands of numeric comparison operators are
766 integers when their expansion is in "skipping" mode (fixes bug
767 introduced by 4.64-PH/07).
769 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
770 child addresses, Exim now panics and dies. Previously, because the count
771 is held in a short int, deliveries were likely to be lost. As such a
772 large number of recipients for a single message is ridiculous
773 (performance will be very, very poor), I have chosen to impose a limit
774 rather than extend the field.
780 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
781 leftover -K file (the existence of which was triggered by #402).
782 While we were at it, introduced process PID as part of the -K
783 filename. This should rule out race conditions when creating
786 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
787 processing considerably. Previous code took too long for large mails,
788 triggering a timeout which in turn triggers #401.
790 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
791 in the DK code in transports.c. sendfile() is not really portable,
792 hence the _LINUX specificness.
794 TF/01 In the add_headers option to the mail command in an Exim filter,
795 there was a bug that Exim would claim a syntax error in any
796 header after the first one which had an odd number of characters
799 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
800 callout verification, Exim cached a "reject" for the entire domain. This
801 is correct for most verifications, but it is not correct for a recipient
802 verification with use_sender or use_postmaster set, because in that case
803 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
804 case of MAIL FROM:<> rejection from other early rejections (e.g.
805 rejection of HELO). When verifying a recipient using a non-null MAIL
806 address, the cache is ignored if it shows MAIL FROM:<> rejection.
807 Whatever the result of the callout, the value of the domain cache is
808 left unchanged (for any other kind of callout, getting as far as trying
809 RCPT means that the domain itself is ok).
811 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
814 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
815 manifest itself as EPIPE rather than ECONNECT. When tidying away a
816 session, the daemon ignores ECONNECT errors and logs others; it now
817 ignores EPIPE as well.
819 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
820 (quoted-printable decoding).
822 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
823 later the small subsequent patch to fix an introduced bug.
825 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
827 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
829 PH/08 An error is now given if message_size_limit is specified negative.
831 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
832 to be given (somewhat) arbitrary names.
834 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
837 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
838 miscellaneous code fixes
840 PH/10 Added the log_reject_target ACL modifier to specify where to log
843 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
844 hostname. This is wrong, because it relates to the incoming message (and
845 probably the interface on which it is arriving) and not to the outgoing
846 callout (which could be using a different interface). This has been
847 changed to use the value of the helo_data option from the smtp transport
848 instead - this is what is used when a message is actually being sent. If
849 there is no remote transport (possible with a router that sets up host
850 addresses), $smtp_active_hostname is used.
852 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
853 tweaks were necessary in order to get it to work (see also 21 below):
854 (a) The code assumed that strncpy() returns a negative number on buffer
855 overflow, which isn't the case. Replaced with Exim's string_format()
857 (b) There were several signed/unsigned issues. I just did the minimum
858 hacking in of casts. There is scope for a larger refactoring.
859 (c) The code used strcasecmp() which is not a standard C function.
860 Replaced with Exim's strcmpic() function.
861 (d) The code set only $1; it now sets $auth1 as well.
862 (e) A simple test gave the error "authentication client didn't specify
863 service in request". It would seem that Dovecot has changed its
864 interface. Fortunately there's a specification; I followed it and
865 changed what the client sends and it appears to be working now.
867 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
870 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
871 address A is aliased to B and C, where B exists and C does not. Without
872 -v the output is "A verified" because verification stops after a
873 successful redirection if more than one address is generated. However,
874 with -v the child addresses are also verified. Exim was outputting "A
875 failed to verify" and then showing the successful verification for C,
876 with its parentage. It now outputs "B failed to verify", showing B's
877 parentage before showing the successful verification of C.
879 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
880 look up a TXT record in a specific list after matching in a combined
883 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
884 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
885 they consult the DNS. I had assumed they would set it the way they
886 wanted; and indeed my experiments on Linux seem to show that in some
887 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
888 To be on the safe side, however, I have now made the interface to
889 host_find_byname() similar to host_find_bydns(), with an argument
890 containing the DNS resolver options. The host_find_byname() function now
891 sets these options at its start, just as host_find_bydns() does. The smtp
892 transport options dns_qualify_single and dns_search_parents are passed to
893 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
894 of host_find_byname() use the default settings of RES_DEFNAMES
895 (qualify_single) but not RES_DNSRCH (search_parents).
897 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
898 spool_read_header() do less string testing, by means of a preliminary
899 switch on the second character of optional "-foo" lines. (This is
900 overdue, caused by the large number of possibilities that now exist.
901 Originally there were few.) While I was there, I also converted the
902 str(n)cmp tests so they don't re-test the leading "-" and the first
903 character, in the hope this might squeeze out yet more improvement.
905 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
906 flag allowing group syntax was set by the header_syntax check but not
907 turned off, possible causing trouble later; (2) The flag was not being
908 set at all for the header_verify test, causing "group"-style headers to
909 be rejected. I have now set it in this case, and also caused header_
910 verify to ignore an empty address taken from a group. While doing this, I
911 came across some other cases where the code for allowing group syntax
912 while scanning a header line wasn't quite right (mostly, not resetting
913 the flag correctly in the right place). These bugs could have caused
914 trouble for malformed header lines. I hope it is now all correct.
916 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
917 with the "reply" argument non-NULL. The code, however (which originally
918 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
919 but it didn't always do it. This confused somebody who was copying the
920 code for some other use. I have removed all the tests.
922 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
923 feature that was used to support insecure browsers during the U.S. crypto
924 embargo. It requires special client support, and Exim is probably the
925 only MTA that supported it -- and would never use it because real RSA is
926 always available. This code has been removed, because it had the bad
927 effect of slowing Exim down by computing (never used) parameters for the
928 RSA_EXPORT functionality.
930 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
931 authenticator to fail if there's a tab character in the incoming data
932 (there should never be unless someone is messing about, as it's supposed
933 to be base64-encoded). Also added, on Timo's advice, the "secured" option
934 if the connection is using TLS or if the remote IP is the same as the
935 local IP, and the "valid-client-cert option" if a client certificate has
938 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
939 authenticators. This can be used for authorization after authentication
940 succeeds. (In the case of plaintext, it servers for both authentication
943 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
944 if any retry times were supplied.
946 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
947 connection, where there is no client IP address to check. In this
948 situation, the verify now always succeeds.
950 PH/25 Applied John Jetmore's -Mset patch.
952 PH/26 Added -bem to be like -Mset, but loading a message from a file.
954 PH/27 In a string expansion for a processed (not raw) header when multiple
955 headers of the same name were present, leading whitespace was being
956 removed from all of them, but trailing whitespace was being removed only
957 from the last one. Now trailing whitespace is removed from each header
958 before concatenation. Completely empty headers in a concatenation (as
961 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
962 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
964 PH/29 [Removed. This was a change that I later backed out, and forgot to
965 correct the ChangeLog entry (that I had efficiently created) before
966 committing the later change.]
968 PH/30 Exim was sometimes attempting to deliver messages that had suffered
969 address errors (4xx response to RCPT) over the same connection as other
970 messages routed to the same hosts. Such deliveries are always "forced",
971 so retry times are not inspected. This resulted in far too many retries
972 for the affected addresses. The effect occurred only when there were more
973 hosts than the hosts_max_try setting in the smtp transport when it had
974 the 4xx errors. Those hosts that it had tried were not added to the list
975 of hosts for which the message was waiting, so if all were tried, there
976 was no problem. Two fixes have been applied:
978 (i) If there are any address or message errors in an SMTP delivery, none
979 of the hosts (tried or untried) are now added to the list of hosts
980 for which the message is waiting, so the message should not be a
981 candidate for sending over the same connection that was used for a
982 successful delivery of some other message. This seems entirely
983 reasonable: after all the message is NOT "waiting for some host".
984 This is so "obvious" that I'm not sure why it wasn't done
985 previously. Hope I haven't missed anything, but it can't do any
986 harm, as the worst effect is to miss an optimization.
988 (ii) If, despite (i), such a delivery is accidentally attempted, the
989 routing retry time is respected, so at least it doesn't keep
990 hammering the server.
992 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
993 in ${readsocket because some servers need this prod.
995 PH/32 Added some extra debug output when updating a wait-xxx database.
997 PH/33 The hint "could be header name not terminated by colon", which has been
998 given for certain expansion errors for a long time, was not being given
999 for the ${if def:h_colon_omitted{... case.
1001 PH/34 The spec says: "With one important exception, whenever a domain list is
1002 being scanned, $domain contains the subject domain." There was at least
1003 one case where this was not true.
1005 PH/35 The error "getsockname() failed: connection reset by peer" was being
1006 written to the panic log as well as the main log, but it isn't really
1007 panic-worthy as it just means the connection died rather early on. I have
1008 removed the panic log writing for the ECONNRESET error when getsockname()
1011 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1012 runs only) independently of the message's sender address. This meant
1013 that, if the 4xx error was in fact related to the sender, a different
1014 message to the same recipient with a different sender could confuse
1015 things. In particualar, this can happen when sending to a greylisting
1016 server, but other circumstances could also provoke similar problems.
1017 I have changed the default so that the retry time for these errors is now
1018 based a combination of the sender and recipient addresses. This change
1019 can be overridden by setting address_retry_include_sender=false in the
1022 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1023 remote server are returned as part of bounce messages. This was not
1024 happening for LMTP over a pipe (the lmtp transport), but now it is the
1025 same for both kinds of LMTP.
1027 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1028 in header lines that were in fact CNAMEs. This is no longer the case.
1030 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1031 and queue runs started by the daemon processed all messages. This has
1032 been fixed so that -R and -S can now usefully be given with -q<time>.
1034 PH/40 Import PCRE release 6.7 (fixes some bugs).
1036 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1038 PH/42 Give an error if -q is specified more than once.
1040 PH/43 Renamed the variables $interface_address and $interface_port as
1041 $received_ip_address and $received_port, to make it clear that these
1042 values apply to message reception, and not to the outgoing interface when
1043 a message is delivered. (The old names remain recognized, of course.)
1045 PH/44 There was no timeout on the connect() call when using a Unix domain
1046 socket in the ${readsocket expansion. There now is.
1048 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1049 be meaningful with "accept".
1051 SC/01 Eximstats V1.43
1052 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1054 SC/02 Eximstats V1.44
1055 Use a glob alias rather than an array ref in the generated
1056 parser. This improves both readability and performance.
1058 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1059 Collect SpamAssassin and rejection statistics.
1060 Don't display local sender or destination tables unless
1061 there is data to show.
1062 Added average volumes into the top table text output.
1064 SC/04 Eximstats V1.46
1065 Collect data on the number of addresses (recipients)
1066 as well as the number of messages.
1068 SC/05 Eximstats V1.47
1069 Added 'Message too big' to the list of mail rejection
1070 reasons (thanks to Marco Gaiarin).
1072 SC/06 Eximstats V1.48
1073 Mainlog lines which have GMT offsets and are too short to
1074 have a flag are now skipped.
1076 SC/07 Eximstats V1.49 (Alain Williams)
1077 Added the -emptyok flag.
1079 SC/08 Eximstats V1.50
1080 Fixes for obtaining the IP address from reject messages.
1082 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1083 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1084 whitesspace changes from 4.64-PH/27
1086 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1089 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1090 are found, allow negative numbers in numeric criteria)
1092 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1094 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1095 to match changes made in 4.64-PH/43
1097 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1099 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1100 "domain" is actually an IP address, and force a failure. This locks out
1101 those revolvers/nameservers that support "A-for-A" lookups, in
1102 contravention of the specifications.
1104 PH/48 When a host name was looked up from an IP address, and the subsequent
1105 forward lookup of the name timed out, the host name was left in
1106 $sender_host_name, contrary to the specification.
1108 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1109 restricted to single-key lookups, Exim was not diagnosing an error if
1110 * or *@ was used with a query-style lookup.
1112 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1114 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1115 long ago, but noone interested enough thought of it. Let's just say that
1116 the "1.1" means that there are some new functions that weren't there at
1117 some point in the past.
1119 PH/51 Error processing for expansion failure of helo_data from an smtp
1120 transport during callout processing was broken.
1122 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1123 tested/used via the -bh/-bhc/-bs options.
1125 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1126 bug, fixed in subsequent PCRE releases).
1128 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1129 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1131 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1137 SC/01 Use a glob alias rather than an array ref in eximstats generated
1138 parser. This improves both readability and performance.
1140 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1141 Don't display local sender or destination tables in eximstats unless
1142 there is data to show.
1143 Added average volumes into the eximstats top table text output.
1145 SC/03 Collect data on the number of addresses (recipients) as well
1146 as the number of messages in eximstats.
1148 TF/01 Correct an error in the documentation for the redirect router. Exim
1149 does (usually) call initgroups() when daemonizing.
1151 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1152 with consistent privilege compared to when running as a daemon.
1154 TF/03 Note in the spec that $authenticated_id is not set for local
1155 submissions from trusted users.
1157 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1158 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1160 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1161 by adding some example configuration directives to the default
1162 configuration file. A little bit of work is required to uncomment the
1163 directives and define how usernames and passwords are checked, but
1164 there is now a framework to start from.
1166 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1167 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1168 without this. I don't know how relevant this is to other LDAP libraries.
1170 PH/02 Add the verb name to the "unknown ACL verb" error.
1172 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1174 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1176 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1177 directory not expanded when it should be if an expanded home directory
1178 was set for the address (which is overridden by the transport).
1180 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1183 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1184 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1185 because it is too late at that time, and has no effect.
1187 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1188 security issue with \' (bugzilla #107). I could not use the
1189 PQescapeStringConn() function, because it needs a PGconn value as one of
1192 PH/08 When testing addresses using -bt, indicate those final addresses that
1193 are duplicates that would not cause an additional delivery. At least one
1194 person was confused, thinking that -bt output corresponded to deliveries.
1195 (Suppressing duplicates isn't a good idea as you lose the information
1196 about possibly different redirections that led to the duplicates.)
1198 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1199 systems where poll() doesn't work, in particular OS X.
1201 PH/10 Added more information to debugging output for retry time not reached.
1203 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1204 operations in malware.c.
1206 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1209 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1210 syslog_duplication set false, log lines that would normally be written
1211 both the the main log and to the reject log were not written to syslog at
1214 PH/14 In the default configuration, change the use of "message" in ACL warn
1215 statements to "add_header".
1217 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1218 not followed by a command (e.g. "seen endif").
1220 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1221 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1224 PH/17 Added extra conditions to the default value of delay_warning_condition
1228 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1229 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1230 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1233 The Auto-Submitted: and various List- headers are standardised, whereas I
1234 don't think Precedence: ever was.
1236 PH/18 Refactored debugging code in route_finduser() to show more information,
1237 in particular, the error code if getpwnam() issues one.
1239 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1240 This is apparently needed in addition to the PH/07 change above to avoid
1241 any possible encoding problems.
1243 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1244 but not after initializing Perl.
1246 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1247 output them only if debugging. By default they are written stderr,
1248 apparently, which is not desirable.
1250 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1253 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1256 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1258 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1259 authenticated or an ident call has been made. Suppress the default
1260 values for $authenticated_id and $authenticated_sender (but permit -oMai
1261 and -oMas) when testing with -bh.
1263 PH/24 Re-jigged the order of the tests in the default configuration so that the
1264 tests for valid domains and recipients precede the DNS black list and CSA
1265 tests, on the grounds that those ones are more expensive.
1267 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1268 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1269 This bug exists in every version of Exim that I still have, right back to
1272 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1273 However, an attempt to turn on case-sensitivity in a regex key by
1274 including (?-i) didn't work because the subject string was already
1275 lowercased, and the effects were non-intuitive. It turns out that a
1276 one-line patch can be used to allow (?-i) to work as expected.
1282 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1283 other effects) broke the use of negated acl sub-conditions.
1285 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1288 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1289 "Deny" causes Exim to reject the incoming connection with a 554 error.
1290 Unfortunately, if there is a major crisis, such as a disk failure,
1291 tcp-wrappers gives "deny", whereas what one would like would be some
1292 kind of temporary error. A kludge has been added to help with this.
1293 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1294 554 error is used if errno is still zero or contains ENOENT (which occurs
1295 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1298 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1300 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1301 errors. Otherwise a message that provokes a temporary error (when other
1302 messages do not) can cause a whole host to time out.
1304 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1305 addresses were routed directly to files or pipes from a redirect router.
1306 File deliveries just didn't batch; pipe deliveries might have suffered
1309 PH/06 A failure to get a lock for a hints database would erroneously always say
1310 "Failed to get write lock", even when it was really a read lock.
1312 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1313 of 0600. This has been changed to use the value of the lockfile_mode
1314 option (which defaults to 0600).
1316 PH/08 Applied small patch from the Sieve maintainer.
1318 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1319 folder from quota calculations, a direct delivery into this folder messed
1320 up the contents of the maildirsize file. This was because the regex was
1321 used only to exclude .Trash (or whatever) when the size of the mailbox
1322 was calculated. There was no check that a delivery was happening into an
1323 excluded directory. This bug has been fixed by ignoring all quota
1324 processing for deliveries into excluded directories.
1326 PH/10 Added the maildirfolder_create_regex option to appendfile.
1332 PH/01 The code for finding all the local interface addresses on a FreeBSD
1333 system running IPv6 was broken. This may well have applied to all BSD
1334 systems, as well as to others that have similar system calls. The broken
1335 code found IPv4 interfaces correctly, but gave incorrect values for the
1336 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1337 that it would not match correctly against @[] and not recognize the IPv6
1340 PH/02 The ipliteral router was not recognizing addresses of the form user@
1341 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1343 PH/03 Added disable_ipv6.
1345 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1346 decoded form, because it is most often used to construct To: headers
1347 lines in autoreplies, and the decoded form may well be syntactically
1348 invalid. However, $reply_address has leading white space removed, and all
1349 newlines turned into spaces so that the autoreply transport does not
1352 PH/05 If group was specified without a user on a router, and no group or user
1353 was specified on a transport, the group from the router was ignored.
1355 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1356 for visible compile-time settings that can be used to change these
1357 numbers, for those that want even more. Backwards compatibility with old
1358 spool files has been maintained. However, going back to a previous Exim
1359 release will lost any variables that are in spool files.
1361 PH/07 Two small changes when running in the test harness: increase delay when
1362 passing a TCP/IP connection to a new process, in case the original
1363 process has to generate a bounce, and remove special handling of
1364 127.0.0.2 (sic), which is no longer necessary.
1366 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1367 be the same on different OS.
