1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 NM/01 Workround for PCRE version dependancy in version reporting
12 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
13 This fixes portability to compilers other than gcc, notably
14 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
16 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
17 makefiles for portability to HP-UX and POSIX correctness.
19 PP/01 Permit LOOKUP_foo enabling on the make command-line.
20 Also via indented variable definition in the Makefile.
21 (Debugging by Oliver Heesakkers).
23 PP/02 Restore caching of spamd results with expanded spamd_address.
24 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
26 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
27 Improves build reliability. Fix from: Frank Elsner
29 NM/02 Fix wide character breakage in the rfc2047 coding
30 Fixes bug 1064. Patch from Andrey N. Oktyabrski
32 NM/03 Allow underscore in dnslist lookups
33 Fixes bug 1026. Patch from Graeme Fowler
35 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
36 Code patches from Adam Ciarcinski of NetBSD.
38 NM/04 Fixed exiqgrep to cope with mailq missing size issue
45 TF/01 Failure to get a lock on a hints database can have serious
46 consequences so log it to the panic log.
48 TF/02 Log LMTP confirmation messages in the same way as SMTP,
49 controlled using the smtp_confirmation log selector.
51 TF/03 Include the error message when we fail to unlink a spool file.
53 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
54 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
55 for maintaining out-of-tree patches for some time.
57 PP/01 Bugzilla 139: Documentation and portability issues.
58 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
59 Handle per-OS dynamic-module compilation flags.
61 PP/02 Let /dev/null have normal permissions.
62 The 4.73 fixes were a little too stringent and complained about the
63 permissions on /dev/null. Exempt it from some checks.
64 Reported by Andreas M. Kirchwitz.
66 PP/03 Report version information for many libraries, including
67 Exim version information for dynamically loaded libraries. Created
68 version.h, now support a version extension string for distributors
69 who patch heavily. Dynamic module ABI change.
71 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
72 privilege escalation vulnerability whereby the Exim run-time user
73 can cause root to append content of the attacker's choosing to
76 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
79 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
80 If dropping privileges for untrusted macros, we disabled normal logging
81 on the basis that it would fail; for the Exim run-time user, this is not
82 the case, and it resulted in successful deliveries going unlogged.
83 Fixed. Reported by Andreas Metzler.
89 PP/01 Date: & Message-Id: revert to normally being appended to a message,
90 only prepend for the Resent-* case. Fixes regression introduced in
91 Exim 4.70 by NM/22 for Bugzilla 607.
93 PP/02 Include check_rfc2047_length in configure.default because we're seeing
94 increasing numbers of administrators be bitten by this.
96 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
98 PP/03 Bugzilla 994: added openssl_options main configuration option.
100 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
102 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
104 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
106 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
107 without a peer certificate, leading to a segfault because of an
108 assumption that peers always have certificates. Be a little more
109 paranoid. Problem reported by Martin Tscholak.
111 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
112 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
113 NB: ClamAV planning to remove STREAM in "middle of 2010".
114 CL also introduces -bmalware, various -d+acl logging additions and
115 more caution in buffer sizes.
117 PP/09 Implemented reverse_ip expansion operator.
119 PP/10 Bugzilla 937: provide a "debug" ACL control.
121 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
123 PP/12 Bugzilla 973: Implement --version.
125 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
127 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
129 PP/15 Bugzilla 816: support multiple condition rules on Routers.
131 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
132 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
133 ignore trailing whitespace.
135 JJ/02 prevent non-panic DKIM error from being sent to paniclog
137 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
140 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
141 Notification from Dr Andrew Aitchison.
143 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
144 ExtendedDetectionInfo response format.
145 Notification from John Horne.
147 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
150 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
151 XSL and documented dependency on system catalogs, with examples of how
154 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
157 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
158 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
159 configuration file which is writeable by the Exim user or group.
161 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
162 of configuration files to cover files specified with the -C option if
163 they are going to be used with root privileges, not just the default
166 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
167 option (effectively making it always true).
169 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
170 files to be used while preserving root privileges.
172 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
173 that rogue child processes cannot use them.
175 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
176 run-time user, instead of root.
178 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
179 Exim run-time user without dropping privileges.
181 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
182 result string, instead of calling string_vformat() twice with the same
185 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
186 for other users. Others should always drop root privileges if they use
187 -C on the command line, even for a whitelisted configure file.
189 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
191 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
197 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
198 $data_path, and $header_path variables; fixed documentation bugs and
201 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
202 exipick to access non-standard spools, including the "frozen" queue
205 NM/01 Bugzilla 965: Support mysql stored procedures.
206 Patch from Alain Williams
208 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
210 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
211 Patch from Andreas Metzler
213 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
214 Patch from Kirill Miazine
216 NM/05 Bugzilla 671: Added umask to procmail example.
218 JJ/03 installed exipick 20100323.0, fixing doc bug
220 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
221 directory. Notification and patch from Dan Rosenberg.
223 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
225 TK/02 Improve log output when DKIM signing operation fails.
227 MH/01 Treat the transport option dkim_domain as a colon separated
228 list, not as a single string, and sign the message with each element,
229 omitting multiple occurences of the same signer.
231 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
232 Bugzilla 985, 986. Patch by Simon Arlott
234 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
235 Patch by Simon Arlott
237 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
238 MBX locking. Notification from Dan Rosenberg.
244 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
246 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
248 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
250 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
252 NM/04 Bugzilla 915: Flag broken perl installation during build.
258 TK/01 Added patch by Johannes Berg that expands the main option
259 "spamd_address" if it starts with a dollar sign.
261 TK/02 Write list of recipients to X-Envelope-Sender header when building
262 the mbox-format spool file for content scanning (suggested by Jakob
265 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
266 (http://www.dcc-servers.net/) support via dccifd. Activated by
267 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
269 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
270 by Mark Daniel Reidel <mr@df.eu>.
272 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
273 When building exim an external PCRE library is now needed -
274 PCRE is a system library on the majority of modern systems.
275 See entry on PCRE_LIBS in EDITME file.
277 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
278 conversation. Added nologin parameter to request.
279 Patch contributed by Kirill Miazine.
281 TF/01 Do not log submission mode rewrites if they do not change the address.
283 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
285 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
286 log files in place. Contributed by Roberto Lima.
288 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
290 TF/03 Bugzilla 615: When checking the local_parts router precondition
291 after a local_part_suffix or local_part_prefix option, Exim now
292 does not use the address's named list lookup cache, since this
293 contains cached lookups for the whole local part.
295 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
296 Robert Millan. Documentation is in experimental-spec.txt.
298 TF/04 Bugzilla 668: Fix parallel build (make -j).
300 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
302 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
303 Patch provided by Jan Srzednicki.
305 TF/05 Leading white space used to be stripped from $spam_report which
306 wrecked the formatting. Now it is preserved.
308 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
309 that they are available at delivery time.
311 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
313 TF/08 TLS error reporting now respects the incoming_interface and
314 incoming_port log selectors.
316 TF/09 Produce a more useful error message if an SMTP transport's hosts
317 setting expands to an empty string.
319 NM/06 Bugzilla 744: EXPN did not work under TLS.
320 Patch provided by Phil Pennock.
322 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
323 Patch provided by Richard Godbee.
325 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
326 acl_smtp_notquit, added index entry.
328 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
329 Patch provided by Eugene Bujak.
331 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
332 accept(). Patch provided by Maxim Dounin.
334 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
335 Patch provided by Phil Pennock.
337 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
339 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
340 Patch provided by Brad "anomie" Jorsch.
342 NM/14 Bugzilla 622: Added timeout setting to transport filter.
343 Patch provided by Dean Brooks.
345 TK/05 Add native DKIM support (does not depend on external libraries).
347 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
348 Patch provided by Graeme Fowler.
350 NM/16 Bugzilla 851: Documentation example syntax fix.
352 NM/17 Changed NOTICE file to remove references to embedded PCRE.
354 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
357 NM/19 Bugzilla 745: TLS version reporting.
358 Patch provided by Phil Pennock.
360 NM/20 Bugzilla 167: bool: condition support.
361 Patch provided by Phil Pennock.
363 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
364 clients. Patch provided by Phil Pennock.
366 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
367 Patch provided by Brad "anomie" Jorsch.
369 NM/23 Bugzilla 687: Fix misparses in eximstats.
370 Patch provided by Heiko Schlittermann.
372 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
373 Patch provided by Heiko Schlittermann.
375 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
376 plus update to original patch.
378 NM/26 Bugzilla 799: Documentation correction for ratelimit.
380 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
381 Patch provided by David Brownlee.
383 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
385 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
387 NM/30 Bugzilla 888: TLS documentation bugfixes.
389 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
391 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
392 Unlike the original bugzilla I have changed all shell scripts in src tree.
394 NM/33 Bugzilla 898: Transport filter timeout fix.
395 Patch by Todd Rinaldo.
397 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
398 Patch by Serge Demonchaux.
400 NM/35 Bugzilla 39: Base64 decode bug fixes.
401 Patch by Jakob Hirsch.
403 NM/36 Bugzilla 909: Correct connect() call in dcc code.
405 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
407 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
409 NM/39 Bugzilla 911: Fixed MakeLinks build script.
415 TK/01 Add preliminary DKIM support. Currently requires a forked version of
416 ALT-N's libdkim that I have put here:
417 http://duncanthrax.net/exim-experimental/
419 Note to Michael Haardt: I had to rename some vars in sieve.c. They
420 were called 'true' and it seems that C99 defines that as a reserved
421 keyword to be used with 'bool' variable types. That means you could
422 not include C99-style headers which use bools without triggering
423 build errors in sieve.c.
425 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
426 as mailq or other aliases. Changed the --help handling significantly
427 to do whats expected. exim_usage() emits usage/help information.
429 SC/01 Added the -bylocaldomain option to eximstats.
431 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
433 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
435 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
441 PH/01 Another patch from the Sieve maintainer.
443 PH/02 When an IPv6 address is converted to a string for single-key lookup
444 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
445 dots are used instead of colons so that keys in lsearch files need not
446 contain colons. This was done some time before quoting was made available
447 in lsearch files. However, iplsearch files do require colons in IPv6 keys
448 (notated using the quote facility) so as to distinguish them from IPv4
449 keys. This meant that lookups for IP addresses in host lists did not work
450 for iplsearch lookups.
452 This has been fixed by arranging for IPv6 addresses to be expressed with
453 colons if the lookup type is iplsearch. This is not incompatible, because
454 previously such lookups could never work.
456 The situation is now rather anomolous, since one *can* have colons in
457 ordinary lsearch keys. However, making the change in all cases is
458 incompatible and would probably break a number of configurations.
460 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
463 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
464 conversion specification without a maximum field width, thereby enabling
465 a rogue spamd server to cause a buffer overflow. While nobody in their
466 right mind would setup Exim to query an untrusted spamd server, an
467 attacker that gains access to a server running spamd could potentially
468 exploit this vulnerability to run arbitrary code as the Exim user.