1369 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1372 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1373 whether --show-vars was specified or not
1375 JJ/02 exipick: Added support for new ACL variable spool format introduced
1378 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1379 syntactically invalid From: or Reply-to: line, and a filter used this to
1380 generate an autoreply, and therefore failed to obtain an address for the
1381 autoreply, Exim could try to deliver to a non-existent relative file
1382 name, causing unrelated and misleading errors. What now happens is that
1383 it logs this as a hard delivery error, but does not attempt to create a
1386 PH/11 The exinext utility has a -C option for testing purposes, but although
1387 the given file was scanned by exinext itself; it wasn't being passed on
1388 when Exim was called.
1390 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1391 an end-of-file indication when reading a command response.
1393 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1394 compiled. In many other places in Exim, IPv6 addresses are always
1395 recognized, so I have changed this. It also means that IPv4 domain
1396 literals of the form [IPV4:n.n.n.n] are now always recognized.
1398 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1399 used if the router is not running as root, for example, when verifying at
1400 ACL time, or when using -bh. The debugging output from this situation was
1401 non-existent - all you got was a failure to exec. I have made two
1404 (a) Failures to set uid/gid, the current directory, or a process leader
1405 in a subprocess such as that created by queryprogram now generate
1406 suitable debugging ouput when -d is set.
1408 (b) The queryprogram router detects when it is not running as root,
1409 outputs suitable debugging information if -d is set, and then runs
1410 the subprocess without attempting to change uid/gid.
1412 PH/15 Minor change to Makefile for building test_host (undocumented testing
1415 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1416 additional section of a DNS packet that returns MX or SRV records.
1417 Instead, it always explicitly searches for A/AAAA records. This avoids
1418 major problems that occur when a DNS server includes only records of one
1419 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1420 fixed another bug: if SRV records were looked up and the corresponding
1421 address records were *not* found in the additional section, the port
1422 values from the SRV records were lost.
1424 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1425 using the correct key (the original address) when searching the retry
1426 rules in order to find which one to use for generating the retry hint.
1428 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1429 adding the default one. Similarly, if it contains a Reply-To: header, the
1430 errors_reply_to option, if set, is not used.
1432 PH/19 When calculating a retry time, Exim used to measure the "time since
1433 failure" by looking at the "first failed" field in the retry record. Now
1434 it does not use this if it is later than than the arrival time of the
1435 message. Instead it uses the arrival time. This makes for better
1436 behaviour in cases where some deliveries succeed, thus re-setting the
1437 "first failed" field. An example is a quota failure for a huge message
1438 when small messages continue to be delivered. Without this change, the
1439 "time since failure" will always be short, possible causing more frequent
1440 delivery attempts for the huge message than are intended.
1441 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1443 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1444 $1, $2, $3) because the numerical variables can be reset during some
1445 expansion items (e.g. "match"), thereby losing the authentication data.
1447 PH/21 Make -bV show the size of off_t variables so that the test suite can
1448 decide whether to run tests for quotas > 2G.
1450 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1451 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1452 filecount value is greater than 2G or if a quota value is greater than 2G
1453 on a system where the size of off_t is not greater than 4, a panic error
1456 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1457 never match. The debug and -bh output now contains an explicit error
1458 message indicating a malformed IPv4 address or mask.
1460 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1461 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1462 PH/23 above applies.
1464 PH/25 Do not write to syslog when running in the test harness. The only
1465 occasion when this arises is a failure to open the main or panic logs
1466 (for which there is an explicit test).
1468 PH/26 Added the /no_tell option to "control=freeze".
1470 PH/27 If a host name lookup failed very early in a connection, for example, if
1471 the IP address matched host_lookup and the reverse lookup yielded a name
1472 that did not have a forward lookup, an error message of the form "no IP
1473 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1474 could be logged. Now it outputs the IP address instead of "NULL".
1476 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1477 allows the sender and the authenticated sender to be set when
1478 submitting a message from within Exim. Since child_open_exim() is
1479 documented for local_scan(), the new function should be too.
1481 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1482 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1483 results in an empty string is now treated as unset.
1485 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1487 PH/31 Added sender_verify_fail logging option.
1489 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1490 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1491 tidied the source and removed it altogether.
1493 PH/33 When a queue run was abandoned because the load average was too high, a
1494 log line was always written; now it is written only if the queue_run log
1495 selector is set. In addition, the log line for abandonment now contains
1496 information about the queue run such as the pid. This is always present
1497 in "start" and "stop" lines but was omitted from the "abandon" line.
1499 PH/34 Omit spaces between a header name and the colon in the error message that
1500 is given when verify = headers_syntax fails (if there are lots of them,
1501 the message gets confusing).
1503 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1504 names, as there are now some PTR records that contain slashes. This check
1505 is only to protect against broken name servers that fall over on strange
1506 characters, so the fact that it applies to all lookups doesn't matter.
1508 PH/36 Now that the new test suite is complete, we can remove some of the
1509 special code in Exim that was needed for the old test suite. For example,
1510 sorting DNS records because real resolvers return them in an arbitrary
1511 order. The new test suite's fake resolver always returns records in the
1514 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1515 bounces) except when queue_only is set, to avoid logging races between
1516 the different processes.
1518 PH/38 Panic-die if .include specifies a non-absolute path.
1520 PH/39 A tweak to the "H" retry rule from its user.
1522 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1523 a label. They prevented compilation on older perls.
1525 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1526 a warning to be raised on newish perls.
1528 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1529 on queue. Changes to match documented behaviour of showing count of
1530 messages matching specified criteria.
1532 PH/40 Changed the default ident timeout from 30s to 5s.
1534 PH/41 Added support for the use of login_cap features, on those BSD systems
1535 that have them, for controlling the resources used by pipe deliveries.
1537 PH/42 The content-scanning code uses fopen() to create files in which to put
1538 message data. Previously it was not paying any attention to the mode of
1539 the files. Exim runs with umask(0) because the rest of the code creates
1540 files with open(), and sets the required mode explicitly. Thus, these
1541 files were ending up world-writeable. This was not a big issue, because,
1542 being within the spool directory, they were not world-accessible. I have
1543 created a function called modefopen, which takes an additional mode
1544 argument. It sets umask(777), creates the file, chmods it to the required
1545 mode, then resets the umask. All the relevant calls to fopen() in the
1546 content scanning code have been changed to use this function.
1548 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1549 to 24 hours. This avoids potential overflow problems when processing G
1550 and H retry rules. I suspect nobody ever tinkers with this value.
1552 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1554 PH/45 When the plaintext authenticator is running as a client, the server's
1555 challenges are checked to ensure they are valid base64 strings. By
1556 default, the authentication attempt is cancelled if an invalid string is
1557 received. Setting client_ignore_invalid_base64 true ignores these errors.
1558 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1559 they are received. Thus, the responses can be made to depend on the
1560 challenges. If an invalid string is ignored, an empty string is placed in
1563 PH/46 Messages that are created by the autoreply transport now contains a
1564 References: header, in accordance with RFCs 2822 and 3834.
1566 PH/47 Added authenticated_sender_force to the smtp transport.
1568 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1570 PH/49 Installed latest patch from the Sieve maintainer.
1572 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1573 was also set, the appendfile transport was unnecessarily scanning a
1574 directory of message files (e.g. for maildir delivery) to find the count
1575 of files (along with the size), even though it did not need this
1576 information. It now does the scan only if it needs to find either the
1577 size of the count of files.
1579 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1581 PH/52 Two bugs concerned with error handling when the smtp transport is
1584 (i) Exim was not creating retry information for temporary errors given
1585 for individual recipients after the DATA command when the smtp transport
1586 was used in LMTP mode. This meant that they could be retried too
1587 frequently, and not timed out correctly.
1589 (ii) Exim was setting the flag that allows error details to be returned
1590 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1591 recipients that were returned after the DATA command.
1593 PH/53 This is related to PH/52, but is more general: for any failing address,
1594 when detailed error information was permitted to be returned to the
1595 sender, but the error was temporary, then after the final timeout, only
1596 "retry timeout exceeded" was returned. Now it returns the full error as
1597 well as "retry timeout exceeded".
1599 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1600 do this, and (what is worse) MTAs that accept it.
1602 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1603 will now be deprecated.
1605 PH/56 New os.c-cygwin from the Cygwin maintainer.
1607 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1608 formats (previously only available in exim formats via -bpr, -bpru,
1609 and -bpra. Now also available in native and exiqgrep formats)
1611 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1612 with very large, slow to parse queues
1614 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1616 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1618 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1619 responses to authentication challenges, though it was showing the
1620 challenges; (ii) I've removed the CR characters from the debug output for
1623 PH/58 Allow for the insertion of a newline as well as a space when a string
1624 is turned into more than one encoded-word during RFC 2047 encoding. The
1625 Sieve code now uses this.
1627 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1628 data_4xx, lost_connection, tls_required.
1630 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1631 message was being sent as an SMTP response.
1633 PH/61 Add -l and -k options to exicyclog.
1635 PH/62 When verifying, if an address was redirected to one new address, so that
1636 verification continued, and the new address failed or deferred after
1637 having set something in $address_data, the value of $address_data was not
1638 passed back to the ACL. This was different to the case when no
1639 redirection occurred. The value is now passed back in both cases.
1641 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1642 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1643 use login_cap.h, so on its own it isn't the distinguishing feature. The
1644 new name refers directly to the setclassresources() function.
1646 PH/65 Added configuration files for NetBSD3.
1648 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1650 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1651 is preferred over IPv4.
1653 PH/68 The bounce_return_message and bounce_return_body options were not being
1654 honoured for bounces generated during the reception of non-SMTP messages.
1655 In particular, this applied to messages rejected by the ACL. This bug has
1656 been fixed. However, if bounce_return_message is true and bounce_return_
1657 body is false, the headers that are returned for a non-SMTP message
1658 include only those that have been read before the error was detected.
1659 (In the case of an ACL rejection, they have all been read.)
1661 PH/69 The HTML version of the specification is now built in a directory called
1662 spec_html instead of spec.html, because the latter looks like a path with
1663 a MIME-type, and this confuses some software.
1665 PH/70 Catch two compiler warnings in sieve.c.
1667 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1668 function verify_get_ident() calls ip_connect() to connect a socket, but
1669 if the "connect()" function timed out, ip_connect() used to close the
1670 socket. However, verify_get_ident() also closes the socket later, and in
1671 between Exim writes to the log, which may get opened at this point. When
1672 the socket was closed in ip_connect(), the log could get the same file
1673 descriptor number as the socket. This naturally causes chaos. The fix is
1674 not to close the socket in ip_connect(); the socket should be closed by
1675 the function that creates it. There was only one place in the code where
1676 this was missing, in the iplookup router, which I don't think anybody now
1677 uses, but I've fixed it anyway.
1679 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1680 well as to direct DNS lookups. Otherwise the handling of names in host
1681 lists is inconsistent and therefore confusing.
1687 PH/01 Two changes to the default runtime configuration:
1689 (1) Move the checks for relay_from_hosts and authenticated clients from
1690 after to before the (commented out) DNS black list checks.
1692 (2) Add control=submission to the relay_from_hosts and authenticated
1693 clients checks, on the grounds that messages accepted by these
1694 statements are most likely to be submissions.
1696 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1698 (1) Generate an error if the third argument for the ${prvs expansion is
1701 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1704 (3) Reset the variables that are obtained from the first argument of
1705 ${prvscheck and used in the second argument before leaving the code,
1706 because their memory is reclaimed, so using them afterwards may do
1709 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1710 one (it's much easier than Tom thought :-).
1712 (5) Because of (4), we can now allow for the use of $prvscheck_result
1713 inside the third argument.
1715 PH/03 For some reason, the default setting of PATH when running a command from
1716 a pipe transport was just "/usr/bin". I have changed it to
1719 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1720 anything to be listed in the output from -bV.
1722 PH/05 When a filter generated an autoreply, the entire To: header line was
1723 quoted in the delivery log line, like this:
1725 => >A.N.Other <ano@some.domain> <original@ddress> ...
1727 This has been changed so that it extracts the operative address. There
1728 may be more than one such address. If so, they are comma-separated, like
1731 => >ano@some.domain,ona@other.domain <original@ddress> ...
1733 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1734 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1735 not being looked up in the rDNS to get a host name, Exim was showing the
1736 IP address twice in Received: lines, even though the IP addresses were
1737 identical. For example:
1739 Received: from [1.2.3.4] (helo=[1.2.3.4])
1741 However, if the real host name was known, it was omitting the HELO data
1742 if it matched the actual IP address. This has been tidied up so that it
1743 doesn't show the same IP address twice.
1745 PH/07 When both +timestamp and +memory debugging was on, the value given by
1746 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1747 called by the debug printing, thereby overwriting the timestamp buffer.
1748 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1750 PH/08 When the original message was included in an autoreply transport, it
1751 always said "this is a copy of the message, including all the headers",
1752 even if body_only or headers_only was set. It now gives an appropriate
1755 PH/09 Applied a patch from the Sieve maintainer which:
1757 o fixes some comments
1758 o adds the (disabled) notify extension core
1759 o adds some debug output for the result of if/elsif tests
1760 o points to the current vacation draft in the documentation
1761 and documents the missing references header update
1765 o fixes a bug in processing the envelope test (when testing
1766 multiple envelope elements, the last element determinted the
1769 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1770 Electronic Mail") by including:
1772 Auto-submitted: auto-generated
1774 in the messages that it generates (bounce messages and others, such as
1775 warnings). In the case of bounce messages for non-SMTP mesages, there was
1776 also a typo: it was using "Auto_submitted" (underscore instead of
1777 hyphen). Since every message generated by Exim is necessarily in response
1778 to another message, thes have all been changed to:
1780 Auto-Submitted: auto-replied
1782 in accordance with these statements in the RFC:
1784 The auto-replied keyword:
1786 - SHOULD be used on messages sent in direct response to another
1787 message by an automatic process,
1789 - MUST NOT be used on manually-generated messages,
1791 - MAY be used on Delivery Status Notifications (DSNs) and Message
1792 Disposition Notifications (MDNs),
1794 - MUST NOT be used on messages generated by automatic or periodic
1795 processes, except for messages which are automatic responses to
1798 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1799 to the default Received: header definition.
1801 PH/12 Added log selector acl_warn_skipped (default on).
1803 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1804 variables because (a) they are in the wrong storage pool and (b) even if
1805 they were copied, it wouldn't work properly because of the caching.
1807 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1808 checking when decoding. Apparently there are clients that generate
1809 overlong encoded strings. Why am I not surprised?
1811 PH/15 If the first argument of "${if match_address" was not empty, but did not
1812 contain an "@" character, Exim crashed. Now it writes a panic log message
1813 and treats the condition as false.
1815 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1817 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1818 extension "envelope-auth". The code is finished and in agreement with
1819 other implementations, but there is no documentation so far and in fact,
1820 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1821 not changing the active code.
1823 Print executed "if" and "elsif" statements when debugging is used. This
1824 helps a great deal to understand what a filter does.
1826 Document more things not specified clearly in RFC3028. I had all this
1827 sorted out, when out of a sudden new issues came to my mind. Oops."
1829 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1832 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1833 canonical form (as documented). However, after a host name lookup from
1834 the IP address, check_host() was doing a simple string comparison with
1835 addresses acquired from the DNS when checking that the found name did
1836 have the original IP as one of its addresses. Since any found IPv6
1837 addresses are likely to be in abbreviated form, the comparison could
1838 fail. Luckily, there already exists a function for doing the comparison
1839 by converting both addresses to binary, so now that is used instead of
1840 the text comparison.
1842 PH/20 There was another similar case to PH/19, when a complete host name was
1843 given in a host list; looking up its IP address could give an abbreviated
1844 form, whereas the current host's name might or might not be abbreviated.
1845 The same fix has been applied.
1851 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1852 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1855 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1856 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1858 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1860 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1861 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1862 Also, for "Auto-Submitted", treat anything other than "no" as
1863 non-personal, in accordance with RFC 3834. (Previously it treated
1864 anything starting "auto-" as non-personal.)
1866 TF/01 The control=submission/name=... option had a problem with syntax
1867 errors if the name included a slash character. The /name= option
1868 now slurps the rest of the string, so it can include any characters
1869 but it must come last in the list of options (after /sender_retain
1872 PH/05 Some modifications to the interface to the fake nameserver for the new
1880 TK/01 Added the "success_on_redirect" address verification option. See
1881 NewStuff for rationale and an example.
1883 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1885 PH/02 Patch to exigrep to allow it to work on syslog lines.
1887 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1888 fread() to skip over the body file's header line, because in Cygwin the
1889 header line is locked and is inaccessible.
1891 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1892 co-exist for some time) to make it clear that it is the Exim ID that is
1893 referenced, not the Message-ID: header line.
1895 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1896 string_format() function, because snprintf() does not exist on all
1899 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1900 ${stat: expansion item.
1902 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1903 protocol synchronization error", to keep the pedants happy.
1905 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1906 well as for IRIX systems, when gcc is being used. See the host.c source
1909 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1911 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1914 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1915 transport and to the smtp transport in LMTP mode.
1917 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1919 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1920 run a filter in a subprocess. This could lead to confusion in subsequent
1921 lookups in the parent process. There should also be a search_tidyup() at
1922 the end of the subprocess.
1924 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1925 only if the host matched helo_try_verify_hosts, which caused the
1926 verification to occur when the EHLO/HELO command was issued. The ACL just
1927 tested the remembered result. Now, if a previous verification attempt has
1928 not happened, "verify = helo" does it there and then.
1930 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1932 TK/03 Fix log output including CR from clamd.
1934 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1935 exist provoked a memory error which could cause a segfault.
1937 PH/15 Installed PCRE 6.2
1939 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1941 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1942 of the problem. Specifically, suggested +O2 rather than +O1 for the
1945 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1947 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1948 record, the port was not correctly logged when the outgoing_port log
1949 selector was set (it logged the transort's default port).
1951 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1952 fallback_hosts, and "hosts" in the smtp transport.
1954 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1955 host errors such as "Connection refused".
1957 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1958 authentication with radiusclient 0.4.9:
1960 - Error returned from rc_read_config was caught wrongly
1961 - Username/password not passed on to radius server due to wrong length.
1963 The presumption is that some radiusclient API changes for 4.51/PH/17
1964 were not taken care of correctly. The code is still untested by me (my
1965 Linux distribution still has 0.3.2 of radiusclient), but it was
1966 contributed by a Radius user.
1968 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1969 expanding the "port" option of the smtp transport.
1971 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1972 while reading a MIME header. Thanks to Tom Hughes for a patch.
1974 PH/24 Include config.h inside local_scan.h so that configuration settings are
1977 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1978 that in an ACL for RCPT (for example), you can examine exactly what was
1981 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1982 commands, but it was not correctly comparing the address with the actual
1983 client host address. Thus, it would show the EHLO address in Received:
1984 header lines when this was not necessary.
1986 PH/27 Added the % operator to ${eval:}.
1988 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1989 it should be ignoring failures (because with -C, for example, it has lost
1990 privilege). It wasn't ignoring creation failures other than "already
1993 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1996 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1997 because an input line was too long, either on its own, or by virtue of
1998 too many continuations, the temporary file was not being removed, and the
1999 return code was incorrect.
2001 PH/31 Missing "BOOL" in function definition in filtertest.c.
2003 PH/32 Applied Sieve patches from the maintainer.
2005 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2007 PH/33 Added "verify = not_blind".
2009 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2010 Local/Makefile (with some defaults set). These are used in built scripts
2011 such as exicyclog, but they have never been used in the exim_install
2012 script (though there are many overriding facilities there). I have
2013 arranged that the exim_install script now takes note of these two
2016 PH/35 Installed configuration files for Dragonfly.
2018 PH/36 When a locally submitted message by a trusted user did not contain a
2019 From: header, and the sender address was obtained from -f or from an SMTP
2020 MAIL command, and the trusted user did not use -F to supply a sender
2021 name, $originator_name was incorrectly used when constructing a From:
2022 header. Furthermore, $originator_name was used for submission mode
2023 messages from external hosts without From: headers in a similar way,
2024 which is clearly wrong.
2026 PH/37 Added control=suppress_local_fixups.
2028 PH/38 When log_selector = +received_sender was set, and the addition of the
2029 sender made the log line's construction buffer exactly full, or one byte
2030 less than full, an overflow happened when the terminating "\n" was
2033 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2034 when the result of a list match is failure because a DNS lookup failed.
2036 PH/40 RM_COMMAND is now used in the building process.
2038 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2039 the "build-* directories that it finds.
2041 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2042 address in a domain literal was a prefix of an interface address.
2044 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2045 when verifying a sender address, unless rewrite_headers is false.
2047 PH/44 Wrote a long comment about why errors_to addresses are verified as
2048 recipients, not senders.
2050 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2051 the ratelimit ACL was added.
2053 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2055 PH/46 Added extra information about PostgreSQL errors to the error string.
2057 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2058 avoiding the need to install special zones in a real server. This is
2059 backwards compatible; if it can't find the fake resolver, it drops back.
2060 Thus, both old and new test suites can be run.
2062 TF/02 Added util/ratelimit.pl
2064 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2065 clock is set back in time.
2067 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2068 Candler <B.Candler@pobox.com>.
2070 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2071 OK for addresses that are the result of redirections.
2073 PH/48 A number of further additions for the benefit of the new test suite,
2074 including a fake gethostbyname() that interfaces to the fake DNS resolver
2077 TF/06 The fix for widen_domains has also been applied to qualify_single and
2078 search_parents which are the other dnslookup options that can cause
2081 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2084 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2086 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2087 service" if it encountered an MX record with an empty target hostname.
2088 The message is now "an MX or SRV record indicated no SMTP service".
2090 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2091 if the DNS of the sending site is misconfigured. This is quite a
2092 common situation. This change restores the behaviour of treating a
2093 helo verification defer as a failure.
2095 PH/51 If self=fail was set on a router, the bounce message did not include the
2096 actual error message.
2102 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2104 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2105 command itself ended in error, the underlying message about the transport
2106 filter timeout was being overwritten with the pipe command error. Now the
2107 underlying error message should be appended to the second error message.
2109 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2111 PH/02 Reduce the amount of output that "make" produces by default. Full output
2112 can still be requested.
2114 PH/03 The warning log line about a condition test deferring for a "warn" verb
2115 was being output only once per connection, rather than after each
2116 occurrence (because it was using the same function as for successful
2117 "warn" verbs). This seems wrong, so I have changed it.
2119 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2120 it should not have, which might have caused a crash in the right
2121 circumstances, but probably never did.
2123 PH/04 Installed a modified version of Tony Finch's patch to make submission
2124 mode fix the return path as well as the Sender: header line, and to
2125 add a /name= option so that you can make the user's friendly name appear
2128 TF/03 Added the control = fakedefer ACL modifier.
2130 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2131 Mark Lowes for thorough testing.
2133 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2135 TK/03 Merged latest SRS patch from Miles Wilton.
2137 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2138 with the definition in sysexits.h (which is #included earlier).
2139 Fortunately, Exim does not actually use EX_OK. The code used to try to
2140 preserve the sysexits.h value, by assumimg that macro definitions were
2141 scanned for macro replacements. I have been disabused of this notion,
2142 so now the code just undefines EX_OK before #including unistd.h.
2144 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2145 in the smtp transport. When a block could not be written in a single
2146 write() function, the timeout was being re-applied to each part-write.
2147 This seems wrong - if the receiver was accepting one byte at a time it
2148 would take for ever. The timeout is now adjusted when this happens. It
2149 doesn't have to be particularly precise.
2151 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2152 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2154 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2155 without a domain if the check to <postmaster@domain> fails.
2157 SC/01 Eximstats: added -xls and the ability to specify output files
2158 (patch written by Frank Heydlauf).
2160 SC/02 Eximstats: use FileHandles for outputing results.
2162 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2164 SC/04 Eximstats: fixed display of large numbers with -nvr option
2166 SC/05 Eximstats: fixed merging of reports with empty tables.
2168 SC/06 Eximstats: added the -include_original_destination flag
2170 SC/07 Eximstats: removed tabs and trailing whitespace.
2172 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2174 TK/06 MBOX spool code: Add real "From " MBOX separator line
2175 so the .eml file is really in mbox format (even though
2176 most programs do not really care). Patch from Alex Miller.
2178 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2179 The latter is generated from $received_to and is only set if the
2180 message has one envelope recipient. SA can use these headers,
2181 obviously out-of-the-box. Patch from Alex Miller.
2183 PH/08 The ${def test on a variable was returning false if the variable's
2184 value was "0", contrary to what the specification has always said!
2185 The result should be true unless the variable is empty.
2187 PH/09 The syntax error of a character other than { following "${if
2188 def:variable_name" (after optional whitespace) was not being diagnosed.
2189 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2190 accidental colon was present, for example, could give incorrect results.
2192 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2193 result is used (not including appendfile, where other changes are about
2196 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2197 This involved changing a lot of size variables from int to off_t. It
2198 should work with maildirs and everything.
2200 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2201 spamd dying while we are connected to it.
2203 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2206 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2207 function for BDB 4.3.
2209 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2211 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2212 the sources for PCRE have been re-arranged and more files are now
2215 PH/15 The code I had for printing potentially long long variables in PH/11
2216 above was not the best (it lost precision). The length of off_t variables
2217 is now inspected at build time, and an appropriate printing format (%ld
2218 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2219 to be "long long int" or "long int". This is needed for the internal
2220 formatting function string_vformat().
2222 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2223 the configuration file to be ":syslog", then the script "guesses" where
2224 the logs files are, rather than using the compiled in default. In our
2225 case the guess is not the same as the compiled default, so the script
2226 suddenly stopped working when I started to use syslog. The patch checks
2227 to see if log_file_path is "". If so, it attempts to read it from exim
2228 with no configuration file to get the compiled in version, before it
2229 falls back to the previous guessing code."
2231 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2232 implementing BATV in an Exim configuration. See NewStuff for the gory
2235 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2236 Makefile that are specific to HP-UX.
2238 PH/18 If the "use_postmaster" option was set for a recipient callout together
2239 with the "random" option, the postmaster address was used as the MAIL
2240 FROM address for the random test, but not for the subsequent recipient
2241 test. It is now used for both.
2243 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2244 patch removes a few documentation additions to RFC 3028, because the
2245 latest draft now contains them. It adds the new en;ascii-case comparator
2246 and a new error check for 8bit text in MIME parts. Comparator and
2247 require names are now matched exactly. I enabled the subaddress
2248 extension, but it is not well tested yet (read: it works for me)."
2250 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2251 rework some of the code of TK/09 above to avoid the hardwired use of
2252 "%lld" and "long long". Replaced the call to snprintf() with a call to
2255 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2256 records point to non-existent hosts", "retry timeout exceeded", and
2257 "retry time not reached for any host after a long failure period".
2259 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2260 experimental DomainKeys support:
2262 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2263 (2) On an error such as an illegally used "control", the wrong name for
2264 the control was given.
2266 These problems did NOT occur unless DomainKeys support was compiled.
2268 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2270 PH/24 Added ${if match_ip condition.
2272 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2273 databases so that it will be absolutely obvious if a crash occurs in the
2274 DB library. This is a regular occurrence (often caused by mis-matched
2277 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2278 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2279 picked up on a user's system that detects such things. There doesn't seem
2280 to be a gcc warning option for this - only an attribute that has to be
2281 put on the function's prototype. It seems that in Fedora Core 4 they have
2282 set this on a number of new functions. No doubt there will be more in due
2285 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2286 specify a transport. However, if an address that was verified by such a
2287 router was the subject of a callout, Exim crashed because it tried to
2288 read the rcpt_include_affixes from the non-existent transport. Now it
2289 just assumes that the setting of that option is false. This bug was
2290 introduced by 4.51/PH/31.
2292 PH/28 Changed -d+all to exclude +memory, because that information is very
2293 rarely of interest, but it makes the output a lot bigger. People tend to
2294 do -d+all out of habit.
2296 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2297 code in os-type was giving problems when libc.so lives in lib64, like on
2300 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2301 aren't the modern standard, and it seems that some systems' include files
2302 don't always have them. Exim was already checking for some of the newer
2303 ones like T_AAAA, and defining it itself. I've added checks for all the
2304 record types that Exim uses.
2306 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2307 not automatically generating a new one, as it is supposed to. This
2308 prevented TLS from working. If the file did exist, but contained invalid
2309 data, a new version was generated, as expected. It was only the case of a
2310 non-existent file that was broken.
2312 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2313 with a change in libdomainkeys > 0.64.
2315 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2316 from DNS. If the selector record carries the flag, it now has
2317 precedence over the domain-wide flag.
2319 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2321 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2322 the use of a port name that isn't defined in /etc/services) occurred, the
2323 message was deferred as in a normal delivery, and thus remained on the
2324 spool, instead of being failed because of the mua_wrapper setting. This
2325 is now fixed, and I tidied up some of the mua_wrapper messages at the
2328 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2329 the messages in a hash of arrays rather than using individual hashes.
2330 This is a bit cleaner and results in dramatic memory savings, albeit
2331 at a slight CPU cost.
2333 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2334 as requested by Marc Sherman.
2336 SC/10 Eximstats: added histograms for user specified patterns as requested
2339 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2341 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2342 fopen() in the content-scanning modules that did not already have it.
2348 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2349 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2351 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2353 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2355 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2356 to test Sieve filters that use "vacation".
2358 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2359 that changes the way the GnuTLS parameters are stored in the cache file.
2360 The new format can be generated externally. For backward compatibility,
2361 if the data in the cache doesn't make sense, Exim assumes it has read an
2362 old-format file, and it generates new data and writes a new file. This
2363 means that you can't go back to an older release without removing the
2366 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2367 work if there are any delivery delays because "one_time" forces the
2368 parent to be marked "delivered", so its unseen clone is never tried
2369 again. For this reason, Exim now forbids the simultaneous setting of
2372 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2373 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2374 entry if you want to know the details. The fix, however, neglected to
2375 consider the case when local delivery batching is involved. The test for
2376 "previously delivered" was not happening when checking to see if an
2377 address could be batched with a previous (undelivered) one; under
2378 certain circumstances this could lead to multiple deliveries to the same
2381 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2382 in its include files, and this causes problems building Exim.
2384 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2385 header_syntax) but Exim was just ignoring anything given after a slash.
2386 In particular, this caused confusion with an attempt to use "verify =
2387 reverse_host_lookup/defer_ok". An error is now given when options are
2388 supplied for verify items that do not have them. (Maybe reverse_host_
2389 lookup should have a defer_ok option, but that's a different point.)
2391 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2392 defined by RFC 821) to 2048, because there were problems with some AUTH
2393 commands, and RFC 1869 says the size should be increased for extended
2394 SMTP commands that take arguments.
2396 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2399 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2400 "unknown" error; now it says that the functionality isn't in the binary.
2402 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2403 an address' error message when a string expansion fails (syntax or
2404 whatever). Otherwise the password may appear in the log. Following change
2405 PH/42 below, there is no longer a chance of it appearing in a bounce
2408 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2410 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2411 of its addresses was ever tried. (Bugzilla bug #2.)
2413 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2414 the result incorrectly in the debug output. (It correctly added a newline
2415 to what was transported.)
2417 TF/01 Added $received_time.
2419 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2420 commented out examples of how to interface to a virus scanner and to
2421 SpamAssassin. Also added commented examples of av_scanner and
2422 spamd_address settings.
2424 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2425 and controls are allowed in which ACLs. There were a couple of minor
2426 errors. Some of the entries in the conditions table (which is a table of
2427 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2428 negation of where the condition IS allowed.
2430 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2432 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2433 header file does not have a version number, so I've had to invent a new
2434 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2435 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2436 radiusclient), but it was contributed by a Radius user.
2438 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2439 files or directories, for OpenSSL.
2441 PH/19 When an Exim process that is running as root has to create an Exim log
2442 file, it does so in a subprocess that runs as exim:exim so as to get the
2443 ownership right at creation (otherwise, other Exim processes might see
2444 the file with the wrong ownership). There was no test for failure of this
2445 fork() call, which would lead to the process getting stuck as it waited
2446 for a non-existent subprocess. Forks do occasionally fail when resources
2447 run out. I reviewed all the other calls to fork(); they all seem to check
2450 PH/20 When checking for unexpected SMTP input at connect time (before writing
2451 the banner), Exim was not dealing correctly with a non-positive return
2452 from the read() function. If the client had disconnected by this time,
2453 the result was a log entry for a synchronization error with an empty
2454 string after "input=" when read() returned zero. If read() returned -1
2455 (an event I could not check), uninitialized data bytes were printed.
2456 There were reports of junk text (parts of files, etc) appearing after
2459 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2461 PH/22 Added support for macro redefinition, and (re)definition in between
2462 driver and ACL definitions.
2464 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2465 forgetting to use the resulting value; it was using the unexpanded value.
2467 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2468 hadn't been configured. The fix is from Juergen Kreileder, who
2469 understands it better than I do:
2471 "Here's what I see happening with three configured cyrus_sasl
2472 authenticators configured (plain, login, cram-md5):
2474 On startup auth_cyrus_sasl_init() gets called for each of these.
2475 This means three calls to sasl_listmech() without a specified mech_list.
2476 => SASL tests which mechs of all available mechs actually work
2477 => three warnings about OTP not working
2478 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2480 With the patch, sasl_listmech() also gets called three times. But now
2481 SASL's mech_list option is set to the server_mech specified in the the
2482 authenticator. Or in other words, the answer from sasl_listmech()
2483 gets limited to just the mech you're testing for (which is different
2485 => the return list contains just 'plain' or 'login', 'cram-md5' or
2486 nothing depending on the value of ob->server_mech.
2488 I've just tested the patch: Authentication still works fine,
2489 unavailable mechs specified in the exim configuration are still
2490 caught, and the auth.log warnings about OTP are gone."
2492 PH/25 When debugging is enabled, the contents of the command line are added
2493 to the debugging output, even when log_selector=+arguments is not
2496 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2497 answer is "GNU", and only if the return is "GNU/something" is the answer
2500 PH/27 $acl_verify_message is now set immediately after the failure of a
2501 verification in an ACL, and so is available in subsequent modifiers. In
2502 particular, the message can be preserved by coding like this:
2504 warn !verify = sender
2505 set acl_m0 = $acl_verify_message
2507 Previously, $acl_verify_message was set only while expanding "message"
2508 and "log_message" when a very denied access.
2510 PH/28 Modified OS/os.c-Linux with
2512 -#ifndef OS_LOAD_AVERAGE
2513 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2515 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2516 nomenclature these days.)
2518 PH/29 Installed patch from the Sieve maintainer that adds the options
2519 sieve_useraddress and sieve_subaddress to the redirect router.
2521 PH/30 In these circumstances:
2522 . Two addresses routed to the same list of hosts;
2523 . First host does not offer TLS;
2524 . First host accepts first address;
2525 . First host gives temporary error to second address;
2526 . Second host offers TLS and a TLS session is established;
2527 . Second host accepts second address.
2528 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2529 and peerdn, if requested) that were in fact used only for the second
2532 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2533 attention to any local part prefix or suffix that was matched by the
2534 router that accepted the address. It now behaves in the same way as it
2535 does for delivery: the affixes are removed from the local part unless
2536 rcpt_include_affixes is set on the transport.
2538 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2539 timeout during the DATA phase of an incoming message.
2541 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2542 applied a patch sanctioned by the Sieve maintainer.
2544 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2545 the uid or gid is negative. A case of a negative gid caused this to be
2546 noticed. The fix allows for either to be negative.
2548 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2549 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2552 PH/36 The change PH/12 above was broken. Fixed it.
2554 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2555 the grounds that routing the same address twice would always produce the
2556 same answer. This might have been true once, but it is certainly no
2557 longer true now. Routing a child address may depend on the previous
2558 routing that produced that child. Some complicated redirection strategies
2559 went wrong when messages had multiple recipients, and made Exim's
2560 behaviour dependent on the order in which the addresses were given.
2562 I have moved the duplicate checking until after the routing is complete.
2563 Exim scans the addresses that are assigned to local and remote
2564 transports, and removes any duplicates. This means that more work will be
2565 done, as duplicates will always all be routed, but duplicates are
2566 presumably rare, so I don't expect this is of any significance.
2568 For deliveries to pipes, files, and autoreplies, the duplicate checking
2569 still happens during the routing process, since they are not going to be
2572 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2573 It corrects a timeout issue with spamd. This is Ian's comment: "The
2574 background is that sometimes spamd either never reads data from a
2575 connection it has accepted, or it never writes response data. The exiscan
2576 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2577 blindly assumes that writes won't block so it may never time out."