470 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
471 $primary_hostname instead of what libspf2 thinks the hosts name is.
473 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
474 a directory entry by the name of the lookup key. Previously, if a
475 symlink pointed to a non-existing file or a file in a directory that
476 Exim lacked permissions to read, a lookup for a key matching that
477 symlink would fail. Now it is enough that a matching directory entry
478 exists, symlink or not. (Bugzilla 503.)
480 PH/03 The body_linecount and body_zerocount variables are now exported in the
483 PH/04 Added the $dnslist_matched variable.
485 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
486 This means they are set thereafter only if the connection becomes
489 PH/06 Added the client_condition to authenticators so that some can be skipped
490 by clients under certain conditions.
492 PH/07 The error message for a badly-placed control=no_multiline_responses left
493 "_responses" off the end of the name.
495 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
497 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
498 (without spaces) instead of just copying the configuration text.
500 PH/10 Added the /noupdate option to the ratelimit ACL condition.
502 PH/11 Added $max_received_linelength.
504 PH/12 Added +ignore_defer and +include_defer to host lists.
506 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
507 way in which PCRE > 7.0 is built.
509 PH/14 Implemented queue_only_load_latch.
511 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
512 MAIL command. The effect was to mangle the value on 64-bit systems.
514 PH/16 Another patch from the Sieve maintainer.
516 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
518 PH/18 If a system quota error occurred while trying to create the file for
519 a maildir delivery, the message "Mailbox is full" was not appended to the
520 bounce if the delivery eventually timed out. Change 4.67/27 below applied
521 only to a quota excession during the actual writing of the file.
523 PH/19 It seems that peer DN values may contain newlines (and other non-printing
524 characters?) which causes problems in log lines. The DN values are now
525 passed through string_printing() before being added to log lines.
527 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
528 and InterBase are left for another time.)
530 PH/21 Added message_body_newlines option.
532 PH/22 Guard against possible overflow in moan_check_errorcopy().
534 PH/23 POSIX allows open() to be a macro; guard against that.
536 PH/24 If the recipient of an error message contained an @ in the local part
537 (suitably quoted, of course), incorrect values were put in $domain and
538 $local_part during the evaluation of errors_copy.
544 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
545 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
548 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
549 issue a MAIL command.
551 PH/02 In an ACL statement such as
553 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
555 if a client was not listed at all, or was listed with a value other than
556 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
557 the condition was not true (as it should be), so access was not denied.
558 The bug was that the ! inversion was incorrectly passed on to the second
559 item. This has been fixed.
561 PH/03 Added additional dnslists conditions == and =& which are different from
562 = and & when the dns lookup returns more than one IP address.
564 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
565 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
567 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
568 FSYNC, which compiles an option called disable_fsync that allows for
569 bypassing fsync(). The documentation is heavily laced with warnings.
571 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
573 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
574 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
575 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
576 including adding "make clean"; (3) Added -fPIC when compiling the test
577 dynamically loaded module, to get rid of a warning.
579 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
580 message fails, move_frozen_messages = true and ignore_bounce_errors_after
581 = 0s. The bug is otherwise harmless.
583 PH/07 There was a bug in the dovecot authenticator such that the value of
584 $auth1 could be overwritten, and so not correctly preserved, after a
585 successful authentication. This usually meant that the value preserved by
586 the server_setid option was incorrect.
588 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
590 PH/09 Installed PCRE release 7.0.
592 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
593 run for batched SMTP input. It is now run at the start of every message
594 in the batch. While fixing this I discovered that the process information
595 (output by running exiwhat) was not always getting set for -bs and -bS
596 input. This is fixed, and it now also says "batched" for BSMTP.
598 PH/11 Added control=no_pipelining.
600 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
601 patch, slightly modified), and move the expansion of helo_data till after
602 the connection is made in the smtp transport (so it can use these
605 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
607 PH/14 Added log_selector = +pid.
609 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
611 PH/16 Add ${if forany and ${if forall.
613 PH/17 Added dsn_from option to vary the From: line in DSNs.
615 PH/18 Flush SMTP output before performing a callout, unless control =
616 no_callout_flush is set.
618 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
619 was true (the default) a successful delivery failed to delete the retry
620 item, thus causing premature timeout of the address. The bug is now
623 PH/20 Added hosts_avoid_pipelining to the smtp transport.
625 PH/21 Long custom messages for fakedefer and fakereject are now split up
626 into multiline reponses in the same way that messages for "deny" and
627 other ACL rejections are.
629 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
630 with slight modification.
632 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
633 draft, changing the syntax and factoring some duplicate code".
635 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
636 for deliveries of the second and subsequent messages over the same SMTP
639 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
640 ${reduce, with only minor "tidies".
642 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
644 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
645 expansion side effects.
647 PH/27 When a message times out after an over-quota error from an Exim-imposed
648 quota, the bounce message says "mailbox is full". This message was not
649 being given when it was a system quota that was exceeded. It now should
652 MH/03 Made $recipients available in local_scan(). local_scan() already has
653 better access to the recipient list through recipients_list[], but
654 $recipients can be useful in postmaster-provided expansion strings.
656 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
657 in the case of a MAIL command with additional options following the
658 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
659 were accidentally chopped off.
661 PH/29 SMTP synchronization checks are implemented when a command is read -
662 there is a check that no more input is waiting when there shouldn't be
663 any. However, for some commands, a delay in an ACL can mean that it is
664 some time before the response is written. In this time, more input might
665 arrive, invalidly. So now there are extra checks after an ACL has run for
666 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
667 pipelining has not been advertised.
669 PH/30 MH's patch to allow iscntrl() characters to be list separators.
671 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
672 returned in the SMTP response when smtp_return_error_details was false.
675 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
676 instead of the C I/O that was originally supplied, because problems were
679 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
680 Exim which did not show up earlier: it was assuming that a call to
681 SSL_CTX_set_info_callback() might give an error value. In fact, there is
682 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
683 was a macro that became an assignment, so it seemed to work. This has
684 changed to a proper function call with a void return, hence the compile
685 error. Exim's code has been fixed.
687 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
690 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
692 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
694 PH/37 If a message is not accepted after it has had an id assigned (e.g.
695 because it turns out to be too big or there is a timeout) there is no
696 "Completed" line in the log. When some messages of this type were
697 selected by exigrep, they were listed as "not completed". Others were
698 picked up by some special patterns. I have improved the selection
699 criteria to be more general.
701 PH/38 The host_find_failed option in the manualroute router can now be set
702 to "ignore", to completely ignore a host whose IP address cannot be
703 found. If all hosts are ignored, the behaviour is controlled by the new
704 host_all_ignored option.
706 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
707 homing or because of multiple MX records with /mx) generated more than
708 one IP address, and the following item turned out to be the local host,
709 all the secondary addresses of the first item were incorrectly removed
710 from the list, along with the local host and any following hosts (which
711 is what is supposed to happen).
713 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
714 whoever called Exim into the -H file. In the case of the daemon it was
715 behaving confusingly. When first started, it used values for whoever
716 started the daemon, but after a SIGHUP it used the Exim user (because it
717 calls itself on a restart). I have changed the code so that it now always
720 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
721 message are rejected with the same error (e.g. no authentication or bad
722 sender address), and a DATA command is nevertheless sent (as can happen
723 with PIPELINING or a stupid MUA), the error message that was given to the
724 RCPT commands is included in the rejection of the DATA command. This is
725 intended to be helpful for MUAs that show only the final error to their
728 PH/42 Another patch from the Sieve maintainer.
730 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
731 Eximstats - Fixed some broken HTML links and added missing column headers
733 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
734 columns for Rejects, Temp Rejects, Ham, and Spam rows.
736 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
738 PH/43 Yet another patch from the Sieve maintainer.
740 PH/44 I found a way to check for a TCP/IP connection going away before sending
741 the response to the final '.' that terminates a message, but only in the
742 case where the client has not sent further data following the '.'
743 (unfortunately, this is allowed). However, in many cases there won't be
744 any further data because there won't be any more messages to send. A call
745 to select() can be used: if it shows that the input is "ready", there is
746 either input waiting, or the socket has been closed. An attempt to read
747 the next input character can distinguish the two cases. Previously, Exim
748 would have sent an OK response which the client would never have see.
749 This could lead to message repetition. This fix should cure that, at
750 least in a lot of common cases.
752 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
753 advertised in response to EHLO.
759 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
760 fixed by 4.65/MH/01 (is this a record?) are fixed:
762 (i) An empty string was always treated as zero by the numeric comparison
763 operators. This behaviour has been restored.
765 (ii) It is documented that the numeric comparison operators always treat
766 their arguments as decimal numbers. This was broken in that numbers
767 starting with 0 were being interpreted as octal.
769 While fixing these problems I realized that there was another issue that
770 hadn't been noticed. Values of message_size_limit (both the global option
771 and the transport option) were treated as octal if they started with 0.
772 The documentation was vague. These values are now always treated as
773 decimal, and I will make that clear in the documentation.
779 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
780 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
783 MH/01 Don't check that the operands of numeric comparison operators are
784 integers when their expansion is in "skipping" mode (fixes bug
785 introduced by 4.64-PH/07).
787 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
788 child addresses, Exim now panics and dies. Previously, because the count
789 is held in a short int, deliveries were likely to be lost. As such a
790 large number of recipients for a single message is ridiculous
791 (performance will be very, very poor), I have chosen to impose a limit
792 rather than extend the field.
798 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
799 leftover -K file (the existence of which was triggered by #402).
800 While we were at it, introduced process PID as part of the -K
801 filename. This should rule out race conditions when creating
804 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
805 processing considerably. Previous code took too long for large mails,
806 triggering a timeout which in turn triggers #401.
808 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
809 in the DK code in transports.c. sendfile() is not really portable,
810 hence the _LINUX specificness.
812 TF/01 In the add_headers option to the mail command in an Exim filter,
813 there was a bug that Exim would claim a syntax error in any
814 header after the first one which had an odd number of characters
817 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
818 callout verification, Exim cached a "reject" for the entire domain. This
819 is correct for most verifications, but it is not correct for a recipient
820 verification with use_sender or use_postmaster set, because in that case
821 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
822 case of MAIL FROM:<> rejection from other early rejections (e.g.
823 rejection of HELO). When verifying a recipient using a non-null MAIL
824 address, the cache is ignored if it shows MAIL FROM:<> rejection.
825 Whatever the result of the callout, the value of the domain cache is
826 left unchanged (for any other kind of callout, getting as far as trying
827 RCPT means that the domain itself is ok).
829 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
832 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
833 manifest itself as EPIPE rather than ECONNECT. When tidying away a
834 session, the daemon ignores ECONNECT errors and logs others; it now
835 ignores EPIPE as well.
837 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
838 (quoted-printable decoding).
840 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
841 later the small subsequent patch to fix an introduced bug.
843 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
845 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
847 PH/08 An error is now given if message_size_limit is specified negative.
849 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
850 to be given (somewhat) arbitrary names.
852 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
855 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
856 miscellaneous code fixes
858 PH/10 Added the log_reject_target ACL modifier to specify where to log
861 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
862 hostname. This is wrong, because it relates to the incoming message (and
863 probably the interface on which it is arriving) and not to the outgoing
864 callout (which could be using a different interface). This has been
865 changed to use the value of the helo_data option from the smtp transport
866 instead - this is what is used when a message is actually being sent. If
867 there is no remote transport (possible with a router that sets up host
868 addresses), $smtp_active_hostname is used.
870 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
871 tweaks were necessary in order to get it to work (see also 21 below):
872 (a) The code assumed that strncpy() returns a negative number on buffer
873 overflow, which isn't the case. Replaced with Exim's string_format()
875 (b) There were several signed/unsigned issues. I just did the minimum
876 hacking in of casts. There is scope for a larger refactoring.
877 (c) The code used strcasecmp() which is not a standard C function.
878 Replaced with Exim's strcmpic() function.
879 (d) The code set only $1; it now sets $auth1 as well.
880 (e) A simple test gave the error "authentication client didn't specify
881 service in request". It would seem that Dovecot has changed its
882 interface. Fortunately there's a specification; I followed it and
883 changed what the client sends and it appears to be working now.
885 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
888 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
889 address A is aliased to B and C, where B exists and C does not. Without
890 -v the output is "A verified" because verification stops after a
891 successful redirection if more than one address is generated. However,
892 with -v the child addresses are also verified. Exim was outputting "A
893 failed to verify" and then showing the successful verification for C,
894 with its parentage. It now outputs "B failed to verify", showing B's
895 parentage before showing the successful verification of C.
897 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
898 look up a TXT record in a specific list after matching in a combined
901 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
902 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
903 they consult the DNS. I had assumed they would set it the way they
904 wanted; and indeed my experiments on Linux seem to show that in some
905 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
906 To be on the safe side, however, I have now made the interface to
907 host_find_byname() similar to host_find_bydns(), with an argument
908 containing the DNS resolver options. The host_find_byname() function now
909 sets these options at its start, just as host_find_bydns() does. The smtp
910 transport options dns_qualify_single and dns_search_parents are passed to
911 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
912 of host_find_byname() use the default settings of RES_DEFNAMES
913 (qualify_single) but not RES_DNSRCH (search_parents).
915 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
916 spool_read_header() do less string testing, by means of a preliminary
917 switch on the second character of optional "-foo" lines. (This is
918 overdue, caused by the large number of possibilities that now exist.
919 Originally there were few.) While I was there, I also converted the
920 str(n)cmp tests so they don't re-test the leading "-" and the first
921 character, in the hope this might squeeze out yet more improvement.
923 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
924 flag allowing group syntax was set by the header_syntax check but not
925 turned off, possible causing trouble later; (2) The flag was not being
926 set at all for the header_verify test, causing "group"-style headers to
927 be rejected. I have now set it in this case, and also caused header_
928 verify to ignore an empty address taken from a group. While doing this, I
929 came across some other cases where the code for allowing group syntax
930 while scanning a header line wasn't quite right (mostly, not resetting
931 the flag correctly in the right place). These bugs could have caused
932 trouble for malformed header lines. I hope it is now all correct.
934 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
935 with the "reply" argument non-NULL. The code, however (which originally
936 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
937 but it didn't always do it. This confused somebody who was copying the
938 code for some other use. I have removed all the tests.
940 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
941 feature that was used to support insecure browsers during the U.S. crypto
942 embargo. It requires special client support, and Exim is probably the
943 only MTA that supported it -- and would never use it because real RSA is
944 always available. This code has been removed, because it had the bad
945 effect of slowing Exim down by computing (never used) parameters for the
946 RSA_EXPORT functionality.
948 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
949 authenticator to fail if there's a tab character in the incoming data
950 (there should never be unless someone is messing about, as it's supposed
951 to be base64-encoded). Also added, on Timo's advice, the "secured" option
952 if the connection is using TLS or if the remote IP is the same as the
953 local IP, and the "valid-client-cert option" if a client certificate has
956 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
957 authenticators. This can be used for authorization after authentication
958 succeeds. (In the case of plaintext, it servers for both authentication
961 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
962 if any retry times were supplied.
964 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
965 connection, where there is no client IP address to check. In this
966 situation, the verify now always succeeds.
968 PH/25 Applied John Jetmore's -Mset patch.
970 PH/26 Added -bem to be like -Mset, but loading a message from a file.
972 PH/27 In a string expansion for a processed (not raw) header when multiple
973 headers of the same name were present, leading whitespace was being
974 removed from all of them, but trailing whitespace was being removed only
975 from the last one. Now trailing whitespace is removed from each header
976 before concatenation. Completely empty headers in a concatenation (as
979 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
980 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
982 PH/29 [Removed. This was a change that I later backed out, and forgot to
983 correct the ChangeLog entry (that I had efficiently created) before
984 committing the later change.]
986 PH/30 Exim was sometimes attempting to deliver messages that had suffered
987 address errors (4xx response to RCPT) over the same connection as other
988 messages routed to the same hosts. Such deliveries are always "forced",
989 so retry times are not inspected. This resulted in far too many retries
990 for the affected addresses. The effect occurred only when there were more
991 hosts than the hosts_max_try setting in the smtp transport when it had
992 the 4xx errors. Those hosts that it had tried were not added to the list
993 of hosts for which the message was waiting, so if all were tried, there
994 was no problem. Two fixes have been applied:
996 (i) If there are any address or message errors in an SMTP delivery, none
997 of the hosts (tried or untried) are now added to the list of hosts
998 for which the message is waiting, so the message should not be a
999 candidate for sending over the same connection that was used for a
1000 successful delivery of some other message. This seems entirely
1001 reasonable: after all the message is NOT "waiting for some host".
1002 This is so "obvious" that I'm not sure why it wasn't done
1003 previously. Hope I haven't missed anything, but it can't do any
1004 harm, as the worst effect is to miss an optimization.
1006 (ii) If, despite (i), such a delivery is accidentally attempted, the
1007 routing retry time is respected, so at least it doesn't keep
1008 hammering the server.
1010 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1011 in ${readsocket because some servers need this prod.
1013 PH/32 Added some extra debug output when updating a wait-xxx database.
1015 PH/33 The hint "could be header name not terminated by colon", which has been
1016 given for certain expansion errors for a long time, was not being given
1017 for the ${if def:h_colon_omitted{... case.
1019 PH/34 The spec says: "With one important exception, whenever a domain list is
1020 being scanned, $domain contains the subject domain." There was at least
1021 one case where this was not true.
1023 PH/35 The error "getsockname() failed: connection reset by peer" was being
1024 written to the panic log as well as the main log, but it isn't really
1025 panic-worthy as it just means the connection died rather early on. I have
1026 removed the panic log writing for the ECONNRESET error when getsockname()
1029 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1030 runs only) independently of the message's sender address. This meant
1031 that, if the 4xx error was in fact related to the sender, a different
1032 message to the same recipient with a different sender could confuse
1033 things. In particualar, this can happen when sending to a greylisting
1034 server, but other circumstances could also provoke similar problems.
1035 I have changed the default so that the retry time for these errors is now
1036 based a combination of the sender and recipient addresses. This change
1037 can be overridden by setting address_retry_include_sender=false in the
1040 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1041 remote server are returned as part of bounce messages. This was not
1042 happening for LMTP over a pipe (the lmtp transport), but now it is the
1043 same for both kinds of LMTP.
1045 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1046 in header lines that were in fact CNAMEs. This is no longer the case.
1048 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1049 and queue runs started by the daemon processed all messages. This has
1050 been fixed so that -R and -S can now usefully be given with -q<time>.
1052 PH/40 Import PCRE release 6.7 (fixes some bugs).
1054 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1056 PH/42 Give an error if -q is specified more than once.
1058 PH/43 Renamed the variables $interface_address and $interface_port as
1059 $received_ip_address and $received_port, to make it clear that these
1060 values apply to message reception, and not to the outgoing interface when
1061 a message is delivered. (The old names remain recognized, of course.)
1063 PH/44 There was no timeout on the connect() call when using a Unix domain
1064 socket in the ${readsocket expansion. There now is.
1066 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1067 be meaningful with "accept".
1069 SC/01 Eximstats V1.43
1070 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1072 SC/02 Eximstats V1.44
1073 Use a glob alias rather than an array ref in the generated
1074 parser. This improves both readability and performance.
1076 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1077 Collect SpamAssassin and rejection statistics.
1078 Don't display local sender or destination tables unless
1079 there is data to show.
1080 Added average volumes into the top table text output.
1082 SC/04 Eximstats V1.46
1083 Collect data on the number of addresses (recipients)
1084 as well as the number of messages.
1086 SC/05 Eximstats V1.47
1087 Added 'Message too big' to the list of mail rejection
1088 reasons (thanks to Marco Gaiarin).
1090 SC/06 Eximstats V1.48
1091 Mainlog lines which have GMT offsets and are too short to
1092 have a flag are now skipped.
1094 SC/07 Eximstats V1.49 (Alain Williams)
1095 Added the -emptyok flag.
1097 SC/08 Eximstats V1.50
1098 Fixes for obtaining the IP address from reject messages.
1100 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1101 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1102 whitesspace changes from 4.64-PH/27
1104 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1107 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1108 are found, allow negative numbers in numeric criteria)
1110 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1112 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1113 to match changes made in 4.64-PH/43
1115 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1117 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1118 "domain" is actually an IP address, and force a failure. This locks out
1119 those revolvers/nameservers that support "A-for-A" lookups, in
1120 contravention of the specifications.
1122 PH/48 When a host name was looked up from an IP address, and the subsequent
1123 forward lookup of the name timed out, the host name was left in
1124 $sender_host_name, contrary to the specification.
1126 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1127 restricted to single-key lookups, Exim was not diagnosing an error if
1128 * or *@ was used with a query-style lookup.
1130 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1132 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1133 long ago, but noone interested enough thought of it. Let's just say that
1134 the "1.1" means that there are some new functions that weren't there at
1135 some point in the past.
1137 PH/51 Error processing for expansion failure of helo_data from an smtp
1138 transport during callout processing was broken.
1140 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1141 tested/used via the -bh/-bhc/-bs options.
1143 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1144 bug, fixed in subsequent PCRE releases).
1146 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1147 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1149 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1155 SC/01 Use a glob alias rather than an array ref in eximstats generated
1156 parser. This improves both readability and performance.
1158 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1159 Don't display local sender or destination tables in eximstats unless
1160 there is data to show.
1161 Added average volumes into the eximstats top table text output.
1163 SC/03 Collect data on the number of addresses (recipients) as well
1164 as the number of messages in eximstats.
1166 TF/01 Correct an error in the documentation for the redirect router. Exim
1167 does (usually) call initgroups() when daemonizing.
1169 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1170 with consistent privilege compared to when running as a daemon.
1172 TF/03 Note in the spec that $authenticated_id is not set for local
1173 submissions from trusted users.