2579 PH/39 Allow G after quota size as well as K and M.
2581 PH/40 The value set for $authenticated_id in an authenticator may not contain
2582 binary zeroes or newlines because the value is written to log lines and
2583 to spool files. There was no check on this. Now the value is run through
2584 the string_printing() function so that such characters are converted to
2585 printable escape sequences.
2587 PH/41 $message_linecount is a new variable that contains the total number of
2588 lines in the message. Compare $body_linecount, which is the count for the
2591 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2592 bounce and delay warning messages, except in certain special cases, which
2595 (a) An SMTP error message from a remote host;
2596 (b) A message specified in a :fail: redirection;
2597 (c) A message specified in a "fail" command in a system filter;
2598 (d) A message specified in a FAIL return from the queryprogram router;
2599 (e) A message specified by the cannot_route_message router option.
2601 In these cases only, Exim does include the error details in bounce and
2602 warning messages. There are also a few cases where bland messages such
2603 as "unrouteable address" or "local delivery error" are given.
2605 PH/43 $value is now also set for the "else" part of a ${run expansion.
2607 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2608 being worked on, but at least Exim now implements the latest version to
2611 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2612 process to complete was treated as a delivery failure, a timeout while
2613 writing the message to the pipe was logged, but erroneously treated as a
2614 successful delivery. Such timeouts include transport filter timeouts. For
2615 consistency with the overall process timeout, these timeouts are now
2616 treated as errors, giving rise to delivery failures by default. However,
2617 there is now a new Boolean option for the pipe transport called
2618 timeout_defer, which, if set TRUE, converts the failures into defers for
2619 both kinds of timeout. A transport filter timeout is now identified in
2622 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2623 systems where "make" and "gmake" are different, calling "gmake" at top
2624 level broke things. I've arranged for the value of $(MAKE) to be passed
2625 from the Makefile to this script so that it can call the same version of
2629 A note about Exim versions 4.44 and 4.50
2630 ----------------------------------------
2632 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2633 changes of various kinds. As a consequence, a big documentation update was
2634 needed. This delayed the release for rather longer than seemed good, especially
2635 in the light of a couple of (minor) security issues. Therefore, the changes
2636 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2637 release. So 4.44 and 4.50 are in effect two different branches that both start
2640 I have left the 4.50 change log unchanged; it contains all the changes since
2641 4.43. The change log for 4.44 is below; many of its items are identical to
2642 those for 4.50. This seems to be the most sensible way to preserve the
2643 historical information.
2649 1. Minor wording change to the doc/README.SIEVE file.
2651 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2652 computation of the current number of files was incorrect.
2654 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2655 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2658 4. Give more explanation in the error message when the command for a transport
2659 filter fails to execute.
2661 5. There are several places where Exim runs a non-Exim command in a
2662 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2663 was being done only for the command run by the queryprogram router. It is
2664 now done for all such subprocesses. The other cases are: ${run, transport
2665 filters, and the commands run by the lmtp and pipe transports.
2667 6. Added CONFIGURE_GROUP build-time option.
2669 7. Some older OS have a limit of 256 on the maximum number of file
2670 descriptors. Exim was using setrlimit() to set 1000 as a large value
2671 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2672 systems. I've change it so that if it can't get 1000, it tries for 256.
2674 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2675 was an oversight, and furthermore, ever since the addition of extra
2676 controls (e.g. 4.43/32), the checks on when to allow different forms of
2677 "control" were broken. There should now be diagnostics for all cases when a
2678 control that does not make sense is encountered.
2680 9. Added the /retain_sender option to "control=submission".
2682 10. $recipients is now available in the predata ACL (oversight).
2684 11. Tidy the search cache before the fork to do a delivery from a message
2685 received from the command line. Otherwise the child will trigger a lookup
2686 failure and thereby defer the delivery if it tries to use (for example) a
2687 cached ldap connection that the parent has called unbind on.
2689 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2690 of $address_data from the recipient verification was clobbered by the
2691 sender verification.
2693 13. The value of address_data from a sender verification is now available in
2694 $sender_address_data in subsequent conditions in the ACL statement.
2696 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2698 15. Added a new option "connect=<time>" to callout options, to set a different
2701 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2702 was its contents. (It was OK if the option was not defined at all.)
2704 17. A "Completed" log line is now written for messages that are removed from
2705 the spool by the -Mrm option.
2707 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2708 information about exactly what failed.
2710 19. Added -dd to debug only the daemon process.
2712 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2713 handles timeouts, both on the server side and network timeouts. Renamed the
2714 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2716 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2717 It is now set to "smtps".
2719 22. $host_address is now set to the target address during the checking of
2720 ignore_target_hosts.
2722 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2723 being passed; this would have caused $sender_host_name to have been used if
2724 matching the list had actually called for a host name (not very likely,
2725 since this list is usually IP addresses). A host name is now passed as
2728 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2729 code that specifies a non-restarting handler (typically sigaction() in
2730 modern systems) in an attempt to fix a rare and obscure crash bug.
2732 25. Narrowed the window for a race in the daemon that could cause it to ignore
2733 SIGCHLD signals. This is not a major problem, because they are used only to
2734 wake it up if nothing else does.
2736 26. A malformed maildirsize file could cause Exim to calculate negative values
2737 for the mailbox size or file count. Odd effects could occur as a result.
2738 The maildirsize information is now recalculated if the size or filecount
2741 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2742 support for a long time. Removed HAVE_SYS_VFS_H.
2744 28. Installed the latest version of exipick from John Jetmore.
2746 29. In an address list, if the pattern was not a regular expression, an empty
2747 subject address (from a bounce message) matched only if the pattern was an
2748 empty string. Non-empty patterns were not even tested. This was the wrong
2749 because it is perfectly reasonable to use an empty address as part of a
2750 database query. An empty address is now tested by patterns that are
2751 lookups. However, all the other forms of pattern expect the subject to
2752 contain a local part and a domain, and therefore, for them, an empty
2753 address still always fails if the pattern is not itself empty.
2755 30. Exim went into a mad DNS loop when attempting to do a callout where the
2756 host was specified on an smtp transport, and looking it up yielded more
2757 than one IP address.
2759 31. Re-factored the code for checking spool and log partition space into a
2760 function that finds that data and another that does the check. The former
2761 is then used to implement four new variables: $spool_space, $log_space,
2762 $spool_inodes, and $log_inodes.
2764 32. The RFC2047 encoding function was originally intended for short strings
2765 such as real names; it was not keeping to the 75-character limit for
2766 encoded words that the RFC imposes. It now respects the limit, and
2767 generates multiple encoded words if necessary. To be on the safe side, I
2768 have increased the buffer size for the ${rfc2047: expansion operator from
2771 33. It is now permitted to omit both strings after an "if" condition; if the
2772 condition is true, the result is "true". As before, when the second string
2773 is omitted, a false condition yields an empty string. This makes it less
2774 cumbersome to write custom ACL and router conditions.
2776 34. Failure to deliver a bounce message always caused it to be frozen, even if
2777 there was an errors_to setting on the router. The errors_to setting is now
2780 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2781 canonical form (fully expanded) before being placed in
2782 $sender_host_address.
2784 36. The table in the code that translates DNS record types into text (T_A to
2785 "A" for instance) was missing entries for NS and CNAME. It is just possible
2786 that this could have caused confusion if both these types were looked up
2787 for the same domain, because the text type is used as part of Exim's
2788 per-process caching. But the chance of anyone hitting this buglet seems
2791 37. The dnsdb lookup has been extended in a number of ways.
2793 (1) There is a new type, "zns", which walks up the domain tree until it
2794 finds some nameserver records. It should be used with care.
2796 (2) There is a new type, "mxh", which is like "mx" except that it returns
2797 just the host names, not the priorities.
2799 (3) It is now possible to give a list of domains (or IP addresses) to be
2800 looked up. The behaviour when one of the lookups defers can be
2801 controlled by a keyword.
2803 (4) It is now possible to specify the separator character for use when
2804 multiple records are returned.
2806 38. The dnslists ACL condition has been extended: it is now possible to supply
2807 a list of IP addresses and/or domains to be looked up in a particular DNS
2810 39. Added log_selector=+queue_time_overall.
2812 40. When running the queue in the test harness, wait just a tad after forking a
2813 delivery process, to get repeatability of debugging output.
2815 41. Include certificate and key file names in error message when GnuTLS fails
2816 to set them up, because the GnuTLS error message doesn't include the name
2817 of the failing file when there is a problem reading it.
2819 42. Allow both -bf and -bF in the same test run.
2821 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2823 44. The "Exiscan patch" is now merged into the mainline Exim source.
2825 45. Sometimes the final signoff response after QUIT could fail to get
2826 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2827 < 0 before doing a fflush(). This bug looks as though it goes back to the
2828 introduction of TLS in release 3.20, but "sometimes" must have been rare
2829 because the tests only now provoked it.
2831 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2832 (this can affect the format of dates).
2834 47. exim_tidydb, when checking for the continued existence of a message for
2835 which it has found a message-specific retry record, was not finding
2836 messages that were in split spool directories. Consequently, it was
2837 deleting retry records that should have stayed in existence.
2839 48. Steve fixed some bugs in eximstats.
2841 49. The SPA authentication driver was not abandoning authentication and moving
2842 on to the next authenticator when an expansion was forced to fail,
2843 contradicting the general specification for all authenticators. Instead it
2844 was generating a temporary error. It now behaves as specified.
2846 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2847 (the order specifies the preference for clients). The order is now AES256,
2848 AES128, 3DES, ARCFOUR128.
2850 51. Small patch to Sieve code - explicitly set From: when generating an
2853 52. Exim crashed if a remote delivery caused a very long error message to be
2854 recorded - for instance if somebody sent an entire SpamAssassin report back
2855 as a large number of 550 error lines. This bug was coincidentally fixed by
2856 increasing the size of one of Exim's internal buffers (big_buffer) that
2857 happened as part of the Exiscan merge. However, to be on the safe side, I
2858 have made the code more robust (and fixed the comments that describe what
2861 53. Now that there can be additional text after "Completed" in log lines (if
2862 the queue_time_overall log selector is set), a one-byte patch to exigrep
2863 was needed to allow it to recognize "Completed" as not the last thing in
2866 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2867 patch that reportedly fixes this has been added. I am not expert enough to
2868 create a test for it. This is what the patch creator wrote:
2870 "I found a little strange behaviour of ldap code when working with
2871 Windows 2003 AD Domain, where users was placed in more than one
2872 Organization Units. When I tried to give exim partial DN, the exit code
2873 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2874 But simultaneously result of request was absolutely normal ldap result,
2875 so I produce this patch..."
2877 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2878 REFERENCE, so I have modified the code to exclude the patch when that macro
2881 55. Some experimental protocols are using DNS PTR records for new purposes. The
2882 keys for these records are domain names, not reversed IP addresses. The
2883 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2884 leaves it alone. Component reversal etc. now happens only for IP addresses.
2887 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2889 57. Double the size of the debug message buffer (to 2048) so that more of very
2890 long debug lines gets shown.
2892 58. The exicyclog utility now does better if the number of log files to keep
2893 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2895 59. Two changes related to the smtp_active_hostname option:
2897 (1) $smtp_active_hostname is now available as a variable.
2898 (2) The default for smtp_banner uses $smtp_active_hostname instead
2899 of $primary_hostname.
2901 60. The host_aton() function is supposed to be passed a string that is known
2902 to be a valid IP address. However, in the case of IPv6 addresses, it was
2903 not checking this. This is a hostage to fortune. Exim now panics and dies
2904 if the condition is not met. A case was found where this could be provoked
2905 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2906 components; fortuitously, this particular loophole had already been fixed
2907 by change 4.50/55 above.
2909 If there are any other similar loopholes, the new check in host_aton()
2910 itself should stop them being exploited. The report I received stated that
2911 data on the command line could provoke the exploit when Exim was running as
2912 exim, but did not say which command line option was involved. All I could
2913 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2914 running as the user.
2917 61. There was a buffer overflow vulnerability in the SPA authentication code
2918 (which came originally from the Samba project). I have added a test to the
2919 spa_base64_to_bits() function which I hope fixes it.
2922 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2923 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2925 63. The daemon start-up calls getloadavg() while still root for those OS that
2926 need the first call to be done as root, but it missed one case: when
2927 deliver_queue_load_max is set with deliver_drop_privilege. This is
2928 necessary for the benefit of the queue runner, because there is no re-exec
2929 when deliver_drop_privilege is set.
2931 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2932 This has been fixed.
2934 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2935 was in use, was not putting the data itself into the right store pool;
2936 consequently, it could be overwritten for a subsequent message in the same
2937 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2940 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2942 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2943 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2944 the function were treating the return as a boolean value, which happened to
2945 work because 0=false and not-0=true, but is not correct code.
2947 68. The host_aton() function was not handling scoped IPv6 addresses (those
2948 with, for example, "%eth0" on the end) correctly.
2950 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2951 negated items (that is, ~something) in unsigned ints. Some compilers
2952 apparently mutter when there is no cast.
2954 70. If an address verification called from an ACL failed, and did not produce a
2955 user-specific message (i.e. there was only a "system" message), nothing was
2956 put in $acl_verify_message. In this situation, it now puts the system
2959 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2960 session; change 4.31/43 added the unwanted input to the log line - except
2961 that it did not do this in the start of session case. It now does.
2963 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2964 This is wrong and can cause the other end to generate a synchronization
2965 error if it is another Exim or anything else that does the synchronization
2966 check. A QUIT command is no longer sent after a timeout.
2968 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2969 during host lookups.
2971 74. The defer_ok option of callout verification was not working if it was used
2972 when verifying addresses in header lines, that is, for this case:
2974 verify = header_sender/callout=defer_ok
2976 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2977 those file descriptors could be used for SMTP connections. If anything
2978 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2979 could be sent to the SMTP client, causing chaos. The daemon now opens
2980 stdin, stdout, and stderr to /dev/null when it puts itself into the
2983 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2984 log by default. The user can override this with suitable Perl magic.
2986 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2987 the log message when discard triggers, was not working for the DATA ACL or
2988 for the non-SMTP ACL.
2990 78. Error message wording change in sieve.c.
2992 79. If smtp_accept_max_per_host was set, the number of connections could be
2993 restricted to fewer than expected, because the daemon was trying to set up
2994 a new connection before checking whether the processes handling previous
2995 connections had finished. The check for completed processes is now done
2996 earlier. On busy systems, this bug wouldn't be noticed because something
2997 else would have woken the daemon, and it would have reaped the completed
3000 80. If a message was submitted locally by a user whose login name contained one
3001 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3002 It caused a spool format error. I have fixed the spool reading code. A
3003 related problem was that the "from" clause in the Received: line became
3004 illegal because of the space(s). It is now covered by ${quote_local_part.
3006 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3009 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3011 83. Patch from Sieve maintainer to fix unterminated string problem in
3012 "vacation" handling.
3014 84. Some minor changes to the Linux configuration files to help with other
3015 OS variants using glibc.
3017 85. One more patch for Sieve to update vacation handling to latest spec.
3020 ----------------------------------------------------
3021 See the note above about the 4.44 and 4.50 releases.
3022 ----------------------------------------------------
3028 1. Change 4.43/35 introduced a bug that caused file counts to be
3029 incorrectly computed when quota_filecount was set in an appendfile
3032 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3033 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3036 3. Give more explanation in the error message when the command for a transport
3037 filter fails to execute.
3039 4. There are several places where Exim runs a non-Exim command in a
3040 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3041 was being done only for the command run by the queryprogram router. It is
3042 now done for all such subprocesses. The other cases are: ${run, transport
3043 filters, and the commands run by the lmtp and pipe transports.
3045 5. Some older OS have a limit of 256 on the maximum number of file
3046 descriptors. Exim was using setrlimit() to set 1000 as a large value
3047 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3048 systems. I've change it so that if it can't get 1000, it tries for 256.
3050 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3051 was an oversight, and furthermore, ever since the addition of extra
3052 controls (e.g. 4.43/32), the checks on when to allow different forms of
3053 "control" were broken. There should now be diagnostics for all cases when a
3054 control that does not make sense is encountered.
3056 7. $recipients is now available in the predata ACL (oversight).
3058 8. Tidy the search cache before the fork to do a delivery from a message
3059 received from the command line. Otherwise the child will trigger a lookup
3060 failure and thereby defer the delivery if it tries to use (for example) a
3061 cached ldap connection that the parent has called unbind on.
3063 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3064 of $address_data from the recipient verification was clobbered by the
3065 sender verification.
3067 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3068 was its contents. (It was OK if the option was not defined at all.)
3070 11. A "Completed" log line is now written for messages that are removed from
3071 the spool by the -Mrm option.
3073 12. $host_address is now set to the target address during the checking of
3074 ignore_target_hosts.
3076 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3077 being passed; this would have caused $sender_host_name to have been used if
3078 matching the list had actually called for a host name (not very likely,
3079 since this list is usually IP addresses). A host name is now passed as
3082 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3083 code that specifies a non-restarting handler (typically sigaction() in
3084 modern systems) in an attempt to fix a rare and obscure crash bug.
3086 15. Narrowed the window for a race in the daemon that could cause it to ignore
3087 SIGCHLD signals. This is not a major problem, because they are used only to
3088 wake it up if nothing else does.
3090 16. A malformed maildirsize file could cause Exim to calculate negative values
3091 for the mailbox size or file count. Odd effects could occur as a result.
3092 The maildirsize information is now recalculated if the size or filecount
3095 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3096 support for a long time. Removed HAVE_SYS_VFS_H.
3098 18. Updated exipick to current release from John Jetmore.
3100 19. Allow an empty sender to be matched against a lookup in an address list.
3101 Previously the only cases considered were a regular expression, or an
3104 20. Exim went into a mad DNS lookup loop when doing a callout where the
3105 host was specified on the transport, if the DNS lookup yielded more than
3108 21. The RFC2047 encoding function was originally intended for short strings
3109 such as real names; it was not keeping to the 75-character limit for
3110 encoded words that the RFC imposes. It now respects the limit, and
3111 generates multiple encoded words if necessary. To be on the safe side, I
3112 have increased the buffer size for the ${rfc2047: expansion operator from
3115 22. Failure to deliver a bounce message always caused it to be frozen, even if
3116 there was an errors_to setting on the router. The errors_to setting is now
3119 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3120 canonical form (fully expanded) before being placed in
3121 $sender_host_address.
3123 24. Updated eximstats to version 1.33
3125 25. Include certificate and key file names in error message when GnuTLS fails
3126 to set them up, because the GnuTLS error message doesn't include the name
3127 of the failing file when there is a problem reading it.