1175 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1176 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1178 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1179 by adding some example configuration directives to the default
1180 configuration file. A little bit of work is required to uncomment the
1181 directives and define how usernames and passwords are checked, but
1182 there is now a framework to start from.
1184 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1185 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1186 without this. I don't know how relevant this is to other LDAP libraries.
1188 PH/02 Add the verb name to the "unknown ACL verb" error.
1190 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1192 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1194 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1195 directory not expanded when it should be if an expanded home directory
1196 was set for the address (which is overridden by the transport).
1198 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1201 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1202 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1203 because it is too late at that time, and has no effect.
1205 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1206 security issue with \' (bugzilla #107). I could not use the
1207 PQescapeStringConn() function, because it needs a PGconn value as one of
1210 PH/08 When testing addresses using -bt, indicate those final addresses that
1211 are duplicates that would not cause an additional delivery. At least one
1212 person was confused, thinking that -bt output corresponded to deliveries.
1213 (Suppressing duplicates isn't a good idea as you lose the information
1214 about possibly different redirections that led to the duplicates.)
1216 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1217 systems where poll() doesn't work, in particular OS X.
1219 PH/10 Added more information to debugging output for retry time not reached.
1221 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1222 operations in malware.c.
1224 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1227 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1228 syslog_duplication set false, log lines that would normally be written
1229 both the the main log and to the reject log were not written to syslog at
1232 PH/14 In the default configuration, change the use of "message" in ACL warn
1233 statements to "add_header".
1235 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1236 not followed by a command (e.g. "seen endif").
1238 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1239 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1242 PH/17 Added extra conditions to the default value of delay_warning_condition
1246 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1247 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1248 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1251 The Auto-Submitted: and various List- headers are standardised, whereas I
1252 don't think Precedence: ever was.
1254 PH/18 Refactored debugging code in route_finduser() to show more information,
1255 in particular, the error code if getpwnam() issues one.
1257 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1258 This is apparently needed in addition to the PH/07 change above to avoid
1259 any possible encoding problems.
1261 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1262 but not after initializing Perl.
1264 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1265 output them only if debugging. By default they are written stderr,
1266 apparently, which is not desirable.
1268 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1271 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1274 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1276 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1277 authenticated or an ident call has been made. Suppress the default
1278 values for $authenticated_id and $authenticated_sender (but permit -oMai
1279 and -oMas) when testing with -bh.
1281 PH/24 Re-jigged the order of the tests in the default configuration so that the
1282 tests for valid domains and recipients precede the DNS black list and CSA
1283 tests, on the grounds that those ones are more expensive.
1285 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1286 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1287 This bug exists in every version of Exim that I still have, right back to
1290 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1291 However, an attempt to turn on case-sensitivity in a regex key by
1292 including (?-i) didn't work because the subject string was already
1293 lowercased, and the effects were non-intuitive. It turns out that a
1294 one-line patch can be used to allow (?-i) to work as expected.
1300 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1301 other effects) broke the use of negated acl sub-conditions.
1303 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1306 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1307 "Deny" causes Exim to reject the incoming connection with a 554 error.
1308 Unfortunately, if there is a major crisis, such as a disk failure,
1309 tcp-wrappers gives "deny", whereas what one would like would be some
1310 kind of temporary error. A kludge has been added to help with this.
1311 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1312 554 error is used if errno is still zero or contains ENOENT (which occurs
1313 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1316 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1318 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1319 errors. Otherwise a message that provokes a temporary error (when other
1320 messages do not) can cause a whole host to time out.
1322 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1323 addresses were routed directly to files or pipes from a redirect router.
1324 File deliveries just didn't batch; pipe deliveries might have suffered
1327 PH/06 A failure to get a lock for a hints database would erroneously always say
1328 "Failed to get write lock", even when it was really a read lock.
1330 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1331 of 0600. This has been changed to use the value of the lockfile_mode
1332 option (which defaults to 0600).
1334 PH/08 Applied small patch from the Sieve maintainer.
1336 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1337 folder from quota calculations, a direct delivery into this folder messed
1338 up the contents of the maildirsize file. This was because the regex was
1339 used only to exclude .Trash (or whatever) when the size of the mailbox
1340 was calculated. There was no check that a delivery was happening into an
1341 excluded directory. This bug has been fixed by ignoring all quota
1342 processing for deliveries into excluded directories.
1344 PH/10 Added the maildirfolder_create_regex option to appendfile.
1350 PH/01 The code for finding all the local interface addresses on a FreeBSD
1351 system running IPv6 was broken. This may well have applied to all BSD
1352 systems, as well as to others that have similar system calls. The broken
1353 code found IPv4 interfaces correctly, but gave incorrect values for the
1354 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1355 that it would not match correctly against @[] and not recognize the IPv6
1358 PH/02 The ipliteral router was not recognizing addresses of the form user@
1359 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1361 PH/03 Added disable_ipv6.
1363 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1364 decoded form, because it is most often used to construct To: headers
1365 lines in autoreplies, and the decoded form may well be syntactically
1366 invalid. However, $reply_address has leading white space removed, and all
1367 newlines turned into spaces so that the autoreply transport does not
1370 PH/05 If group was specified without a user on a router, and no group or user
1371 was specified on a transport, the group from the router was ignored.
1373 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1374 for visible compile-time settings that can be used to change these
1375 numbers, for those that want even more. Backwards compatibility with old
1376 spool files has been maintained. However, going back to a previous Exim
1377 release will lost any variables that are in spool files.
1379 PH/07 Two small changes when running in the test harness: increase delay when
1380 passing a TCP/IP connection to a new process, in case the original
1381 process has to generate a bounce, and remove special handling of
1382 127.0.0.2 (sic), which is no longer necessary.
1384 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1385 be the same on different OS.
1387 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1390 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1391 whether --show-vars was specified or not
1393 JJ/02 exipick: Added support for new ACL variable spool format introduced
1396 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1397 syntactically invalid From: or Reply-to: line, and a filter used this to
1398 generate an autoreply, and therefore failed to obtain an address for the
1399 autoreply, Exim could try to deliver to a non-existent relative file
1400 name, causing unrelated and misleading errors. What now happens is that
1401 it logs this as a hard delivery error, but does not attempt to create a
1404 PH/11 The exinext utility has a -C option for testing purposes, but although
1405 the given file was scanned by exinext itself; it wasn't being passed on
1406 when Exim was called.
1408 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1409 an end-of-file indication when reading a command response.
1411 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1412 compiled. In many other places in Exim, IPv6 addresses are always
1413 recognized, so I have changed this. It also means that IPv4 domain
1414 literals of the form [IPV4:n.n.n.n] are now always recognized.
1416 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1417 used if the router is not running as root, for example, when verifying at
1418 ACL time, or when using -bh. The debugging output from this situation was
1419 non-existent - all you got was a failure to exec. I have made two
1422 (a) Failures to set uid/gid, the current directory, or a process leader
1423 in a subprocess such as that created by queryprogram now generate
1424 suitable debugging ouput when -d is set.
1426 (b) The queryprogram router detects when it is not running as root,
1427 outputs suitable debugging information if -d is set, and then runs
1428 the subprocess without attempting to change uid/gid.
1430 PH/15 Minor change to Makefile for building test_host (undocumented testing
1433 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1434 additional section of a DNS packet that returns MX or SRV records.
1435 Instead, it always explicitly searches for A/AAAA records. This avoids
1436 major problems that occur when a DNS server includes only records of one
1437 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1438 fixed another bug: if SRV records were looked up and the corresponding
1439 address records were *not* found in the additional section, the port
1440 values from the SRV records were lost.
1442 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1443 using the correct key (the original address) when searching the retry
1444 rules in order to find which one to use for generating the retry hint.
1446 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1447 adding the default one. Similarly, if it contains a Reply-To: header, the
1448 errors_reply_to option, if set, is not used.
1450 PH/19 When calculating a retry time, Exim used to measure the "time since
1451 failure" by looking at the "first failed" field in the retry record. Now
1452 it does not use this if it is later than than the arrival time of the
1453 message. Instead it uses the arrival time. This makes for better
1454 behaviour in cases where some deliveries succeed, thus re-setting the
1455 "first failed" field. An example is a quota failure for a huge message
1456 when small messages continue to be delivered. Without this change, the
1457 "time since failure" will always be short, possible causing more frequent
1458 delivery attempts for the huge message than are intended.
1459 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1461 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1462 $1, $2, $3) because the numerical variables can be reset during some
1463 expansion items (e.g. "match"), thereby losing the authentication data.
1465 PH/21 Make -bV show the size of off_t variables so that the test suite can
1466 decide whether to run tests for quotas > 2G.
1468 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1469 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1470 filecount value is greater than 2G or if a quota value is greater than 2G
1471 on a system where the size of off_t is not greater than 4, a panic error
1474 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1475 never match. The debug and -bh output now contains an explicit error
1476 message indicating a malformed IPv4 address or mask.
1478 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1479 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1480 PH/23 above applies.
1482 PH/25 Do not write to syslog when running in the test harness. The only
1483 occasion when this arises is a failure to open the main or panic logs
1484 (for which there is an explicit test).
1486 PH/26 Added the /no_tell option to "control=freeze".
1488 PH/27 If a host name lookup failed very early in a connection, for example, if
1489 the IP address matched host_lookup and the reverse lookup yielded a name
1490 that did not have a forward lookup, an error message of the form "no IP
1491 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1492 could be logged. Now it outputs the IP address instead of "NULL".
1494 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1495 allows the sender and the authenticated sender to be set when
1496 submitting a message from within Exim. Since child_open_exim() is
1497 documented for local_scan(), the new function should be too.
1499 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1500 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1501 results in an empty string is now treated as unset.
1503 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1505 PH/31 Added sender_verify_fail logging option.
1507 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1508 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1509 tidied the source and removed it altogether.
1511 PH/33 When a queue run was abandoned because the load average was too high, a
1512 log line was always written; now it is written only if the queue_run log
1513 selector is set. In addition, the log line for abandonment now contains
1514 information about the queue run such as the pid. This is always present
1515 in "start" and "stop" lines but was omitted from the "abandon" line.
1517 PH/34 Omit spaces between a header name and the colon in the error message that
1518 is given when verify = headers_syntax fails (if there are lots of them,
1519 the message gets confusing).
1521 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1522 names, as there are now some PTR records that contain slashes. This check
1523 is only to protect against broken name servers that fall over on strange
1524 characters, so the fact that it applies to all lookups doesn't matter.
1526 PH/36 Now that the new test suite is complete, we can remove some of the
1527 special code in Exim that was needed for the old test suite. For example,
1528 sorting DNS records because real resolvers return them in an arbitrary
1529 order. The new test suite's fake resolver always returns records in the
1532 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1533 bounces) except when queue_only is set, to avoid logging races between
1534 the different processes.
1536 PH/38 Panic-die if .include specifies a non-absolute path.
1538 PH/39 A tweak to the "H" retry rule from its user.
1540 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1541 a label. They prevented compilation on older perls.
1543 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1544 a warning to be raised on newish perls.
1546 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1547 on queue. Changes to match documented behaviour of showing count of
1548 messages matching specified criteria.
1550 PH/40 Changed the default ident timeout from 30s to 5s.
1552 PH/41 Added support for the use of login_cap features, on those BSD systems
1553 that have them, for controlling the resources used by pipe deliveries.
1555 PH/42 The content-scanning code uses fopen() to create files in which to put
1556 message data. Previously it was not paying any attention to the mode of
1557 the files. Exim runs with umask(0) because the rest of the code creates
1558 files with open(), and sets the required mode explicitly. Thus, these
1559 files were ending up world-writeable. This was not a big issue, because,
1560 being within the spool directory, they were not world-accessible. I have
1561 created a function called modefopen, which takes an additional mode
1562 argument. It sets umask(777), creates the file, chmods it to the required
1563 mode, then resets the umask. All the relevant calls to fopen() in the
1564 content scanning code have been changed to use this function.
1566 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1567 to 24 hours. This avoids potential overflow problems when processing G
1568 and H retry rules. I suspect nobody ever tinkers with this value.
1570 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1572 PH/45 When the plaintext authenticator is running as a client, the server's
1573 challenges are checked to ensure they are valid base64 strings. By
1574 default, the authentication attempt is cancelled if an invalid string is
1575 received. Setting client_ignore_invalid_base64 true ignores these errors.
1576 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1577 they are received. Thus, the responses can be made to depend on the
1578 challenges. If an invalid string is ignored, an empty string is placed in
1581 PH/46 Messages that are created by the autoreply transport now contains a
1582 References: header, in accordance with RFCs 2822 and 3834.
1584 PH/47 Added authenticated_sender_force to the smtp transport.
1586 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1588 PH/49 Installed latest patch from the Sieve maintainer.
1590 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1591 was also set, the appendfile transport was unnecessarily scanning a
1592 directory of message files (e.g. for maildir delivery) to find the count
1593 of files (along with the size), even though it did not need this
1594 information. It now does the scan only if it needs to find either the
1595 size of the count of files.
1597 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1599 PH/52 Two bugs concerned with error handling when the smtp transport is
1602 (i) Exim was not creating retry information for temporary errors given
1603 for individual recipients after the DATA command when the smtp transport
1604 was used in LMTP mode. This meant that they could be retried too
1605 frequently, and not timed out correctly.
1607 (ii) Exim was setting the flag that allows error details to be returned
1608 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1609 recipients that were returned after the DATA command.
1611 PH/53 This is related to PH/52, but is more general: for any failing address,
1612 when detailed error information was permitted to be returned to the
1613 sender, but the error was temporary, then after the final timeout, only
1614 "retry timeout exceeded" was returned. Now it returns the full error as
1615 well as "retry timeout exceeded".
1617 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1618 do this, and (what is worse) MTAs that accept it.
1620 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1621 will now be deprecated.
1623 PH/56 New os.c-cygwin from the Cygwin maintainer.
1625 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1626 formats (previously only available in exim formats via -bpr, -bpru,
1627 and -bpra. Now also available in native and exiqgrep formats)
1629 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1630 with very large, slow to parse queues
1632 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1634 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1636 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1637 responses to authentication challenges, though it was showing the
1638 challenges; (ii) I've removed the CR characters from the debug output for
1641 PH/58 Allow for the insertion of a newline as well as a space when a string
1642 is turned into more than one encoded-word during RFC 2047 encoding. The
1643 Sieve code now uses this.
1645 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1646 data_4xx, lost_connection, tls_required.
1648 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1649 message was being sent as an SMTP response.
1651 PH/61 Add -l and -k options to exicyclog.
1653 PH/62 When verifying, if an address was redirected to one new address, so that
1654 verification continued, and the new address failed or deferred after
1655 having set something in $address_data, the value of $address_data was not
1656 passed back to the ACL. This was different to the case when no
1657 redirection occurred. The value is now passed back in both cases.
1659 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1660 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1661 use login_cap.h, so on its own it isn't the distinguishing feature. The
1662 new name refers directly to the setclassresources() function.
1664 PH/65 Added configuration files for NetBSD3.
1666 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1668 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1669 is preferred over IPv4.
1671 PH/68 The bounce_return_message and bounce_return_body options were not being
1672 honoured for bounces generated during the reception of non-SMTP messages.
1673 In particular, this applied to messages rejected by the ACL. This bug has
1674 been fixed. However, if bounce_return_message is true and bounce_return_
1675 body is false, the headers that are returned for a non-SMTP message
1676 include only those that have been read before the error was detected.
1677 (In the case of an ACL rejection, they have all been read.)
1679 PH/69 The HTML version of the specification is now built in a directory called
1680 spec_html instead of spec.html, because the latter looks like a path with
1681 a MIME-type, and this confuses some software.
1683 PH/70 Catch two compiler warnings in sieve.c.
1685 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1686 function verify_get_ident() calls ip_connect() to connect a socket, but
1687 if the "connect()" function timed out, ip_connect() used to close the
1688 socket. However, verify_get_ident() also closes the socket later, and in
1689 between Exim writes to the log, which may get opened at this point. When
1690 the socket was closed in ip_connect(), the log could get the same file
1691 descriptor number as the socket. This naturally causes chaos. The fix is
1692 not to close the socket in ip_connect(); the socket should be closed by
1693 the function that creates it. There was only one place in the code where
1694 this was missing, in the iplookup router, which I don't think anybody now
1695 uses, but I've fixed it anyway.
1697 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1698 well as to direct DNS lookups. Otherwise the handling of names in host
1699 lists is inconsistent and therefore confusing.
1705 PH/01 Two changes to the default runtime configuration:
1707 (1) Move the checks for relay_from_hosts and authenticated clients from
1708 after to before the (commented out) DNS black list checks.
1710 (2) Add control=submission to the relay_from_hosts and authenticated
1711 clients checks, on the grounds that messages accepted by these
1712 statements are most likely to be submissions.
1714 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1716 (1) Generate an error if the third argument for the ${prvs expansion is
1719 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1722 (3) Reset the variables that are obtained from the first argument of
1723 ${prvscheck and used in the second argument before leaving the code,
1724 because their memory is reclaimed, so using them afterwards may do
1727 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1728 one (it's much easier than Tom thought :-).
1730 (5) Because of (4), we can now allow for the use of $prvscheck_result
1731 inside the third argument.
1733 PH/03 For some reason, the default setting of PATH when running a command from
1734 a pipe transport was just "/usr/bin". I have changed it to
1737 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1738 anything to be listed in the output from -bV.
1740 PH/05 When a filter generated an autoreply, the entire To: header line was
1741 quoted in the delivery log line, like this:
1743 => >A.N.Other <ano@some.domain> <original@ddress> ...
1745 This has been changed so that it extracts the operative address. There
1746 may be more than one such address. If so, they are comma-separated, like
1749 => >ano@some.domain,ona@other.domain <original@ddress> ...
1751 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1752 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1753 not being looked up in the rDNS to get a host name, Exim was showing the
1754 IP address twice in Received: lines, even though the IP addresses were
1755 identical. For example:
1757 Received: from [1.2.3.4] (helo=[1.2.3.4])
1759 However, if the real host name was known, it was omitting the HELO data
1760 if it matched the actual IP address. This has been tidied up so that it
1761 doesn't show the same IP address twice.
1763 PH/07 When both +timestamp and +memory debugging was on, the value given by
1764 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1765 called by the debug printing, thereby overwriting the timestamp buffer.
1766 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1768 PH/08 When the original message was included in an autoreply transport, it
1769 always said "this is a copy of the message, including all the headers",
1770 even if body_only or headers_only was set. It now gives an appropriate
1773 PH/09 Applied a patch from the Sieve maintainer which:
1775 o fixes some comments
1776 o adds the (disabled) notify extension core
1777 o adds some debug output for the result of if/elsif tests
1778 o points to the current vacation draft in the documentation
1779 and documents the missing references header update
1783 o fixes a bug in processing the envelope test (when testing
1784 multiple envelope elements, the last element determinted the
1787 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1788 Electronic Mail") by including:
1790 Auto-submitted: auto-generated
1792 in the messages that it generates (bounce messages and others, such as
1793 warnings). In the case of bounce messages for non-SMTP mesages, there was
1794 also a typo: it was using "Auto_submitted" (underscore instead of
1795 hyphen). Since every message generated by Exim is necessarily in response
1796 to another message, thes have all been changed to:
1798 Auto-Submitted: auto-replied
1800 in accordance with these statements in the RFC:
1802 The auto-replied keyword:
1804 - SHOULD be used on messages sent in direct response to another
1805 message by an automatic process,
1807 - MUST NOT be used on manually-generated messages,
1809 - MAY be used on Delivery Status Notifications (DSNs) and Message
1810 Disposition Notifications (MDNs),
1812 - MUST NOT be used on messages generated by automatic or periodic
1813 processes, except for messages which are automatic responses to
1816 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1817 to the default Received: header definition.
1819 PH/12 Added log selector acl_warn_skipped (default on).
1821 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1822 variables because (a) they are in the wrong storage pool and (b) even if
1823 they were copied, it wouldn't work properly because of the caching.
1825 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1826 checking when decoding. Apparently there are clients that generate
1827 overlong encoded strings. Why am I not surprised?
1829 PH/15 If the first argument of "${if match_address" was not empty, but did not
1830 contain an "@" character, Exim crashed. Now it writes a panic log message
1831 and treats the condition as false.
1833 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1835 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1836 extension "envelope-auth". The code is finished and in agreement with
1837 other implementations, but there is no documentation so far and in fact,
1838 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1839 not changing the active code.
1841 Print executed "if" and "elsif" statements when debugging is used. This
1842 helps a great deal to understand what a filter does.
1844 Document more things not specified clearly in RFC3028. I had all this
1845 sorted out, when out of a sudden new issues came to my mind. Oops."
1847 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1850 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1851 canonical form (as documented). However, after a host name lookup from
1852 the IP address, check_host() was doing a simple string comparison with
1853 addresses acquired from the DNS when checking that the found name did
1854 have the original IP as one of its addresses. Since any found IPv6
1855 addresses are likely to be in abbreviated form, the comparison could
1856 fail. Luckily, there already exists a function for doing the comparison
1857 by converting both addresses to binary, so now that is used instead of
1858 the text comparison.
1860 PH/20 There was another similar case to PH/19, when a complete host name was
1861 given in a host list; looking up its IP address could give an abbreviated
1862 form, whereas the current host's name might or might not be abbreviated.
1863 The same fix has been applied.
1869 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1870 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1873 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1874 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1876 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1878 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1879 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1880 Also, for "Auto-Submitted", treat anything other than "no" as
1881 non-personal, in accordance with RFC 3834. (Previously it treated
1882 anything starting "auto-" as non-personal.)