3129 26. Expand error message when OpenSSL has problems setting up cert/key files.
3132 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3133 (this can affect the format of dates).
3135 28. exim_tidydb, when checking for the continued existence of a message for
3136 which it has found a message-specific retry record, was not finding
3137 messages that were in split spool directories. Consequently, it was
3138 deleting retry records that should have stayed in existence.
3140 29. eximstats updated to version 1.35
3141 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3142 1.35 - bugfix such that pie charts by volume are generated correctly
3144 30. The SPA authentication driver was not abandoning authentication and moving
3145 on to the next authenticator when an expansion was forced to fail,
3146 contradicting the general specification for all authenticators. Instead it
3147 was generating a temporary error. It now behaves as specified.
3149 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3150 (the order specifies the preference for clients). The order is now AES256,
3151 AES128, 3DES, ARCFOUR128.
3153 31. Small patch to Sieve code - explicitly set From: when generating an
3156 32. Exim crashed if a remote delivery caused a very long error message to be
3157 recorded - for instance if somebody sent an entire SpamAssassin report back
3158 as a large number of 550 error lines. This bug was coincidentally fixed by
3159 increasing the size of one of Exim's internal buffers (big_buffer) that
3160 happened as part of the Exiscan merge. However, to be on the safe side, I
3161 have made the code more robust (and fixed the comments that describe what
3164 33. Some experimental protocols are using DNS PTR records for new purposes. The
3165 keys for these records are domain names, not reversed IP addresses. The
3166 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3167 leaves it alone. Component reversal etc. now happens only for IP addresses.
3170 34. The host_aton() function is supposed to be passed a string that is known
3171 to be a valid IP address. However, in the case of IPv6 addresses, it was
3172 not checking this. This is a hostage to fortune. Exim now panics and dies
3173 if the condition is not met. A case was found where this could be provoked
3174 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3175 components; fortuitously, this particular loophole had already been fixed
3176 by change 4.50/55 or 4.44/33 above.
3178 If there are any other similar loopholes, the new check in host_aton()
3179 itself should stop them being exploited. The report I received stated that
3180 data on the command line could provoke the exploit when Exim was running as
3181 exim, but did not say which command line option was involved. All I could
3182 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3183 running as the user.
3186 35. There was a buffer overflow vulnerability in the SPA authentication code
3187 (which came originally from the Samba project). I have added a test to the
3188 spa_base64_to_bits() function which I hope fixes it.
3191 36. The daemon start-up calls getloadavg() while still root for those OS that
3192 need the first call to be done as root, but it missed one case: when
3193 deliver_queue_load_max is set with deliver_drop_privilege. This is
3194 necessary for the benefit of the queue runner, because there is no re-exec
3195 when deliver_drop_privilege is set.
3197 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3198 was in use, was not putting the data itself into the right store pool;
3199 consequently, it could be overwritten for a subsequent message in the same
3200 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3203 38. Sometimes the final signoff response after QUIT could fail to get
3204 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3205 < 0 before doing a fflush(). This bug looks as though it goes back to the
3206 introduction of TLS in release 3.20, but "sometimes" must have been rare
3207 because the tests only now provoked it.
3213 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3214 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3215 Now it returns an integer. A number of calls were still expecting a T/F
3216 return. Fortuitously, in all cases, the tests worked in OK situations,
3217 which is the norm. However, things would have gone wrong on any write
3218 failures on the smtp file descriptor. This function is used when sending
3219 messages over SMTP and also when doing verify callouts.
3221 2. When Exim is called to do synchronous delivery of a locally submitted
3222 message (the -odf or -odi options), it no longer closes stderr before doing
3225 3. Implemented the mua_wrapper option.
3227 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3229 5. Implemented the functions header_remove(), header_testname(),
3230 header_add_at_position(), and receive_remove_recipient(), and exported them
3233 6. If an ACL "warn" statement specified the addition of headers, Exim already
3234 inserted X-ACL-Warn: at the start if there was no header name. However, it
3235 was not making this test for the second and subsequent header lines if
3236 there were newlines in the string. This meant that an invalid header could
3237 be inserted if Exim was badly configured.
3239 7. Allow an ACL "warn" statement to add header lines at the start or after all
3240 the Received: headers, as well as at the end.
3242 8. Added the rcpt_4xx retry error code.
3244 9. Added postmaster_mailfrom=xxx to callout verification option.
3246 10. Added mailfrom=xxxx to the callout verification option, for verify=
3249 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3250 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3252 12. Inserted some casts to stop certain compilers warning when using pointer
3253 differences as field lengths or precisions in printf-type calls (mostly
3254 affecting debugging statements).
3256 13. Added optional readline() support for -be (dynamically loaded).
3258 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3259 same clock tick as a message's arrival, so that its received time was the
3260 same as the "first fail" time on the retry record, and that message
3261 remained on the queue past the ultimate address timeout, every queue runner
3262 would try a delivery (because it was past the ultimate address timeout) but
3263 after another failure, the ultimate address timeout, which should have then
3264 bounced the address, did not kick in. This was a "< instead of <=" error;
3265 in most cases the first failure would have been in the next clock tick
3266 after the received time, and all would be well.
3268 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3269 being recognized when the domain list was tested by the match_domain
3270 condition in an expansion string.
3272 16. Added the ${str2b64: operator.
3274 17. Exim was always calling setrlimit() to set a large limit for the number of
3275 processes, without checking whether the existing limit was already
3276 adequate. (It did check for the limit on file descriptors.) Furthermore,
3277 errors from getrlimit() and setrlimit() were being ignored. Now they are
3278 logged to the main and panic logs, but Exim does carry on, to try to do its
3279 job under whatever limits there are.
3281 18. Imported PCRE 5.0.
3283 19. Trivial typo in log message " temporarily refused connection" (the leading
3286 20. If the log selector return_path_on_delivery was set and an address was
3287 redirected to /dev/null, the delivery process crashed because it assumed
3288 that a return path would always be set for a "successful" delivery. In this
3289 case, the whole delivery is bypassed as an optimization, and therefore no
3292 21. Internal re-arrangement: the function for sending a challenge and reading
3293 a response while authentication was assuming a zero-terminated challenge
3294 string. It's now changed to take a pointer and a length, to allow for
3295 binary data in such strings.
3297 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3299 23. Exim was not respecting finduser_retries when seeking the login of the
3300 uid under which it was called; it was always trying 10 times. (The default
3301 setting of finduser_retries is zero.) Also, it was sleeping after the final
3302 failure, which is pointless.
3304 24. Implemented tls_on_connect_ports.
3306 25. Implemented acl_smtp_predata.
3308 26. If the domain in control=submission is set empty, Exim assumes that the
3309 authenticated id is a complete email address when it generates From: or
3310 Sender: header lines.
3312 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3313 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3314 chown and chgrp in /bin and hostname in /usr/bin.
3316 28. Exim was keeping the "process log" file open after each use, just as it
3317 does for the main log. This opens the possibility of it remaining open for
3318 long periods when the USR1 signal hits a daemon. Occasional processlog
3319 errors were reported, that could have been caused by this. Anyway, it seems
3320 much more sensible not to leave this file open at all, so that is what now
3323 29. The long-running daemon process does not normally write to the log once it
3324 has entered its main loop, and it closes the log before doing so. This is
3325 so that log files can straightforwardly be renamed and moved. However,
3326 there are a couple of unusual error situations where the daemon does write
3327 log entries, and I had neglected to close the log afterwards.
3329 30. The text of an SMTP error response that was received during a remote
3330 delivery was being truncated at 512 bytes. This is too short for some of
3331 the long messages that one sometimes sees. I've increased the limit to
3334 31. It is now possible to make retry rules that apply only when a message has a
3335 specific sender, in particular, an empty sender.
3337 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3338 it possible to be selective about when SMTP synchronization is enforced.
3340 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3342 32. Implemented hosts_connection_nolog.
3344 33. Added an ACL for QUIT.
3346 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3349 35. Added mailbox_size and mailbox_filecount to appendfile.
3351 36. Added control = no_multiline_responses to ACLs.
3353 37. There was a bug in the logic of the code that waits for the clock to tick
3354 in the case where the clock went backwards by a substantial amount such
3355 that the microsecond fraction of "now" was more than the microsecond
3356 fraction of "then" (but the whole seconds number was less).
3358 38. Added support for the libradius Radius client library this is found on
3359 FreeBSD (previously only the radiusclient library was supported).
3365 1. When certain lookups returned multiple values in the form name=value, the
3366 quoting of the values was not always being done properly. Specifically:
3367 (a) If the value started with a double quote, but contained no whitespace,
3369 (b) If the value contained whitespace other than a space character (i.e.
3370 tabs or newlines or carriage returns) it was not quoted.
3371 This fix has been applied to the mysql and pgsql lookups by writing a
3372 separate quoting function and calling it from the lookup code. The fix
3373 should probably also be applied to nisplus, ibase and oracle lookups, but
3374 since I cannot test any of those, I have not disturbed their existing code.
3376 2. A hit in the callout cache for a specific address caused a log line with no
3377 reason for rejecting RCPT. Now it says "Previous (cached) callout
3378 verification failure".
3380 3. There was an off-by-one bug in the queryprogram router. An over-long
3381 return line was truncated at 256 instead of 255 characters, thereby
3382 overflowing its buffer with the terminating zero. As well as fixing this, I
3383 have increased the buffer size to 1024 (and made a note to document this).
3385 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3386 when Exim is waiting for an SMTP response from a remote server, Exim
3387 restarts its select() call on the socket, thereby resetting its timeout.
3388 This is not a problem when such interrupts are rare. Somebody set up a cron
3389 job to run exiwhat every 2 minutes, which is less than the normal select()
3390 timeout (5 or 10 minutes). This meant that the select() timeout never
3391 kicked in because it was always reset. I have fixed this by comparing the
3392 time when an interrupt arrives with the time at the start of the first call
3393 to select(). If more time than the timeout has elapsed, the interrupt is
3394 treated as a timeout.
3396 5. Some internal re-factoring in preparation for the addition of Sieve
3397 extensions (by MH). In particular, the "personal" test is moved to a
3398 separate function, and given an option for scanning Cc: and Bcc: (which is
3399 not set for Exim filters).
3401 6. When Exim created an email address using the login of the caller as the
3402 local part (e.g. when creating a From: or Sender: header line), it was not
3403 quoting the local part when it contained special characters such as @.
3405 7. Installed new OpenBSD configuration files.
3407 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3408 try to make them clearer.
3410 9. Callout options, other than the timeout value, were being ignored when
3411 verifying sender addresses in header lines. For example, when using
3413 verify = header_sender/callout=no_cache
3415 the cache was (incorrectly) being used.
3417 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3418 only the Cygwin environment.
3420 11. When return_path_on_delivery was set as a log selector, if different remote
3421 addresses in the same message used different return paths and parallel
3422 remote delivery occurred, the wrong values would sometimes be logged.
3423 (Whenever a remote delivery process finished, the return path value from
3424 the most recently started remote delivery process was logged.)
3426 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3427 lines when AUTH and/or TLS are in use. This is the "received protocol"
3428 field. Exim used to use "asmtp" for authenticated SMTP, without any
3429 indication (in the protocol name) for TLS use. Now it follows the RFC and
3430 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3431 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3432 names appear in log lines as well as in Received: header lines.
3434 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3435 extensions, and comparison tests, and to fix some bugs.
3437 14. Changes to the "personal" filter test:
3439 (1) The test was buggy in that it was just doing the equivalent of
3440 "contains" tests on header lines. For example, if a user's address was
3441 anne@some.where, the "personal" test would incorrectly be true for
3443 To: susanne@some.where
3445 This test is now done by extracting each address from the header in turn,
3446 and checking the entire address. Other tests that are part of "personal"
3447 are now done using regular expressions (for example, to check local parts
3448 of addresses in From: header lines).
3450 (2) The list of non-personal local parts in From: addresses has been
3451 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3452 taken from the Sieve specification recommendations.
3454 (3) If the message contains any header line starting with "List-" it is
3455 treated as non-personal.
3457 (4) The test for "circular" in the Subject: header line has been removed
3458 because it now seems ill-conceived.
3460 15. Minor typos in src/EDITME comments corrected.
3462 16. Installed latest exipick from John Jetmore.
3464 17. If headers_add on a router specified a text string that was too long for
3465 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3466 of string_sprintf() is now avoided.
3468 18. $message_body_size was not set (it was always zero) when running the DATA
3469 ACL and the local_scan() function.
3471 19. For the "mail" command in an Exim filter, no default was being set for
3472 the once_repeat time, causing a random time value to be used if "once" was
3473 specified. (If the value happened to be <= 0, no repeat happened.) The
3474 default is now 0s, meaning "never repeat". The "vacation" command was OK
3475 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3476 (I found it when inspecting the code).
3478 20. There is now an overall timeout for performing a callout verification. It
3479 defaults to 4 times the callout timeout, which applies to individual SMTP
3480 commands during the callout. The overall timeout applies when there is more
3481 than one host that can be tried. The timeout is checked before trying the
3482 next host. This prevents very long delays if there are a large number of
3483 hosts and all are timing out (e.g. when the network connections are timing
3484 out). The value of the overall timeout can be changed by specifying an
3485 additional sub-option for "callout", called "maxwait". For example:
3487 verify = sender/callout=5s,maxwait=20s
3489 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3490 to the end before writing, but this should make it even safer).
3492 22. Exim was forgetting that it had advertised PIPELINING for the second and
3493 subsequent messages on an SMTP connection. It was also not resetting its
3494 memory on STARTTLS and an internal HELO.
3496 23. When Exim logs an SMTP synchronization error within a session, it now
3497 records whether PIPELINING has been advertised or not.
3499 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3500 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3501 rather than long int.
3503 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3505 26. Added the never_mail option to autoreply.
3511 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3512 crash if the getsockname() call failed; this can happen if a connection is
3513 closed very soon after it is established. The problem was simply in the
3514 order in which certain operations were done, causing Exim to try to write
3515 to the SMTP stream before it had set up the file descriptor. The bug has
3516 been fixed by making things happen in the correct order.
3522 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3523 before the connection was closed, thus losing the rejection response.
3525 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3526 some early Solaris releases, but causes trouble in current releases where
3527 socklen_t is defined.
3529 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3532 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3535 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3536 that releasing the top of it at the end releases what was used for sub-
3537 expansions (unless the block got too big). However, discard this block if
3538 the first thing is a variable or header, so that we can use its block when
3539 it is dynamic (useful for very large $message_headers, for example).
3541 6. Lookups now cache *every* query, not just the most recent. A new, separate
3542 store pool is used for this. It can be recovered when all lookup caches are
3543 flushed. Lookups now release memory at the end of their result strings.
3544 This has involved some general refactoring of the lookup sources.
3546 7. Some code has been added to the store_xxx() functions to reduce the amount
3547 of flapping under certain conditions.
3549 8. log_incoming_interface used to affect only the <= reception log lines. Now
3550 it causes the local interface and port to be added to several more SMTP log
3551 lines, for example "SMTP connection from", and rejection lines.
3553 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3555 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3557 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3558 could be overwritten at the end of the current message (or the start of a
3559 new message if it was set in a HELO ACL). The value is now preserved for
3560 the duration of the SMTP connection.
3562 12. If a transport had a headers_rewrite setting, and a matching header line
3563 contained an unqualified address, that address was qualified, even if it
3564 did not match any rewriting rules. The underlying bug was that the values
3565 of the flags that permit the existence of unqualified sender and recipient
3566 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3567 non-local messages, and by -bnq for local messages) were not being
3568 preserved with the message after it was received.
3570 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3571 "next input=" as part of the text comprising the host identity instead of
3572 the correct text. The code was using the same buffer for two different
3573 strings. However, depending on which order the printing function evaluated
3574 its arguments, the bug did not always show up. Under Linux, for example, my
3575 test suite worked just fine.
3577 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3578 doesn't work with some older versions of Perl. It has been changed to "my",
3579 which in any case is probably the better facility to use.
3581 15. A really picky compiler found some instances of statements for creating
3582 error messages that either had too many or two few arguments for the format
3585 16. The size of the buffer for calls to the DNS resolver has been increased
3586 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3587 for addresses that have a lot of PTR records. This alleviates a problem; it
3588 does not fully solve it.
3590 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3591 buffer now truncates the list and logs the incident, which is the same
3592 action as happens when Exim is looking up a host name and its aliases.
3593 Previously in this situation something unpredictable would happen;
3594 sometimes it was "internal error: store_reset failed".
3596 18. If a server dropped the connection unexpectedly when an Exim client was
3597 using GnuTLS and trying to read a response, the client delivery process
3598 crashed while trying to generate an error log message.
3600 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3601 string, for example:
3603 warn message = H1: something\nH2: something
3605 the text was added as a single header line from Exim's point of view
3606 though it ended up OK in the delivered message. However, searching for the
3607 second and subsequent header lines using $h_h2: did not work. This has been
3608 fixed. Similarly, if a system filter added multiple headers in this way,
3609 the routers could not see them.
3611 20. Expanded the error message when iplsearch is called with an invalid key to
3612 suggest using net-iplsearch in a host list.
3614 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3615 ACLs are no longer actually imposed (and a message to that effect is
3618 22. If a "gecos" field in a passwd entry contained escaped characters, in
3619 particular, if it contained a \" sequence, Exim got it wrong when building
3620 a From: or a Sender: header from that name. A second bug also caused
3621 incorrect handling when an unquoted " was present following a character
3622 that needed quoting.
3624 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3625 was not being matched caselessly.
3627 24. Arranged for all hyphens in the exim.8 source to be escaped with
3630 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3631 properly. Recipient callout cache records were still being keyed to include
3632 the sender, even when use_sender was set false. This led to far more
3633 callouts that were necessary. The sender is no longer included in the key
3634 when use_sender is false.
3636 26. Added "control = submission" modifier to ACLs.
3638 27. Added the ${base62d: operator to decode base 62 numbers.
3640 28. dnsdb lookups can now access SRV records.
3642 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3643 the configuration file.
3645 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3646 (-v) mode. This makes the output for a verbose queue run more intelligible.
3648 31. Added a use_postmaster feature to recipient callouts.
3650 32. Added the $body_zerocount variable, containing the number of binary zero
3651 bytes in the message body.
3653 33. The time of last modification of the "new" subdirectory is now used as the
3654 "mailbox time last read" when there is a quota error for a maildir
3657 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3659 35. Added +ignore_unknown as a special item in host lists.