1884 TF/01 The control=submission/name=... option had a problem with syntax
1885 errors if the name included a slash character. The /name= option
1886 now slurps the rest of the string, so it can include any characters
1887 but it must come last in the list of options (after /sender_retain
1890 PH/05 Some modifications to the interface to the fake nameserver for the new
1898 TK/01 Added the "success_on_redirect" address verification option. See
1899 NewStuff for rationale and an example.
1901 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1903 PH/02 Patch to exigrep to allow it to work on syslog lines.
1905 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1906 fread() to skip over the body file's header line, because in Cygwin the
1907 header line is locked and is inaccessible.
1909 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1910 co-exist for some time) to make it clear that it is the Exim ID that is
1911 referenced, not the Message-ID: header line.
1913 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1914 string_format() function, because snprintf() does not exist on all
1917 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1918 ${stat: expansion item.
1920 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1921 protocol synchronization error", to keep the pedants happy.
1923 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1924 well as for IRIX systems, when gcc is being used. See the host.c source
1927 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1929 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1932 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1933 transport and to the smtp transport in LMTP mode.
1935 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1937 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1938 run a filter in a subprocess. This could lead to confusion in subsequent
1939 lookups in the parent process. There should also be a search_tidyup() at
1940 the end of the subprocess.
1942 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1943 only if the host matched helo_try_verify_hosts, which caused the
1944 verification to occur when the EHLO/HELO command was issued. The ACL just
1945 tested the remembered result. Now, if a previous verification attempt has
1946 not happened, "verify = helo" does it there and then.
1948 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1950 TK/03 Fix log output including CR from clamd.
1952 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1953 exist provoked a memory error which could cause a segfault.
1955 PH/15 Installed PCRE 6.2
1957 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1959 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1960 of the problem. Specifically, suggested +O2 rather than +O1 for the
1963 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1965 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1966 record, the port was not correctly logged when the outgoing_port log
1967 selector was set (it logged the transort's default port).
1969 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1970 fallback_hosts, and "hosts" in the smtp transport.
1972 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1973 host errors such as "Connection refused".
1975 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1976 authentication with radiusclient 0.4.9:
1978 - Error returned from rc_read_config was caught wrongly
1979 - Username/password not passed on to radius server due to wrong length.
1981 The presumption is that some radiusclient API changes for 4.51/PH/17
1982 were not taken care of correctly. The code is still untested by me (my
1983 Linux distribution still has 0.3.2 of radiusclient), but it was
1984 contributed by a Radius user.
1986 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1987 expanding the "port" option of the smtp transport.
1989 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1990 while reading a MIME header. Thanks to Tom Hughes for a patch.
1992 PH/24 Include config.h inside local_scan.h so that configuration settings are
1995 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1996 that in an ACL for RCPT (for example), you can examine exactly what was
1999 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2000 commands, but it was not correctly comparing the address with the actual
2001 client host address. Thus, it would show the EHLO address in Received:
2002 header lines when this was not necessary.
2004 PH/27 Added the % operator to ${eval:}.
2006 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2007 it should be ignoring failures (because with -C, for example, it has lost
2008 privilege). It wasn't ignoring creation failures other than "already
2011 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2014 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2015 because an input line was too long, either on its own, or by virtue of
2016 too many continuations, the temporary file was not being removed, and the
2017 return code was incorrect.
2019 PH/31 Missing "BOOL" in function definition in filtertest.c.
2021 PH/32 Applied Sieve patches from the maintainer.
2023 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2025 PH/33 Added "verify = not_blind".
2027 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2028 Local/Makefile (with some defaults set). These are used in built scripts
2029 such as exicyclog, but they have never been used in the exim_install
2030 script (though there are many overriding facilities there). I have
2031 arranged that the exim_install script now takes note of these two
2034 PH/35 Installed configuration files for Dragonfly.
2036 PH/36 When a locally submitted message by a trusted user did not contain a
2037 From: header, and the sender address was obtained from -f or from an SMTP
2038 MAIL command, and the trusted user did not use -F to supply a sender
2039 name, $originator_name was incorrectly used when constructing a From:
2040 header. Furthermore, $originator_name was used for submission mode
2041 messages from external hosts without From: headers in a similar way,
2042 which is clearly wrong.
2044 PH/37 Added control=suppress_local_fixups.
2046 PH/38 When log_selector = +received_sender was set, and the addition of the
2047 sender made the log line's construction buffer exactly full, or one byte
2048 less than full, an overflow happened when the terminating "\n" was
2051 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2052 when the result of a list match is failure because a DNS lookup failed.
2054 PH/40 RM_COMMAND is now used in the building process.
2056 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2057 the "build-* directories that it finds.
2059 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2060 address in a domain literal was a prefix of an interface address.
2062 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2063 when verifying a sender address, unless rewrite_headers is false.
2065 PH/44 Wrote a long comment about why errors_to addresses are verified as
2066 recipients, not senders.
2068 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2069 the ratelimit ACL was added.
2071 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2073 PH/46 Added extra information about PostgreSQL errors to the error string.
2075 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2076 avoiding the need to install special zones in a real server. This is
2077 backwards compatible; if it can't find the fake resolver, it drops back.
2078 Thus, both old and new test suites can be run.
2080 TF/02 Added util/ratelimit.pl
2082 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2083 clock is set back in time.
2085 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2086 Candler <B.Candler@pobox.com>.
2088 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2089 OK for addresses that are the result of redirections.
2091 PH/48 A number of further additions for the benefit of the new test suite,
2092 including a fake gethostbyname() that interfaces to the fake DNS resolver
2095 TF/06 The fix for widen_domains has also been applied to qualify_single and
2096 search_parents which are the other dnslookup options that can cause
2099 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2102 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2104 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2105 service" if it encountered an MX record with an empty target hostname.
2106 The message is now "an MX or SRV record indicated no SMTP service".
2108 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2109 if the DNS of the sending site is misconfigured. This is quite a
2110 common situation. This change restores the behaviour of treating a
2111 helo verification defer as a failure.
2113 PH/51 If self=fail was set on a router, the bounce message did not include the
2114 actual error message.
2120 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2122 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2123 command itself ended in error, the underlying message about the transport
2124 filter timeout was being overwritten with the pipe command error. Now the
2125 underlying error message should be appended to the second error message.
2127 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2129 PH/02 Reduce the amount of output that "make" produces by default. Full output
2130 can still be requested.
2132 PH/03 The warning log line about a condition test deferring for a "warn" verb
2133 was being output only once per connection, rather than after each
2134 occurrence (because it was using the same function as for successful
2135 "warn" verbs). This seems wrong, so I have changed it.
2137 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2138 it should not have, which might have caused a crash in the right
2139 circumstances, but probably never did.
2141 PH/04 Installed a modified version of Tony Finch's patch to make submission
2142 mode fix the return path as well as the Sender: header line, and to
2143 add a /name= option so that you can make the user's friendly name appear
2146 TF/03 Added the control = fakedefer ACL modifier.
2148 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2149 Mark Lowes for thorough testing.
2151 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2153 TK/03 Merged latest SRS patch from Miles Wilton.
2155 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2156 with the definition in sysexits.h (which is #included earlier).
2157 Fortunately, Exim does not actually use EX_OK. The code used to try to
2158 preserve the sysexits.h value, by assumimg that macro definitions were
2159 scanned for macro replacements. I have been disabused of this notion,
2160 so now the code just undefines EX_OK before #including unistd.h.
2162 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2163 in the smtp transport. When a block could not be written in a single
2164 write() function, the timeout was being re-applied to each part-write.
2165 This seems wrong - if the receiver was accepting one byte at a time it
2166 would take for ever. The timeout is now adjusted when this happens. It
2167 doesn't have to be particularly precise.
2169 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2170 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2172 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2173 without a domain if the check to <postmaster@domain> fails.
2175 SC/01 Eximstats: added -xls and the ability to specify output files
2176 (patch written by Frank Heydlauf).
2178 SC/02 Eximstats: use FileHandles for outputing results.
2180 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2182 SC/04 Eximstats: fixed display of large numbers with -nvr option
2184 SC/05 Eximstats: fixed merging of reports with empty tables.
2186 SC/06 Eximstats: added the -include_original_destination flag
2188 SC/07 Eximstats: removed tabs and trailing whitespace.
2190 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2192 TK/06 MBOX spool code: Add real "From " MBOX separator line
2193 so the .eml file is really in mbox format (even though
2194 most programs do not really care). Patch from Alex Miller.
2196 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2197 The latter is generated from $received_to and is only set if the
2198 message has one envelope recipient. SA can use these headers,
2199 obviously out-of-the-box. Patch from Alex Miller.
2201 PH/08 The ${def test on a variable was returning false if the variable's
2202 value was "0", contrary to what the specification has always said!
2203 The result should be true unless the variable is empty.
2205 PH/09 The syntax error of a character other than { following "${if
2206 def:variable_name" (after optional whitespace) was not being diagnosed.
2207 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2208 accidental colon was present, for example, could give incorrect results.
2210 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2211 result is used (not including appendfile, where other changes are about
2214 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2215 This involved changing a lot of size variables from int to off_t. It
2216 should work with maildirs and everything.
2218 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2219 spamd dying while we are connected to it.
2221 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2224 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2225 function for BDB 4.3.
2227 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2229 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2230 the sources for PCRE have been re-arranged and more files are now
2233 PH/15 The code I had for printing potentially long long variables in PH/11
2234 above was not the best (it lost precision). The length of off_t variables
2235 is now inspected at build time, and an appropriate printing format (%ld
2236 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2237 to be "long long int" or "long int". This is needed for the internal
2238 formatting function string_vformat().
2240 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2241 the configuration file to be ":syslog", then the script "guesses" where
2242 the logs files are, rather than using the compiled in default. In our
2243 case the guess is not the same as the compiled default, so the script
2244 suddenly stopped working when I started to use syslog. The patch checks
2245 to see if log_file_path is "". If so, it attempts to read it from exim
2246 with no configuration file to get the compiled in version, before it
2247 falls back to the previous guessing code."
2249 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2250 implementing BATV in an Exim configuration. See NewStuff for the gory
2253 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2254 Makefile that are specific to HP-UX.
2256 PH/18 If the "use_postmaster" option was set for a recipient callout together
2257 with the "random" option, the postmaster address was used as the MAIL
2258 FROM address for the random test, but not for the subsequent recipient
2259 test. It is now used for both.
2261 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2262 patch removes a few documentation additions to RFC 3028, because the
2263 latest draft now contains them. It adds the new en;ascii-case comparator
2264 and a new error check for 8bit text in MIME parts. Comparator and
2265 require names are now matched exactly. I enabled the subaddress
2266 extension, but it is not well tested yet (read: it works for me)."
2268 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2269 rework some of the code of TK/09 above to avoid the hardwired use of
2270 "%lld" and "long long". Replaced the call to snprintf() with a call to
2273 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2274 records point to non-existent hosts", "retry timeout exceeded", and
2275 "retry time not reached for any host after a long failure period".
2277 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2278 experimental DomainKeys support:
2280 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2281 (2) On an error such as an illegally used "control", the wrong name for
2282 the control was given.
2284 These problems did NOT occur unless DomainKeys support was compiled.
2286 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2288 PH/24 Added ${if match_ip condition.
2290 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2291 databases so that it will be absolutely obvious if a crash occurs in the
2292 DB library. This is a regular occurrence (often caused by mis-matched
2295 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2296 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2297 picked up on a user's system that detects such things. There doesn't seem
2298 to be a gcc warning option for this - only an attribute that has to be
2299 put on the function's prototype. It seems that in Fedora Core 4 they have
2300 set this on a number of new functions. No doubt there will be more in due
2303 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2304 specify a transport. However, if an address that was verified by such a
2305 router was the subject of a callout, Exim crashed because it tried to
2306 read the rcpt_include_affixes from the non-existent transport. Now it
2307 just assumes that the setting of that option is false. This bug was
2308 introduced by 4.51/PH/31.
2310 PH/28 Changed -d+all to exclude +memory, because that information is very
2311 rarely of interest, but it makes the output a lot bigger. People tend to
2312 do -d+all out of habit.
2314 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2315 code in os-type was giving problems when libc.so lives in lib64, like on
2318 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2319 aren't the modern standard, and it seems that some systems' include files
2320 don't always have them. Exim was already checking for some of the newer
2321 ones like T_AAAA, and defining it itself. I've added checks for all the
2322 record types that Exim uses.
2324 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2325 not automatically generating a new one, as it is supposed to. This
2326 prevented TLS from working. If the file did exist, but contained invalid
2327 data, a new version was generated, as expected. It was only the case of a
2328 non-existent file that was broken.
2330 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2331 with a change in libdomainkeys > 0.64.
2333 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2334 from DNS. If the selector record carries the flag, it now has
2335 precedence over the domain-wide flag.
2337 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2339 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2340 the use of a port name that isn't defined in /etc/services) occurred, the
2341 message was deferred as in a normal delivery, and thus remained on the
2342 spool, instead of being failed because of the mua_wrapper setting. This
2343 is now fixed, and I tidied up some of the mua_wrapper messages at the
2346 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2347 the messages in a hash of arrays rather than using individual hashes.
2348 This is a bit cleaner and results in dramatic memory savings, albeit
2349 at a slight CPU cost.
2351 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2352 as requested by Marc Sherman.
2354 SC/10 Eximstats: added histograms for user specified patterns as requested
2357 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2359 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2360 fopen() in the content-scanning modules that did not already have it.
2366 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2367 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2369 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2371 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2373 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2374 to test Sieve filters that use "vacation".
2376 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2377 that changes the way the GnuTLS parameters are stored in the cache file.
2378 The new format can be generated externally. For backward compatibility,
2379 if the data in the cache doesn't make sense, Exim assumes it has read an
2380 old-format file, and it generates new data and writes a new file. This
2381 means that you can't go back to an older release without removing the
2384 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2385 work if there are any delivery delays because "one_time" forces the
2386 parent to be marked "delivered", so its unseen clone is never tried
2387 again. For this reason, Exim now forbids the simultaneous setting of
2390 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2391 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2392 entry if you want to know the details. The fix, however, neglected to
2393 consider the case when local delivery batching is involved. The test for
2394 "previously delivered" was not happening when checking to see if an
2395 address could be batched with a previous (undelivered) one; under
2396 certain circumstances this could lead to multiple deliveries to the same
2399 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2400 in its include files, and this causes problems building Exim.
2402 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2403 header_syntax) but Exim was just ignoring anything given after a slash.
2404 In particular, this caused confusion with an attempt to use "verify =
2405 reverse_host_lookup/defer_ok". An error is now given when options are
2406 supplied for verify items that do not have them. (Maybe reverse_host_
2407 lookup should have a defer_ok option, but that's a different point.)
2409 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2410 defined by RFC 821) to 2048, because there were problems with some AUTH
2411 commands, and RFC 1869 says the size should be increased for extended
2412 SMTP commands that take arguments.
2414 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2417 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2418 "unknown" error; now it says that the functionality isn't in the binary.
2420 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2421 an address' error message when a string expansion fails (syntax or
2422 whatever). Otherwise the password may appear in the log. Following change
2423 PH/42 below, there is no longer a chance of it appearing in a bounce
2426 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2428 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2429 of its addresses was ever tried. (Bugzilla bug #2.)
2431 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2432 the result incorrectly in the debug output. (It correctly added a newline
2433 to what was transported.)
2435 TF/01 Added $received_time.
2437 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2438 commented out examples of how to interface to a virus scanner and to
2439 SpamAssassin. Also added commented examples of av_scanner and
2440 spamd_address settings.
2442 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2443 and controls are allowed in which ACLs. There were a couple of minor
2444 errors. Some of the entries in the conditions table (which is a table of
2445 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2446 negation of where the condition IS allowed.
2448 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2450 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2451 header file does not have a version number, so I've had to invent a new
2452 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2453 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2454 radiusclient), but it was contributed by a Radius user.
2456 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2457 files or directories, for OpenSSL.
2459 PH/19 When an Exim process that is running as root has to create an Exim log
2460 file, it does so in a subprocess that runs as exim:exim so as to get the
2461 ownership right at creation (otherwise, other Exim processes might see
2462 the file with the wrong ownership). There was no test for failure of this
2463 fork() call, which would lead to the process getting stuck as it waited
2464 for a non-existent subprocess. Forks do occasionally fail when resources
2465 run out. I reviewed all the other calls to fork(); they all seem to check
2468 PH/20 When checking for unexpected SMTP input at connect time (before writing
2469 the banner), Exim was not dealing correctly with a non-positive return
2470 from the read() function. If the client had disconnected by this time,
2471 the result was a log entry for a synchronization error with an empty
2472 string after "input=" when read() returned zero. If read() returned -1
2473 (an event I could not check), uninitialized data bytes were printed.
2474 There were reports of junk text (parts of files, etc) appearing after
2477 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2479 PH/22 Added support for macro redefinition, and (re)definition in between
2480 driver and ACL definitions.
2482 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2483 forgetting to use the resulting value; it was using the unexpanded value.
2485 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2486 hadn't been configured. The fix is from Juergen Kreileder, who
2487 understands it better than I do:
2489 "Here's what I see happening with three configured cyrus_sasl
2490 authenticators configured (plain, login, cram-md5):
2492 On startup auth_cyrus_sasl_init() gets called for each of these.
2493 This means three calls to sasl_listmech() without a specified mech_list.
2494 => SASL tests which mechs of all available mechs actually work
2495 => three warnings about OTP not working
2496 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2498 With the patch, sasl_listmech() also gets called three times. But now
2499 SASL's mech_list option is set to the server_mech specified in the the
2500 authenticator. Or in other words, the answer from sasl_listmech()
2501 gets limited to just the mech you're testing for (which is different
2503 => the return list contains just 'plain' or 'login', 'cram-md5' or
2504 nothing depending on the value of ob->server_mech.
2506 I've just tested the patch: Authentication still works fine,
2507 unavailable mechs specified in the exim configuration are still
2508 caught, and the auth.log warnings about OTP are gone."
2510 PH/25 When debugging is enabled, the contents of the command line are added
2511 to the debugging output, even when log_selector=+arguments is not
2514 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2515 answer is "GNU", and only if the return is "GNU/something" is the answer
2518 PH/27 $acl_verify_message is now set immediately after the failure of a
2519 verification in an ACL, and so is available in subsequent modifiers. In
2520 particular, the message can be preserved by coding like this:
2522 warn !verify = sender
2523 set acl_m0 = $acl_verify_message
2525 Previously, $acl_verify_message was set only while expanding "message"
2526 and "log_message" when a very denied access.
2528 PH/28 Modified OS/os.c-Linux with
2530 -#ifndef OS_LOAD_AVERAGE
2531 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2533 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2534 nomenclature these days.)
2536 PH/29 Installed patch from the Sieve maintainer that adds the options
2537 sieve_useraddress and sieve_subaddress to the redirect router.
2539 PH/30 In these circumstances:
2540 . Two addresses routed to the same list of hosts;
2541 . First host does not offer TLS;
2542 . First host accepts first address;
2543 . First host gives temporary error to second address;
2544 . Second host offers TLS and a TLS session is established;
2545 . Second host accepts second address.
2546 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2547 and peerdn, if requested) that were in fact used only for the second
2550 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2551 attention to any local part prefix or suffix that was matched by the
2552 router that accepted the address. It now behaves in the same way as it
2553 does for delivery: the affixes are removed from the local part unless
2554 rcpt_include_affixes is set on the transport.
2556 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2557 timeout during the DATA phase of an incoming message.
2559 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2560 applied a patch sanctioned by the Sieve maintainer.
2562 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2563 the uid or gid is negative. A case of a negative gid caused this to be
2564 noticed. The fix allows for either to be negative.
2566 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2567 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2570 PH/36 The change PH/12 above was broken. Fixed it.
2572 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2573 the grounds that routing the same address twice would always produce the
2574 same answer. This might have been true once, but it is certainly no
2575 longer true now. Routing a child address may depend on the previous
2576 routing that produced that child. Some complicated redirection strategies
2577 went wrong when messages had multiple recipients, and made Exim's
2578 behaviour dependent on the order in which the addresses were given.
2580 I have moved the duplicate checking until after the routing is complete.
2581 Exim scans the addresses that are assigned to local and remote
2582 transports, and removes any duplicates. This means that more work will be
2583 done, as duplicates will always all be routed, but duplicates are
2584 presumably rare, so I don't expect this is of any significance.
2586 For deliveries to pipes, files, and autoreplies, the duplicate checking
2587 still happens during the routing process, since they are not going to be
2590 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2591 It corrects a timeout issue with spamd. This is Ian's comment: "The
2592 background is that sometimes spamd either never reads data from a
2593 connection it has accepted, or it never writes response data. The exiscan
2594 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2595 blindly assumes that writes won't block so it may never time out."
2597 PH/39 Allow G after quota size as well as K and M.
2599 PH/40 The value set for $authenticated_id in an authenticator may not contain
2600 binary zeroes or newlines because the value is written to log lines and
2601 to spool files. There was no check on this. Now the value is run through
2602 the string_printing() function so that such characters are converted to
2603 printable escape sequences.
2605 PH/41 $message_linecount is a new variable that contains the total number of
2606 lines in the message. Compare $body_linecount, which is the count for the
2609 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2610 bounce and delay warning messages, except in certain special cases, which
2613 (a) An SMTP error message from a remote host;
2614 (b) A message specified in a :fail: redirection;
2615 (c) A message specified in a "fail" command in a system filter;
2616 (d) A message specified in a FAIL return from the queryprogram router;
2617 (e) A message specified by the cannot_route_message router option.
2619 In these cases only, Exim does include the error details in bounce and
2620 warning messages. There are also a few cases where bland messages such
2621 as "unrouteable address" or "local delivery error" are given.