3661 36. Code for decoding IPv6 addresses in host lists is now included, even if
3662 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3663 address was recognized as an IP address, but was then not correctly decoded
3664 into binary, causing unexpected and incorrect effects when compared with
3671 1. Very minor rewording of debugging text in manualroute to say "list of
3672 hosts" instead of "hostlist".
3674 2. If verify=header_syntax was set, and a header line with an unqualified
3675 address (no domain) and a large number of spaces between the end of the
3676 name and the colon was received, the reception process suffered a buffer
3677 overflow, and (when I tested it) crashed. This was caused by some obsolete
3678 code that should have been removed. The fix is to remove it!
3680 3. When running in the test harness, delay a bit after writing a bounce
3681 message to get a bit more predictability in the log output.
3683 4. Added a call to search_tidyup() just before forking a reception process. In
3684 theory, someone could use a lookup in the expansion of smtp_accept_max_
3685 per_host which, without the tidyup, could leave open a database connection.
3687 5. Added the variables $recipient_data and $sender_data which get set from a
3688 lookup success in an ACL "recipients" or "senders" condition, or a router
3689 "senders" option, similar to $domain_data and $local_part_data.
3691 6. Moved the writing of debug_print from before to after the "senders" test
3694 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3695 problems for message scanning, either using a data ACL, or using
3696 local_scan() because the Received: header was not generated till after they
3697 were called (in order to set the time as the time of reception completion).
3698 I have revised the way this works. The header is now generated after the
3699 body is received, but before the ACL or local_scan() are called. After they
3700 are run, the timestamp in the header is updated.
3706 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3707 before starting a queue runner without re-exec. This happened only when
3708 deliver_drop_privilege was set or when the Exim user was set to root. The
3709 effect of the bug was that timeouts during subsequent deliveries caused
3710 crashes instead of being properly handled. The handler is now left at its
3711 default (and expected) setting.
3713 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3714 message, again when deliver_drop_privilege is set or Exim is run as root.
3715 The bug described in (1) was not present in this case, but the tidying up
3716 of the other signals was missing. I have made the two cases consistent.
3718 3. The ignore_target_hosts setting on a manualroute router was being ignored
3719 for hosts that were looked up using the /MX notation.
3721 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3724 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3725 operated on the sender address. After changing the $sender_address to <>
3726 for the sender address verify, Exim was re-instated it as the original
3727 (before rewriting) address, but remembering that it had rewritten it, so it
3728 wasn't rewriting it again. This bug also had the effect of breaking the
3729 sender address verification caching when the sender address was rewritten.
3731 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3732 This has been changed so that if the ip literal address matches
3733 ignore_target_hosts, the router declines.
3735 7. Added expansion conditions match_domain, match_address, and match_local_
3736 part (NOT match_host).
3738 8. The placeholder for the Received: header didn't have a length field set.
3740 9. Added code to Exim itself and to exim_lock to test for a specific race
3741 condition that could lead to file corruption when using MBX delivery. The
3742 issue is with the lockfile that is created in /tmp. If this file is removed
3743 after a process has opened it but before that process has acquired a lock,
3744 there is the potential for a second process to recreate the file and also
3745 acquire a lock. This could lead to two Exim processes writing to the file
3746 at the same time. The added code performs the same test as UW imapd; it
3747 checks after acquiring the lock that its file descriptor still refers to
3748 the same named file.
3750 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3751 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3754 11. Added the smtp_active_hostname option. If used, this will typically be made
3755 to depend on the incoming interface address. Because $interface_address is
3756 not set up until the daemon has forked a reception process, error responses
3757 that can happen earlier (such as "too many connections") no longer contain
3760 12. If an expansion in a condition on a "warn" statement fails because a lookup
3761 defers, the "warn" statement is abandoned, and the next ACL statement is
3762 processed. Previously this caused the whole ACL to be aborted.
3764 13. Added the iplsearch lookup type.
3766 14. Added ident_timeout as a log selector.
3768 15. Added tls_certificate_verified as a log selector.
3770 16. Added a global option tls_require_ciphers (compare the smtp transport
3771 option of the same name). This controls incoming TLS connections.
3773 17. I finally figured out how to make tls_require_ciphers do a similar thing
3774 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3775 before starting the TLS session.
3777 18. Tabs are now shown as \t in -bP output.
3779 19. If the log selector return_path_on_delivery was set, Exim crashed when
3780 bouncing a message because it had too many Received: header lines.
3782 20. If two routers both had headers_remove settings, and the first one included
3783 a superfluous trailing colon, the final name in the first list and the
3784 first name in the second list were incorrectly joined into one item (with a
3785 colon in the middle).
3791 1. Added -C and -D options to the exinext utility, mainly to make it easier
3792 to include in the automated testing, but these could be helpful when
3793 multiple configurations are in use.
3795 2. The exinext utility was not formatting the output nicely when there was
3796 an alternate port involved in the retry record key, nor when there was a
3797 message id as well (for retries that were specific to a specific message
3798 and a specific host). It was also confused by IPv6 addresses, because of
3799 the additional colons they contain. I have fixed the IPv4 problem, and
3800 patched it up to do a reasonable job for IPv6.
3802 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3803 delivery, the log line now contains "pipelined" if PIPELINING was used.
3805 4. An SMTP transport process used to panic and die if the bind() call to set
3806 an explicit outgoing interface failed. This has been changed; it is now
3807 treated in the same way as a connect() failure.
3809 5. A reference to $sender_host_name in the part of a conditional expansion
3810 that was being skipped was still causing a DNS lookup. This no longer
3813 6. The def: expansion condition was not recognizing references to header lines
3814 that used bh_ and bheader_.
3816 7. Added the _cache feature to named lists.
3818 8. The code for checking quota_filecount in the appendfile transport was
3819 allowing one more file than it should have been.
3821 9. For compatibility with Sendmail, the command line option
3829 and sets the incoming protocol and host name (for trusted callers). The
3830 host name and its colon can be omitted when only the protocol is to be set.
3831 Note the Exim already has two private options, -pd and -ps, that refer to
3832 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3833 "s", but I don't think that's a major issue.
3835 10. A number of refactoring changes to the code, none of which should affect
3838 (a) The number of logging options was getting close to filling up the
3839 32-bit word that was used as a bit map. I have split them into two classes:
3840 those that are passed in the argument to log_write(), and those that are
3841 only ever tested independently outside of that function. These are now in
3842 separate 32-bit words, so there is plenty of room for expansion again.
3843 There is no change in the user interface or the logging behaviour.
3845 (b) When building, for example, log lines, the code previously used a
3846 macro that called string_cat() twice, in order to add two strings. This is
3847 not really sufficiently general. Furthermore, there was one instance where
3848 it was actually wrong because one of the argument was used twice, and in
3849 one call a function was used. (As it happened, calling the function twice
3850 did not affect the overall behaviour.) The macro has been replaced by a
3851 function that can join an arbitrary number of extra strings onto a growing
3854 (c) The code for expansion conditions now uses a table and a binary chop
3855 instead of a serial search (which was left over from when there were very
3856 few conditions). Also, it now recognizes conditions like "pam" even when
3857 the relevant support is not compiled in: a suitably worded error message is
3858 given if an attempt is made to use such a condition.
3860 11. Added ${time_interval:xxxxx}.
3862 12. A bug was causing one of the ddress fields not to be passed back correctly
3863 from remote delivery subprocesses. The field in question was not being
3864 subsequently used, so this caused to problems in practice.
3866 13. Added new log selectors queue_time and deliver_time.
3868 14. Might have fixed a bug in maildirsizefile handling that threw up
3869 "unexpected character" debug warnings, and recalculated the data
3870 unnecessarily. In any case, I expanded the warning message to give more
3873 15. Added the message "Restricted characters in address" to the statements in
3874 the default ACL that block characters like @ and % in local parts.
3876 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3877 Three changes have been made:
3879 (a) There was a serious bug; a negative response to MAIL caused the whole
3880 recipient domain to be cached as invalid, thereby blocking all messages
3881 to all local parts at the same domain, from all senders. This bug has
3882 been fixed. The domain is no longer cached after a negative response to
3883 MAIL if the sender used is not empty.
3885 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3888 (c) A new callout option, "use_sender" has been added for people who want
3889 the modified behaviour.
3895 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3898 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3899 indeed breaks things for older releases.
3901 3. Added additional logging to the case where there is a problem reading data
3902 from a filter that is running in a subprocess using a pipe, in order to
3903 try to track down a specific problem.
3905 4. Testing facility fudge: when running in the test harness and attempting
3906 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3907 getting "No route to host". Convert this to a timeout.
3909 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3912 6. Some OS don't have socklen_t but use size_t instead. This affects the
3913 fifth argument of getsockopt() amongst other things. This is now
3914 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3915 can be set for individual OS. I have set it for SunOS5, OSF1, and
3916 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3917 some earlier ones do not.
3919 7. Change 4.30/15 was not doing the test caselessly.
3921 8. The standard form for an IPv6 address literal was being rejected by address
3922 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3923 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3924 this, as well as the form without the "IPv6" on the front (but only when
3925 address literals are enabled, of course).
3927 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3929 10. Exim crashed if a message with an empty sender address specified by -f
3930 encountered a router with an errors_to setting. This could be provoked only
3931 by a command such as
3935 where an empty string was supplied; "<>" did not hit this bug.
3937 11. Installed PCRE release 4.5.
3939 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3940 remained set. It is now erased.
3942 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3943 times from message ids (which are base 36 rather than the normal 62).
3945 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3946 were being counted as actual protocol errors, and logged if the log
3947 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3948 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3949 and DATA following a set of rejected RCPTs do not count as protocol errors.
3950 In other words, Exim assumes they were pipelined, though this may not
3951 actually be the case. Of course, in all cases the client gets an
3952 appropriate error code.
3954 15. If a lookup fails in an ACL condition, a message about the failure may
3955 be available; it is used if testing the ACL cannot continue, because most
3956 such messages specify what the cause of the deferral is. However, some
3957 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3958 that caused an old message to be retained and used if a later statement
3959 caused a defer, replacing the real cause of the deferral.
3961 16. If an IP address had so many PTR records that the DNS lookup buffer
3962 was not large enough to hold them, Exim could crash while trying to process
3963 the truncated data. It now detects and logs this case.
3965 17. Further to 4.21/58, another change has been made: if (and only if) the
3966 first line of a message (the first header line) ends with CRLF, a bare LF
3967 in a subsequent header line has a space inserted after it, so as not to
3968 terminate the header.
3970 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3971 unnecessarily, used atoi() instead of strtol(), and didn't check the
3972 termination when getting file sizes from file names by regex.
3974 19. Completely re-implemented the support for maildirsize files, in the light
3975 of a number of problems with the previous contributed implementation
3976 (4.30/29). In particular:
3978 . If the quota is zero, the maildirsize file is maintained, but no quota is
3981 . If the maildir directory does not exist, it is created before any attempt
3982 to write a maildirsize file.
3984 . The quota value in the file is just a cache; if the quota is changed in
3985 the transport, the new value overrides.
3987 . A regular expression is available for excluding directories from the
3990 20. The autoreply transport checks the characters in options that define the
3991 message's headers; it allows continued headers, but it was checking with
3992 isspace() after an embedded newline instead of explicitly looking for a
3995 21. If all the "regular" hosts to which an address was routed had passed their
3996 expiry times, and had not reached their retry times, the address was
3997 bounced, even if fallback hosts were defined. Now Exim should go on to try
4000 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4001 equivalent code in the SMTP transport. Some hosts send humungous responses
4002 to HELO/EHLO, more than 1024 it seems.
4004 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4005 gives compiler warnings in some environments. I've now done it "properly",
4008 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4009 (because of problems with the built-in one) was declared to return uschar *
4010 instead of char *, causing compiler failure.
4012 25. Fixed a file descriptor leak when processing alias/forward files.
4014 26. Fixed a minor format string issue in dbfn.c.
4016 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4018 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4019 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4022 29. When a sender address is verified, it is cached, to save repeating the test
4023 when there is more than one recipient in a message. However, when the
4024 verification involves a callout, it is possible for different callout
4025 options to be set for different recipients. It is too complicated to keep
4026 track of this in the cache, so now Exim always runs a verification when a
4027 callout is required, relying on the callout cache for the optimization.
4028 The overhead is duplication of the address routing, but this should not be
4031 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4032 to be verified with callout=postmaster, and the main callout worked but the
4033 postmaster check failed, the verification correctly failed. However, if a
4034 subsequent RCPT command asked for sender verification *without* the
4035 postmaster check, incorrect caching caused this verification also to fail,
4038 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4039 it was not caching the DNS options (qualify_single, search_parents) that
4040 were used when the lookup failed. A subsequent lookup with different
4041 options therefore always gave the same answer, though there were cases
4042 where it should not have. (Example: a "domains = !$mx_any" option on a
4043 dnslookup router: the "domains" option is always processed without any
4044 widening, but the router might have qualify_single set.) Now Exim uses the
4045 cached value only when the same options are set.
4047 32. Added John Jetmore's "exipick" utility to the distribution.
4049 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4050 than a timeout (e.g. a certificate is required, and is not provided), an
4051 Exim server now closes the connection immediately. Previously it waited for
4052 the client to close - but if the client is SSL, it seems that they each
4053 wait for each other, leading to a delay before one of them times out.
4055 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4056 maintained 0.8.x compatibility because I don't think many are using it, and
4057 it is clearly obsolete.
4059 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4062 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4063 client certificate was expired. A simple patch fixes this, though I don't
4064 understand the full logic of why the verify callback is called multiple
4067 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4068 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4069 which causes problems with some clients (such as the Certicom SSL Plus
4070 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4071 disables the coutermeasure allowing Eudora to connect."
4073 38. Exim was not checking that a write() to a log file succeeded. This could
4074 lead to Bad Things if a log got too big, in particular if it hit a file
4075 size limit. Exim now panics and dies if it cannot write to a log file, just
4076 as it does if it cannot open a log file.
4078 39. Modified OS/Makefile-Linux so that it now contains
4080 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4082 The two -D definitions ensure that Exim is compiled with large file
4083 support, which makes it possible to handle log files that are bigger than
4086 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4087 instance) a domain was checked against a named list that involved a lookup,
4088 causing $domain_data to be set, then another domain was checked against the
4089 same list, then the first domain was re-checked, the value of $domain_data
4090 after the final check could be wrong. In particular, if the second check
4091 failed, it could be set empty. This bug probably also applied to
4094 41. The strip_trailing_dot option was not being applied to the address given
4095 with the -f command-line option.
4097 42. The code for reading a message's header from the spool was incrementing
4098 $received_count, but never initializing it. This meant that the value was
4099 incorrect (doubled) while delivering a message in the same process in which
4100 it was received. In the most common configuration of Exim, this never
4101 happens - a fresh exec is done - but it can happen when
4102 deliver_drop_privilege is set.
4104 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4105 it now includes up to 150 characters of the unexpected data in the log
4108 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4109 and building data strings. The size of both of these buffers was 10 000
4110 bytes - far larger than anybody would *ever* want, thought I. Needless to
4111 say, somebody hit the limit. I have increased the maximum line length to
4112 20 000 and the maximum data length of concatenated lines to 100 000. I have
4113 also fixed two bugs, because there was no checking on these buffers. Tsk,
4114 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4115 buffer is too small.
4117 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4118 lsearch lookups. Now it does.
4120 46. When parsing a route_list item in a manualroute router, a fixed-length
4121 buffer was used for the list of hosts. I made this 1024 bytes long,
4122 thinking that nobody would ever have a list of hosts that long. Wrong.
4123 Somebody had a whole pile of complicated expansion conditions, and the
4124 string was silently truncated, leading to an expansion error. It turns out
4125 that it is easier to change to an unlimited length (owing to other changes
4126 that have happened since this code was originally written) than to build
4127 structure for giving a limitation error. The length of the item that
4128 expands into the list of hosts is now unlimited.
4130 47. The lsearch lookup could not handle data where the length of text line was
4131 more than 4095 characters. Such lines were truncated, leading to shortened
4132 data being returned. It should now handle lines of any length.
4134 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4135 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4138 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4139 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4140 no longer happens in comment lines. A list of replacements is now placed
4141 at the head of all of the source files, except those whose only change is
4142 to replace PERL_COMMAND in the very first #! line.
4144 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4145 messages on the queue, with a bottom-up merge sort, using code contributed
4146 by Michael Haardt. This should make operations like -bp somewhat faster on
4147 large queues. It won't affect queue runners, except when queue_run_in_order
4150 51. Installed eximstats 1.31 in the distribution.
4152 52. Added support for SRV lookups to the dnslookup router.
4154 53. If an ACL referred to $message_body or $message_body_end, the value was not
4155 reset for any messages that followed in the same SMTP session.
4157 54. The store-handling optimization for building very long strings was not
4158 differentiating between the different store pools. I don't think this
4159 actually made any difference in practice, but I've tidied it.
4161 55. While running the routers to verify a sender address, $sender_address
4162 was still set to the sender address. This is wrong, because when routing to
4163 send a bounce to the sender, it would be empty. Therefore, I have changed
4164 it so that, while verifying a sender address, $sender_address is set to <>.
4165 (There is no change to what happens when verifying a recipient address.)
4167 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4168 target A or AAAA records (if not already returned) without resetting the
4169 qualify_single or search_parents options of the DNS resolver. These are
4170 inappropriate in this case because the targets of MX and SRV records must
4171 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4172 target that, when qualified, matched something in the local domain. These
4173 two options are now turned off when doing these lookups.
4175 57. It seems that at least some releases of Reiserfs (which does not have the
4176 concept of a fixed number of inodes) returns zero and not -1 for the
4177 number of available inodes. This interacted badly with check_spool_inodes,
4178 which assumed that -1 was the "no such thing" setting. What I have done is
4179 to check that the total number of inodes is greater than zero before doing
4180 the test of how many are available.
4182 58. When a "warn" ACL statement has a log_message modifier, the message is
4183 remembered, and not repeated. This is to avoid a lot of repetition when a
4184 message has many recipients that cause the same warning to be written.
4185 Howewer, Exim was preserving the list of already written lines for an
4186 entire SMTP session, which doesn't seem right. The memory is now reset if a
4187 new message is started.
4189 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4190 debugging output unless log_rewrite was also set.
4192 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4193 of (int)(handle) when we know that handle contains (void *)(-1).