2623 PH/43 $value is now also set for the "else" part of a ${run expansion.
2625 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2626 being worked on, but at least Exim now implements the latest version to
2629 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2630 process to complete was treated as a delivery failure, a timeout while
2631 writing the message to the pipe was logged, but erroneously treated as a
2632 successful delivery. Such timeouts include transport filter timeouts. For
2633 consistency with the overall process timeout, these timeouts are now
2634 treated as errors, giving rise to delivery failures by default. However,
2635 there is now a new Boolean option for the pipe transport called
2636 timeout_defer, which, if set TRUE, converts the failures into defers for
2637 both kinds of timeout. A transport filter timeout is now identified in
2640 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2641 systems where "make" and "gmake" are different, calling "gmake" at top
2642 level broke things. I've arranged for the value of $(MAKE) to be passed
2643 from the Makefile to this script so that it can call the same version of
2647 A note about Exim versions 4.44 and 4.50
2648 ----------------------------------------
2650 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2651 changes of various kinds. As a consequence, a big documentation update was
2652 needed. This delayed the release for rather longer than seemed good, especially
2653 in the light of a couple of (minor) security issues. Therefore, the changes
2654 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2655 release. So 4.44 and 4.50 are in effect two different branches that both start
2658 I have left the 4.50 change log unchanged; it contains all the changes since
2659 4.43. The change log for 4.44 is below; many of its items are identical to
2660 those for 4.50. This seems to be the most sensible way to preserve the
2661 historical information.
2667 1. Minor wording change to the doc/README.SIEVE file.
2669 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2670 computation of the current number of files was incorrect.
2672 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2673 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2676 4. Give more explanation in the error message when the command for a transport
2677 filter fails to execute.
2679 5. There are several places where Exim runs a non-Exim command in a
2680 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2681 was being done only for the command run by the queryprogram router. It is
2682 now done for all such subprocesses. The other cases are: ${run, transport
2683 filters, and the commands run by the lmtp and pipe transports.
2685 6. Added CONFIGURE_GROUP build-time option.
2687 7. Some older OS have a limit of 256 on the maximum number of file
2688 descriptors. Exim was using setrlimit() to set 1000 as a large value
2689 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2690 systems. I've change it so that if it can't get 1000, it tries for 256.
2692 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2693 was an oversight, and furthermore, ever since the addition of extra
2694 controls (e.g. 4.43/32), the checks on when to allow different forms of
2695 "control" were broken. There should now be diagnostics for all cases when a
2696 control that does not make sense is encountered.
2698 9. Added the /retain_sender option to "control=submission".
2700 10. $recipients is now available in the predata ACL (oversight).
2702 11. Tidy the search cache before the fork to do a delivery from a message
2703 received from the command line. Otherwise the child will trigger a lookup
2704 failure and thereby defer the delivery if it tries to use (for example) a
2705 cached ldap connection that the parent has called unbind on.
2707 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2708 of $address_data from the recipient verification was clobbered by the
2709 sender verification.
2711 13. The value of address_data from a sender verification is now available in
2712 $sender_address_data in subsequent conditions in the ACL statement.
2714 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2716 15. Added a new option "connect=<time>" to callout options, to set a different
2719 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2720 was its contents. (It was OK if the option was not defined at all.)
2722 17. A "Completed" log line is now written for messages that are removed from
2723 the spool by the -Mrm option.
2725 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2726 information about exactly what failed.
2728 19. Added -dd to debug only the daemon process.
2730 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2731 handles timeouts, both on the server side and network timeouts. Renamed the
2732 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2734 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2735 It is now set to "smtps".
2737 22. $host_address is now set to the target address during the checking of
2738 ignore_target_hosts.
2740 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2741 being passed; this would have caused $sender_host_name to have been used if
2742 matching the list had actually called for a host name (not very likely,
2743 since this list is usually IP addresses). A host name is now passed as
2746 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2747 code that specifies a non-restarting handler (typically sigaction() in
2748 modern systems) in an attempt to fix a rare and obscure crash bug.
2750 25. Narrowed the window for a race in the daemon that could cause it to ignore
2751 SIGCHLD signals. This is not a major problem, because they are used only to
2752 wake it up if nothing else does.
2754 26. A malformed maildirsize file could cause Exim to calculate negative values
2755 for the mailbox size or file count. Odd effects could occur as a result.
2756 The maildirsize information is now recalculated if the size or filecount
2759 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2760 support for a long time. Removed HAVE_SYS_VFS_H.
2762 28. Installed the latest version of exipick from John Jetmore.
2764 29. In an address list, if the pattern was not a regular expression, an empty
2765 subject address (from a bounce message) matched only if the pattern was an
2766 empty string. Non-empty patterns were not even tested. This was the wrong
2767 because it is perfectly reasonable to use an empty address as part of a
2768 database query. An empty address is now tested by patterns that are
2769 lookups. However, all the other forms of pattern expect the subject to
2770 contain a local part and a domain, and therefore, for them, an empty
2771 address still always fails if the pattern is not itself empty.
2773 30. Exim went into a mad DNS loop when attempting to do a callout where the
2774 host was specified on an smtp transport, and looking it up yielded more
2775 than one IP address.
2777 31. Re-factored the code for checking spool and log partition space into a
2778 function that finds that data and another that does the check. The former
2779 is then used to implement four new variables: $spool_space, $log_space,
2780 $spool_inodes, and $log_inodes.
2782 32. The RFC2047 encoding function was originally intended for short strings
2783 such as real names; it was not keeping to the 75-character limit for
2784 encoded words that the RFC imposes. It now respects the limit, and
2785 generates multiple encoded words if necessary. To be on the safe side, I
2786 have increased the buffer size for the ${rfc2047: expansion operator from
2789 33. It is now permitted to omit both strings after an "if" condition; if the
2790 condition is true, the result is "true". As before, when the second string
2791 is omitted, a false condition yields an empty string. This makes it less
2792 cumbersome to write custom ACL and router conditions.
2794 34. Failure to deliver a bounce message always caused it to be frozen, even if
2795 there was an errors_to setting on the router. The errors_to setting is now
2798 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2799 canonical form (fully expanded) before being placed in
2800 $sender_host_address.
2802 36. The table in the code that translates DNS record types into text (T_A to
2803 "A" for instance) was missing entries for NS and CNAME. It is just possible
2804 that this could have caused confusion if both these types were looked up
2805 for the same domain, because the text type is used as part of Exim's
2806 per-process caching. But the chance of anyone hitting this buglet seems
2809 37. The dnsdb lookup has been extended in a number of ways.
2811 (1) There is a new type, "zns", which walks up the domain tree until it
2812 finds some nameserver records. It should be used with care.
2814 (2) There is a new type, "mxh", which is like "mx" except that it returns
2815 just the host names, not the priorities.
2817 (3) It is now possible to give a list of domains (or IP addresses) to be
2818 looked up. The behaviour when one of the lookups defers can be
2819 controlled by a keyword.
2821 (4) It is now possible to specify the separator character for use when
2822 multiple records are returned.
2824 38. The dnslists ACL condition has been extended: it is now possible to supply
2825 a list of IP addresses and/or domains to be looked up in a particular DNS
2828 39. Added log_selector=+queue_time_overall.
2830 40. When running the queue in the test harness, wait just a tad after forking a
2831 delivery process, to get repeatability of debugging output.
2833 41. Include certificate and key file names in error message when GnuTLS fails
2834 to set them up, because the GnuTLS error message doesn't include the name
2835 of the failing file when there is a problem reading it.
2837 42. Allow both -bf and -bF in the same test run.
2839 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2841 44. The "Exiscan patch" is now merged into the mainline Exim source.
2843 45. Sometimes the final signoff response after QUIT could fail to get
2844 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2845 < 0 before doing a fflush(). This bug looks as though it goes back to the
2846 introduction of TLS in release 3.20, but "sometimes" must have been rare
2847 because the tests only now provoked it.
2849 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2850 (this can affect the format of dates).
2852 47. exim_tidydb, when checking for the continued existence of a message for
2853 which it has found a message-specific retry record, was not finding
2854 messages that were in split spool directories. Consequently, it was
2855 deleting retry records that should have stayed in existence.
2857 48. Steve fixed some bugs in eximstats.
2859 49. The SPA authentication driver was not abandoning authentication and moving
2860 on to the next authenticator when an expansion was forced to fail,
2861 contradicting the general specification for all authenticators. Instead it
2862 was generating a temporary error. It now behaves as specified.
2864 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2865 (the order specifies the preference for clients). The order is now AES256,
2866 AES128, 3DES, ARCFOUR128.
2868 51. Small patch to Sieve code - explicitly set From: when generating an
2871 52. Exim crashed if a remote delivery caused a very long error message to be
2872 recorded - for instance if somebody sent an entire SpamAssassin report back
2873 as a large number of 550 error lines. This bug was coincidentally fixed by
2874 increasing the size of one of Exim's internal buffers (big_buffer) that
2875 happened as part of the Exiscan merge. However, to be on the safe side, I
2876 have made the code more robust (and fixed the comments that describe what
2879 53. Now that there can be additional text after "Completed" in log lines (if
2880 the queue_time_overall log selector is set), a one-byte patch to exigrep
2881 was needed to allow it to recognize "Completed" as not the last thing in
2884 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2885 patch that reportedly fixes this has been added. I am not expert enough to
2886 create a test for it. This is what the patch creator wrote:
2888 "I found a little strange behaviour of ldap code when working with
2889 Windows 2003 AD Domain, where users was placed in more than one
2890 Organization Units. When I tried to give exim partial DN, the exit code
2891 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2892 But simultaneously result of request was absolutely normal ldap result,
2893 so I produce this patch..."
2895 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2896 REFERENCE, so I have modified the code to exclude the patch when that macro
2899 55. Some experimental protocols are using DNS PTR records for new purposes. The
2900 keys for these records are domain names, not reversed IP addresses. The
2901 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2902 leaves it alone. Component reversal etc. now happens only for IP addresses.
2905 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2907 57. Double the size of the debug message buffer (to 2048) so that more of very
2908 long debug lines gets shown.
2910 58. The exicyclog utility now does better if the number of log files to keep
2911 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2913 59. Two changes related to the smtp_active_hostname option:
2915 (1) $smtp_active_hostname is now available as a variable.
2916 (2) The default for smtp_banner uses $smtp_active_hostname instead
2917 of $primary_hostname.
2919 60. The host_aton() function is supposed to be passed a string that is known
2920 to be a valid IP address. However, in the case of IPv6 addresses, it was
2921 not checking this. This is a hostage to fortune. Exim now panics and dies
2922 if the condition is not met. A case was found where this could be provoked
2923 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2924 components; fortuitously, this particular loophole had already been fixed
2925 by change 4.50/55 above.
2927 If there are any other similar loopholes, the new check in host_aton()
2928 itself should stop them being exploited. The report I received stated that
2929 data on the command line could provoke the exploit when Exim was running as
2930 