4195 61. The Exim daemon panic-logs an error return when it closes the incoming
4196 connection. However "connection reset by peer" seems to be common, and
4197 isn't really an error worthy of noting specially, so that particular error
4200 62. When Exim is trying to find all the local interfaces, it used to panic and
4201 die if the ioctl to get the interface flags failed. However, it seems that
4202 on at least one OS (Solaris 9) it is possible to have an interface that is
4203 included in the list of interfaces, but for which you get a failure error
4204 for this call. This happens when the interface is not "plumbed" into a
4205 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4206 failure of the "get flags" call assumes that the interface is down.
4208 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4209 makes life easier for people who are doing arithmetic on fields extracted
4210 from dates, where you often get leading zeros that should not be
4211 interpreted as octal.
4213 64. Added qualify_domain to the redirect router, to override the global
4216 65. If a pathologically long header line contained very many addresses (the
4217 report of this problem mentioned 10 000) and each of them was rewritten,
4218 Exim could use up a very large amount of memory. (It kept on making new
4219 copies of the header line as it rewrote, and never released the old ones.)
4220 At the expense of a bit more processing, the header rewriting function has
4221 been changed so that it no longer eats memory in this way.
4223 66. The generation of the Received: header has been moved from the time that a
4224 message starts to be received, to the time that it finishes. The timestamp
4225 in the Received: header should now be very close to that of the <= log
4226 line. There are two side-effects of this change:
4228 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4229 logged header lines no longer include the local Received: line, because
4230 it has not yet been created. The same applies to a copy of the message
4231 that is returned to a non-SMTP sender when a message is rejected.
4233 (b) When a filter file is tested using -bf, no additional Received: header
4234 is added to the test message. After some thought, I decided that this
4237 This change does not affect the value of $received_for. It is still set
4238 after address rewriting, but before local_scan() is called.
4240 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4242 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4243 gave an unhelpful panic error message, and a defer error. I have managed to
4244 change this behaviour so that it now rejects any supplied certificate,
4245 which seems right, as the list of acceptable certificates is empty.
4247 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4248 gave an unhelpful defer error. I have not managed to make this reject any
4249 supplied certificates, but the error message it gives is "no certificate
4250 supplied", which is not helpful.
4252 70. exigrep's output now also includes lines that are not associated with any
4253 message, but which match the given pattern. Implemented by a patch from
4254 Martin Sluka, which also tidied up the Perl a bit.
4256 71. Recipient callout verification, like sender verification, was using <> in
4257 the MAIL FROM command. This isn't really the right thing, since the actual
4258 sender may affect whether the remote host accepts the recipient or not. I
4259 have changed it to use the actual sender in the callout; this means that
4260 the cache record is now keyed on a recipient/sender pair, not just the
4261 recipient address. There doesn't seem to be a real danger of callout loops,
4262 since a callout by the remote host to check the sender would use <>.
4263 [SEE ABOVE: changed after hitting problems.]
4265 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4266 temporary errors. However, in the case of such a code being given after
4267 the end of a data transmission (i.e. after ".") Exim was failing to write
4268 a retry record for the message. (Yes, there was some broken host that was
4269 actually sending 8xx at this point.)
4271 73. An unknown lookup type in a host list could cause Exim to panic-die when
4272 the list was checked. (An example that provoked this was putting <; in the
4273 middle of a list instead of at the start.) If this happened during a DATA
4274 ACL check, a -D file could be left lying around. This kind of configuration
4275 error no longer causes Exim to die; instead it causes a defer errror. The
4276 incident is still logged to the main and panic logs.
4278 74. Buglet left over from Exim 3 conversion. The message "too many messages
4279 in one connection" was written to the rejectlog but not the mainlog, except
4280 when address rewriting (yes!) was being logged.
4282 75. Added write_rejectlog option.
4284 76. When a system filter was run not as root (that is, when system_filter_user
4285 was set), the values of the $n variables were not being returned to the
4286 main process; thus, they were not subsequently available in the $sn
4289 77. Added +return_path_on_delivery log selector.
4291 78. A connection timeout was being treated differently from recipients deferred
4292 when testing hosts_max_try with a message that was older than the host's
4293 retry timeout. (The host should not be counted, thus allowing all hosts to
4294 be tried at least once before bouncing.) This may have been the cause of an
4295 occasionally reported bug whereby a message would remain on the queue
4296 longer than the retry timeout, but would be bounced if a delivery was
4297 forced. I say "may" because I never totally pinned down the problem;
4298 setting up timeout/retry tests is difficult. See also the next item.
4300 79. The ultimate address timeout was not being applied to errors that involved
4301 a combination of host plus message (for example, a timeout on a MAIL
4302 command). When an address resolved to a number of possible hosts, and they
4303 were not all tried for each delivery (e.g. because of hosts_max_try), a
4304 message could remain on the queue longer than the retry timeout.
4306 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4309 81. Fixed an obscure SMTP outgoing bug which required at least the following
4310 conditions: (a) there was another message waiting for the same server;
4311 (b) the server returned 5xx to all RCPT commands in the first message so
4312 that the message was not completed; (c) the server dropped the connection
4313 or gave a negative response to the RSET that Exim sends to abort the
4314 transaction. The observed case was a dropped connection after DATA that had
4315 been sent in pipelining mode. That is, the server had advertised PIPELINING
4316 but was not implementing it correctly. The effect of the bug was incorrect
4317 behaviour, such as trying another host, and this could lead to a crash.
4323 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4324 and daemon.c were passed as pointers to ints; they should have been
4325 pointers to socklen_t variables (which are typically unsigned ints).
4327 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4330 3. Fixed a really odd bug that affected only the testing scheme; patching a
4331 certain fixed string in the binary changed the value of another string that
4332 happened to be identical to the end of the original first string.
4334 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4335 name", it returns that address as the IP address. On some operating
4336 systems (e.g. Solaris), it also passes back the IP address string as the
4337 "host name". However, on others (e.g. Linux), it passes back an empty
4338 string. Exim wasn't checking for this, and was changing the host name to an
4339 empty string, assuming it had been canonicized.
4341 5. Although rare, it is permitted to have more than one PTR record for a given
4342 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4343 all the names associated with an address, because they do in Solaris.
4344 However, it seems that they do not in Linux for data that comes from the
4345 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4346 I found this out when I moved to a new Linux workstation and tried to run
4347 the Exim test suite.
4349 To get round this problem I have changed the code so that it now does its
4350 own call to the DNS to look up PTR records when searching for a host name.
4351 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4352 addresses that are only in /etc/hosts are still found.
4354 This behaviour is, however, controlled by an option called host_lookup_
4355 order, which defaults to "bydns:byaddr". If people want to use the other
4356 order, or indeed, just use one or the other means of lookup, they can
4357 specify it in this variable.
4359 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4360 some operating systems, this comes back from gethostbyaddr() as an empty
4361 string, and this is what Exim used to test for. However, it seems that in
4362 other systems, "." is yielded. Exim now tests for this case too.
4364 7. The values of check_spool_space and check_log_space are now held internally
4365 as a number of kilobytes instead of an absolute number of bytes. If a
4366 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4367 kilobyte. This means that much larger values can be stored.
4369 8. Exim monitor: an attempt to get the action menu when not actually pointing
4370 at a message produces an empty menu entitled "No message selected". This
4371 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4372 no entries in it ("Shell widget menu has zero width and/or height"). So I
4373 have added a single, blank menu entry in this case.
4375 9. Added ${quote_local_part.
4377 10. MIME decoding is now applied to the contents of Subject: header lines when
4380 11. Now that a reference to $sender_host_address automatically causes a reverse
4381 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4382 host lookup before query-style lookups in lists that might use this
4383 variable. This has therefore been abolished, and the "net-" prefix is no
4384 longer necessary for query-style lookups.
4386 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4387 have been a typo for LFLAGS, so it has been changed.
4389 13. The install script calls Exim with "-C /dev/null" in order to find the
4390 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4391 to be output. Howeve, since Exim outputs its version number before the
4392 error, it didn't break the script. It just looked ugly. I fixed this by
4393 always allowing "-C /dev/null" if the caller is root.
4395 14. Ignore overlarge ACL variable number when reading spool file - insurance
4396 against a later release with more variables having written the file.
4398 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4399 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4400 this, as well as the form without the "IPv6" on the front.
4402 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4403 OS/Makefile-Darwin file.
4405 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4406 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4408 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4409 variable. However, some OS use ints for inodes. I've added cast to long int
4410 to get rid of the compiler warning.
4412 19. I had forgotten to lock out "/../" in configuration file names when
4413 ALT_CONFIG_PREFIX was set.
4415 20. Routers used for verification do not need to specify transports. However,
4416 if such a router generated a host list, and callout was configured, Exim
4417 crashed, because it could not find a port number from the (non-existent)
4418 transport. It now assumes port 25 in this circumstance.
4420 21. Added the -t option to exigrep.
4422 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4423 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4424 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4425 them is set, LOOKUP_LSEARCH is forced.
4427 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4429 24. Added sender and host information to the "rejected by local_scan()" log
4430 line; previously there was no indication of these.
4432 25. Added .include_if_exists.
4434 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4435 Linux. It turns out that not all file systems support this. Apparently some
4436 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4437 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4438 not supported on the file descriptor, is now ignored when Exim is trying to
4439 sync a directory. This applies only to Linux.
4441 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4443 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4444 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4447 29. Added support for maildirsize files from supplied patch (modified a bit).
4449 30. The use of :fail: followed by an empty string could lead Exim to respond to
4450 sender verification failures with (e.g.):
4452 550 Verification failed for <xxx>
4453 550 Sender verify failed
4455 where the first response line was missing the '-' that indicates it is not
4456 the final line of the response.
4458 31. The loop for finding the name of the user that called Exim had a hardwired
4459 limit of 10; it now uses the value of finduser_retries, which is used for
4460 all other user lookups.
4462 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4465 33. Exim was neglecting to zero errno before one call of strtol() when
4466 expanding a string and expecting an integer value. On some systems this
4467 resulted in spurious "integer overflow" errors. Also, it was casting the
4468 result into an int without checking.
4470 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4471 did not work. The code looks as if it has *never* worked, though it appears
4472 to have been documented since at least releast 1.62. I have made it work.
4474 35. The "timeout_DNS" error in retry rules, also documented since at least
4475 1.62, also never worked. As it isn't clear exactly what this means, and
4476 clearly it isn't a major issue, I have abolished the feature by treating it
4477 as "timeout", and writing a warning to the main and panic logs.
4479 36. The display of retry rules for -brt wasn't always showing the error code
4482 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4483 timeout_connect_A, timeout_connect_MX.
4485 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4486 to the empty sender.
4488 39. The daemon was not analysing the content of -oX till after it had closed
4489 stderr and disconnected from the controlling terminal. This meant that any
4490 syntax errors were only noted on the panic log, and the return code from
4491 the command was 0. By re-arranging the code a little, I've made the
4492 decoding happen first, so such errors now appear on stderr, and the return
4493 code is 1. However, the actual setting up of the sockets still happens in
4494 the disconnected process, so errors there are still only recorded on the
4497 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4498 connections (as happens on some IP stacks) was logged at start up time as
4499 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4500 it from "IPv6 and IPv4", which means that two separate sockets are being
4503 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4504 says whether AF_INET or AF_INET6 was passed as an argument.
4506 42. Exiwhat output was messed up when time zones were included in log
4509 43. Exiwhat now gives more information about the daemon's listening ports,
4510 and whether -tls-on-connect was used.
4512 44. The "port" option of the smtp transport is now expanded.
4514 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4515 silently ignored. Now an error message is written to the main and panic
4518 46. There's a new ACL modifier called "logwrite" which writes to a log file
4519 as soon as it is encountered.
4521 47. Added $local_user_uid and $local_user_gid at routing time.
4523 48. Exim crashed when trying to verify a sender address that was being
4526 49. Exim was recognizing only a space character after ".include". It now also
4527 recognizes a tab character.
4529 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4530 extracting the relevant information from the specification. The man page no
4531 longer contains scrambled data for the -d option, and I've added a section
4532 at the front about calling Exim under different names.
4534 51. Added "extra_headers" argument to the "mail" command in filter files.
4536 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4539 53. Installed eximstats 1.29.
4541 54. Added transport_filter_timeout as a generic transport option.
4543 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4544 Cc: header lines. This was required by RFC 822, but it not required by RFC
4547 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4548 incoming messages that did not have them. Now it does so only if the
4549 message originates locally, that is, if there is no associated remote host
4550 address. When Resent- header lines are present, this applies to the Resent-
4551 lines rather than the non-Resent- lines.
4553 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4554 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4556 58. Messages for configuration errors now include the name of the main
4557 configuration file - useful now that there may be more than one file in a
4558 list (.included file names were always shown).
4560 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4561 for those rare installations that do not start the daemon as root or run it
4562 setuid root. I've cut out the call to initgroups() if the daemon is not
4565 60. The Exim user and group can now be bound into the binary as text strings
4566 that are looked up at the start of Exim's processing.
4568 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4570 62. Added $mailstore_basename variable.
4572 63. Installed patch to sieve.c from Michael Haardt.
4574 64. When Exim failed to open the panic log after failing to open the main log,
4575 the original message it was trying to log was written to stderr and debug
4576 output, but if they were not available (the usual case in production), it
4577 was lost. Now it is written to syslog before the two lines that record the
4578 failures to open the logs.
4580 65. Users' Exim filters run in subprocesses under the user's uid. It is
4581 possible for a "deliver" command or an alias in a "personal" command to
4582 provoke an address rewrite. If logging of address rewriting is configured,
4583 this fails because the process is not running as root or exim. There may be
4584 a better way of dealing with this, but for the moment (because 4.30 needs
4585 to be released), I have disabled address rewrite logging when running a
4586 filter in a non-root, non-exim process.
4592 1. The buildconfig auxiliary program wasn't quoting the value set for
4593 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4594 not defined. This bug was masked in 4.22 by the effect that was fixed in
4597 2. Some messages that were rejected after a message id was allocated were
4598 shown as "incomplete" by exigrep. It no longer does this for messages that
4599 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4601 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4602 have allow_domain_literals set, the ID did not get logged in the <= line.
4603 Domain literals are now always recognized in Message-ID: header lines.
4605 4. The first argument for a ${extract expansion item is the key name or field
4606 number. Leading and trailing spaces in this item were not being ignored,
4607 causing some misleading effects.
4609 5. When deliver_drop_privilege was set, single queue runner processes started
4610 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4611 same command in the process it spins off) were not dropping privilege.
4613 6. When the daemon running as "exim" started a queue runner, it always
4614 re-executed Exim in the spun-off process. This is a waste of effort when
4615 deliver_drop_privilege is set. The new process now just calls the
4616 queue-runner function directly.
4622 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4625 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4626 ignored. Though the use of -oP was forcing the writing of a pid file, it
4627 was always written to the default place.
4629 3. If the message "no IP address found for host xxxx" is generated during
4630 incoming verification, it is now followed by identification of the incoming
4631 connection (so you can more easily find what provoked it).
4633 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4635 5. Added some features to "harden" Exim a bit more against certain attacks:
4637 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4638 be put in Local/Makefile. This is like the never_users runtime option,
4639 but it cannot be overridden. The default setting is "root".
4641 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4642 prefix string with which any file named in a -C command line option
4645 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4646 is retained for -C and -D only if the caller of Exim is root. Without
4647 it, the exim user may also use -C and -D and retain privilege.
4649 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4650 command line option is disabled.
4652 6. Macro names set by the -D option must start with an upper case letter, just
4653 like macro names defined in the configuration file.
4655 7. Added "dereference=" facility to LDAP.
4657 8. Two instances of the typo "uknown" in the source files are fixed.
4659 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4660 the Configure-Makefile script screwed up while processing it.
4662 10. Incorporated PCRE 4.4.
4664 11. The SMTP synchronization check was not operating right at the start of an
4665 SMTP session. For example, it could not catch a HELO sent before the client
4666 waited for the greeting. There is now a check for outstanding input at the
4667 point when the greeting is written. Because of the duplex, asynchronous
4668 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4669 way, but not yet received, when the check is performed.
4671 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4672 on TCP/IP sockets, because this apparently causes some broken clients to
4675 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4676 unchanged) from the Cygwin maintainer.
4678 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4679 was supported instead of testing for mbx. Effectively a typo.
4681 15. The spa authenticator server code was not checking that the input it
4682 received was valid base64.
4684 16. The debug output line for the "set" modifier in ACLs was not showing the
4685 name of the variable that was being set.
4687 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4689 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4690 lookup on its own. Something else was needed to trigger the lookup. For
4691 example, a match in host_lookup or the need for a host name in a host list.
4692 Now, if $sender_host_name is referenced and the host name has not yet been
4693 looked up, a lookup is performed. If the lookup fails, the variable remains
4694 empty, and $host_lookup_failed is set to "1".
4696 19. Added "eqi" as a case-independent comparison operator.
4698 20. The saslauthd authentication condition could segfault if neither service
4699 nor realm was specified.
4701 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4702 error message that was logged was misleading, and incoming SMTP
4703 connections were dropped. The message is now more accurate, and temporary
4704 errors are given to SMTP connections.
4706 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4707 (except RSET) with a 421 error, until QUIT is received. However, it was
4708 failing to send a response to QUIT.
4710 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4711 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4712 ACL is now run after the helo_try_verify_hosts code.
4714 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4715 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4716 case-independent, so other case variants are also recognized). Apparently
4717 some systems use these upper case variants.
4719 25. If more than two messages were waiting for the same host, and a transport
4720 filter was specified for the transport, Exim sent two messages over the
4721 same TCP/IP connection, and then failed with "socket operation on non-
4722 socket" when it tried to send the third.
4724 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4726 27. The extern definition of crypt16() in expand.c was not being excluded when
4727 the OS had its own crypt16() function.
4729 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4730 as a preferred synonym for return_size_limit, both as an option and as an
4733 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4735 30. Changed the default configuration ACL to relax the local part checking rule
4736 for addresses that are not in any local domains. For these addresses,
4737 slashes and pipe symbols are allowed within local parts, but the sequence
4738 /../ is explicitly forbidden.
4740 31. SPA server authentication was not clearing the challenge buffer before
4743 32. log_message in a "warn" ACL statement was writing to the reject log as
4744 well as to the main log, which contradicts the documentation and doesn't
4745 seem right (because no rejection is happening). So I have stopped it.
4747 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4748 However, I am unable to do any testing of this.
4750 34. Fixed an infelicity in the appendfile transport. When checking directories
4751 for a mailbox, to see if any needed to be created, it was accidentally
4752 using path names with one or more superfluous leading slashes; tracing
4753 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4755 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4756 contents are added to the log line that is written for every discarded
4757 recipient. (Previously a log_message setting was ignored.)
4759 36. The ${quote: operator now quotes the string if it is empty.
4761 37. The install script runs exim in order to find its version number. If for
4762 some reason other than non-existence or emptiness, which it checks, it
4763 could not run './exim', it was installing it with an empty version number,
4764 i.e. as "exim-". This error state is now caught, and the installation is
4767 38. An argument was missing from the function that creates an error message
4768 when Exim fails to connect to the socket for saslauthd authentication.
4769 This could cause Exim to crash, or give a corrupted message.
4771 39. Added isip, isip4, and isip6 to ${if conditions.
4773 40. The ACL variables $acl_xx are now saved with the message, and can be
4774 accessed later in routers, transports, and filters.
4776 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4777 strings in the file are not string-expanded.
4779 42. If a MAIL command specified a SIZE value that was too large to fit into an
4780 int variable, the check against message_size_limit failed. Such values are
4781 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4782 day this will have to be increased, but I don't think I want to be around
4783 when emails are that large.
4790 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4791 iconv() is not standard in FreeBSD.
4793 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4794 IPv6 enabled. The observed symptom was a segmentation fault on return from
4795 the function os_common_find_running_interfaces() in src/os.c.
4797 3. In the check_special_case() function in daemon.c I had used "errno" as an
4798 argument name, which causes warnings on some systems. This was basically a
4799 typo, since it was named "eno" in the comments!
4801 4. The code that waits for the clock to tick (at a resolution of some fraction
4802 of a second) so as to ensure message-id uniqueness was always waiting for
4803 at least one whole tick, when it could have waited for less. [This is
4804 almost certainly not relevant at current processor speeds, where it is
4805 unlikely to ever wait at all. But we try to future-proof.]
4807 5. The function that sleeps for a time interval that includes fractions of a
4808 second contained a race. It did not block SIGALRM between setting the
4809 timer, and suspending (a couple of lines later). If the interval was short
4810 and the sigsuspend() was delayed until after it had expired, the suspension
4811 never ended. On busy systems this could lead to processes getting stuck for
4814 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4815 process, before it forks any delivery processes. The open lookup caching
4816 mechanism meant that the open file or database connection was passed into
4817 the delivery process. The problem was that delivery processes always tidy
4818 up cached lookup data. This could cause a problem for the next delivery
4819 process started by the queue runner, because the external queue runner
4820 process does not know about the closure. So the next delivery process
4821 still has data in the lookup cache. In the case of a file lookup, there was
4822 no problem because closing a file descriptor in a subprocess doesn't affect
4823 the parent. However, if the lookup was caching a connection to a database,
4824 the connection was closed, and the second delivery process was likely to
4825 see errors such as "PGSQL: query failed: server closed the connection
4826 unexpectedly". The problem has been fixed by closing all cached lookups
4827 in a queue runner before running a delivery process.
4829 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4830 seem to have the "const" qualifier which it has on other OS. I've
4833 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4834 *of the same type* (client or server) with the same public name that an
4835 error should be diagnosed.
4837 9. When Exim looked up a host name for an IP address, but failed to find the
4838 original IP address when looking up the host name (a safety check), it
4839 output the message "<ip address> does not match any IP for NULL", which was
4840 confusing, to say the least. The bug was that the host name should have
4841 appeared instead of "NULL".
4843 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4844 user that is built into the binary, and the -C or -D options is used, root
4845 privilege is dropped before the configuration file is read. In addition,
4846 logging is switched to stderr instead of the normal log files. If the
4847 configuration then re-defines the Exim user, the unprivileged environment
4848 is probably not what is expected, so Exim logs a panic warning message (but
4851 However, if deliver_drop_privilege is set, the unprivileged state may well
4852 be exactly what is intended, so the warning has been cut out in that case,
4853 and Exim is allowed to try to write to its normal log files.
4859 1. smtp_return_error_details was not giving details for temporary sender
4860 or receiver verification errors.
4862 2. Diagnose a configuration error if two authenticators have the same public
4865 3. Exim used not to create the message log file for a message until the first
4866 delivery attempt. This could be confusing when incoming messages were held
4867 for policy or load reasons. The message log file is now created at the time
4868 the message is received, and an initial "Received" line is written to it.
4870 4. The automatically generated man page for command line options had a minor
4871 bug that caused no ill effects; however, a more serious problem was that
4872 the procedure for building the man page automatically didn't always
4873 operate. Consequently, release 4.20 contains an out-of-date version. This
4874 shouldn't happen again.
4876 5. When building Exim with embedded Perl support, the script that builds the
4877 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4878 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4880 6. The freeze_tell option was not being used for messages that were frozen on
4881 arrival, either by an ACL or by local_scan().
4883 7. Added the smtp_incomplete_transaction log selector.
4885 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4886 was accepting AUTH without a new EHLO.
4888 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4889 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4892 10. Logging of TCP/IP connections (when configured) now happens in the main
4893 daemon process instead of the child process, so that the TCP/IP connection
4894 count is more accurate (but it can never be perfect).
4896 11. The use of "drop" in a nested ACL was not being handled correctly in the
4897 outer ACL. Now, if condition failure induced by the nested "drop" causes
4898 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4899 or "require"), the connection is dropped.
4901 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4902 that yield "discard" can now be used with an "accept" or a "discard" verb,
4903 but an error is generated for any others (because I can't see a useful way
4904 to define what should happen).
4906 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4907 are now processed in the same way as lines in the Exim configuration file.
4908 In particular, continuation lines are supported.
4910 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4912 15. Added -ti meaning -t -i.
4914 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4915 domains, and warn by logging if others are found.
4917 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4918 returned from GIFCONF when Exim is trying to find the list of interfaces on
4919 a host. The code in os.c has been modified to copy each ifreq to an aligned
4920 structure in all cases.
4922 Also, in some cases, the returned ifreq's were being copied to a 'struct
4923 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4924 means the last couple of bytes of an IPv6 address could be chopped if the
4925 ifreq contained only a normal sockaddr (14 bytes storage).
4927 18. Named domain lists were not supported in the hosts_treat_as_local option.
4928 An entry such as +xxxx was not recognized, and was treated as a literal
4931 19. Ensure that header lines added by a DATA ACL are included in the reject log
4932 if the ACL subsequently rejects the message.
4934 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4935 MD5 (which is deprecated).
4937 21. When testing a filter file using -bf, Exim was writing a message when it
4938 took the sender from a "From " line in the message, but it was not doing so
4939 when it took $return_path from a Return-Path: header line. It now does.
4941 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4942 with a valid header line field name (a series of printing characters
4943 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4945 23. Changed "disc" in the source to "disk" to conform to the documentation and
4946 the book and for uniformity.
4948 24. Ignore Sendmail's -Ooption=value command line item.
4950 25. When execve() failed while trying to run a command in a pipe transport,
4951 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4952 could be confused with a return value of 69 from the command itself. This
4953 has been changed to 127, the value the shell returns if it is asked to run
4954 a non-existent command. The wording for the related log line suggests a
4955 non-existent command as the problem.
4957 26. If received_header_text expands to an empty string, do not add a Received:
4958 header line to the message. (Well, it adds a token one on the spool, but
4959 marks it "old" so that it doesn't get used or transmitted.)
4961 27. Installed eximstats 1.28 (addition of -nt option).
4963 28. There was no check for failure on the call to getsockname() in the daemon
4964 code. This can fail if there is a shortage of resources on the system, with
4965 ENOMEM, for example. A temporary error is now given on failure.
4967 29. Contrary to the C standard, it seems that in some environments, the
4968 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4969 program. Exim now does this explicitly; it affects the formatting of
4970 timestamps using strftime().
4972 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4973 complaints. I've now initialized all the variables, to avoid this.
4975 32. Header lines added by a system filter were not being "seen" during
4976 transport-time rewrites.
4978 33. The info_callback() function passed to OpenSSL is set up with type void
4979 (*)(SSL *, int, int), as described somewhere. However, when calling the
4980 function (actually a macro) that sets it up, the type void(*)() is
4981 expected. I've put in a cast to prevent warnings from picky compilers.
4983 34. If a DNS black list lookup found a CNAME record, but there were no A
4984 records associated with the domain it pointed at, Exim crashed.
4986 35. If a DNS black list lookup returned more than one A record, Exim ignored
4987 all but the first. It now scans all returned addresses if a particular IP
4988 value is being sought. In this situation, the contents of the
4989 $dnslist_value variable are a list of all the addresses, separated by a
4992 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4993 to accept a host name and all its aliases if the forward lookup for any of
4994 them yielded the IP address of the incoming connection. Now it accepts only
4995 those names whose forward lookup yields the correct IP address. Any other
4996 names are discarded. This closes a loophole whereby a rogue DNS
4997 administrator could create reverse DNS records to break through a
4998 wildcarded host restriction in an ACL.
5000 37. If a user filter or a system filter that ran in a subprocess used any of
5001 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5002 the wrong values were passed to the pipe command ($thisaddress had the
5003 value of $0, $0 had the value of $1, etc). This bug was introduced by
5004 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5006 38. Improved the line breaking for long SMTP error messages from ACLs.
5007 Previously, if there was no break point between 40 and 75 characters, Exim
5008 left the rest of the message alone. Two changes have been made: (a) I've
5009 reduced the minimum length to 35 characters; (b) if it can't find a break
5010 point between 35 and 75 characters, it looks ahead and uses the first one
5011 that it finds. This may give the occasional overlong line, but at least the
5012 remaining text gets split now.
5014 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5015 file descriptors might be low, and that setting 1000 would always raise it.
5016 It turns out that in some environments, the limit is already over 1000 and
5017 that lowering it causes trouble. So now Exim takes care not to decrease it.
5019 40. When delivering a message, the value of $return_path is set to $sender_
5020 address at the start of routing (routers may change the value). By an
5021 oversight, this default was not being set up when an address was tested by
5022 -bt or -bv, which affected the outcome if any router or filter referred to
5025 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5026 log only when "message" or "log_message" are set. However, if one of the
5027 conditions was an address verification, or a call to a nested ACL, the
5028 messages generated by the underlying test were being passed through. This
5029 no longer happens. The underlying message is available in $acl_verify_
5030 message for both "message" and "log_message" expansions, so it can be
5031 passed through if needed.
5033 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5034 new expansion $bh_ to give the encoded byte string without charset
5035 translation. Translation happens only if iconv() is available; HAVE_ICONV
5036 indicates this at build time. HEADERS_CHARSET gives the charset to
5037 translate to; headers_charset can change it in the configuration, and
5038 "headers charset" can change it in an individual filter file.
5040 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5041 that creates RFC 2047 encoded "words" labels them as that charset instead
5042 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5043 expansion operator; (ii) when Exim creates a From: line for a local
5044 message; (iii) when a header line is rewritten to include a "phrase" part.
5046 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5047 buggy, causing it to skip the first lines of messages whose message ID
5048 ended in 'D'. This would not have bitten before Exim release 4.14, because
5049 message IDs were unlikely to end in 'D' before then. The effect was to have
5050 incorrect size information for certain domains.
5052 45. #include "config.h" was missing at the start of the crypt16.c module. This
5053 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5056 46. If there was a timeout during a "random" callout check, Exim treated it as
5057 a failure of the random address, and carried on sending RSET and the real
5058 address. If the delay was just some slowness somewhere, the response to the
5059 original RCPT would be taken as a response to RSET and so on, causing
5060 mayhem of various kinds.
5062 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5063 when I implemented it. It didn't allow for the fact that some option values
5064 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5065 the right test for positive values.
5067 48. Domain names in DNS records are case-independent. Exim always looks them up
5068 in lower case. Some resolvers return domain names in exactly the case they
5069 appear in the zone file, that is, they may contain uppercase letters. Not
5070 all resolvers do this - some return always lower case. Exim was treating a
5071 change of case by a resolver as a change of domain, similar to a widening
5072 of a domain abbreviation. This triggered its re-routing code and so it was
5073 trying to route what was effectively the same domain again. This normally
5074 caused routing to fail (because the router wouldn't handle the domain
5075 twice). Now Exim checks for this case specially, and just changes the
5076 casing of the domain that it ultimately uses when it transmits the message
5079 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5082 50. If a filter generated a file delivery with a non-absolute name (possible if
5083 no home directory exists for the router), the forbid_file option was not
5086 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5087 the existing equality matching.
5089 52. Exim was using ints instead of ino_t variables in some places where it was
5090 dealing with inode numbers.
5092 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5093 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5094 called TMPDIR, and if it finds it is different, it changes its value.
5096 54. The smtp_printf() function is now made available to local_scan() so
5097 additional output lines can be written before returning. There is also an
5098 smtp_fflush() function to enable the detection of a dropped connection.
5099 The variables smtp_input and smtp_batched_input are exported to
5102 55. Changed the default runtime configuration: the message "Unknown user"
5103 has been removed from the ACL, and instead placed on the localuser router,
5104 using the cannot_route_message feature. This means that any verification
5105 failures that generate their own messages won't get overridden. Similarly,
5106 the "Unrouteable address" message that was in the ACL for unverifiable
5107 relay addresses has also been removed.
5109 56. Added hosts_avoid_esmtp to the smtp transport.
5111 57. The exicyclog script was not checking for the esoteric option
5112 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5113 will work only if exicyclog is run under the appropriate euid.
5115 58. Following a discussion on the list, the rules by which Exim recognises line
5116 endings on incoming messages have been changed. The -dropcr and drop_cr
5117 options are now no-ops, retained only for backwards compatibility. The
5118 following line terminators are recognized: LF CRLF CR. However, special
5119 processing applies to CR:
5121 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5122 nor a local message in the state where . is a terminator.
5124 (ii) If a bare CR is encountered in a header line, an extra space is added
5125 after the line terminator so as not to end the header. The reasoning
5126 behind this is that bare CRs in header lines are most likely either
5127 to be mistakes, or people trying to play silly games.
5129 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5130 was being incorrectly given as 18 bytes larger than it should have been.
5131 This is a VOB (very old bug).
5133 60. This may never have affected anything current, but just in case it has:
5134 When the local host is found other than at the start of a list of hosts,
5135 the local host, those with the same MX, and any that follow, are discarded.
5136 When the list in question was part of a longer list of hosts, the following
5137 hosts (not currently being processed) were also being discarded. This no
5138 longer happens. I'm not sure if this situation could ever has previously
5141 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5144 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5145 header. This is something that is recommended in a new Internet Draft, and
5146 is something that is documented as being done by Sendmail. There are two
5147 possible values. For messages generated by the autoreply transport, Exim
5150 Auto-Submitted: auto-replied
5152 whereas for all other generated messages (e.g. bounces) it adds
5154 Auto-Submitted: auto-generated
5156 63. The "personal" condition in filters now includes a test for the
5157 Auto-Submitted: header. If it contains the string "auto-" the message it
5158 not considered personal.
5160 64. Added rcpt_include_affixes as a generic transport option.
5162 65. Added queue_only_override (default true).
5164 66. Added the syslog_duplication option.
5166 67. If what should have been the first header line of a message consisted of
5167 a space followed by a colon, Exim was mis-interpreting it as a header line.
5168 It isn't of course - it is syntactically invalid and should therefore be
5169 treated as the start of the message body. The misbehaviour could have
5170 caused a number of strange effects, including loss of data in subsequent
5171 header lines, and spool format errors.
5173 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5174 client host had authenticated. This control can now be exercised by an ACL
5175 for more flexibility.
5177 69. By default, callouts do not happen when testing with -bh. There is now a
5178 variant, -bhc, which does actually run the callout code, including
5179 consulting and updating the callout cache.
5181 70. Added support for saslauthd authentication, courtesy of Alexander
5184 71. If statvfs() failed on the spool or log directories while checking their
5185 size for availability, Exim confusingly gave the error "space shortage".
5186 Furthermore, in debugging mode it crashed with a floating point exception.
5187 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5188 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5189 serious problem, Exim now writes to the main and panic logs when this
5190 happens, with details of the failure. It then refuses to accept the
5191 incoming message, giving the message "spool directory problem" or "log
5192 directory problem" with a 421 code for SMTP messages.
5194 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5195 0, 1, and 2 exist, because some OS complain for execs without them (see
5196 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5197 descriptors. However, the code omitted to check that the open succeeded,
5198 causing mysterious errors if for some reason the permissions on /dev/null
5199 got screwed. Now Exim writes a message to the main and panic logs, and
5200 bombs out if it can't open /dev/null.
5202 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5203 interact so that it is all more flexible. It is supposed to remain
5204 backwards compatible. Also added extra_local_interfaces.
5206 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5207 to bomb out with an assertion failure - to the client this appears as a
5208 connection drop. This problem occurs in the part of the code that was taken
5209 from the Samba project. Fortunately, the assertion is in a very simple
5210 function, so I have fixed this by reproducing the function inline in the
5211 one place where it is called, and arranging for authentication to fail
5212 instead of killing the process with assert().
5214 75. The SPA client code was not working when the server requested OEM rather
5215 than Unicode encoding.
5217 76. Added code to make require_files with a specific uid setting more usable in
5218 the case where statting the file as root fails - usually a non-root-mounted
5219 NFS file system. When this happens and the failure is EACCES, Exim now
5220 forks a subprocess and does the per-uid checking as the relevant uid.
5222 77. Added process_log_path.
5224 78. If log_file_path was not explicitly set, a setting of check_log_space or
5225 check_log_inodes was ignored.
5227 79. If a space check for the spool or log partitions fails, the incident is now
5228 logged. Of course, in the latter case the data may get lost...
5230 80. Added the %p formatting code to string_format() so that it can be used to
5231 print addresses in debug_print(). Adjusted all the address printing in the
5232 debugging in store.c to use %p rather than %d.
5234 81. There was a concern that a line of code in smtp_in.c could overflow a
5235 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5236 initially expressed, the concern was not well-founded, because trailing
5237 spaces are removed early. However, if the trailing spaces were followed by
5238 a NULL, they did not get removed, so the overflow was possible. Two fixes
5241 (a) I re-wrote the offending code in a cleaner fashion.
5242 (b) If an incoming SMTP command contains a NULL character, it is rejected
5245 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5246 runs initgroups(), so that if the Exim user is in any additional groups,
5247 they will be used during message reception.
5253 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob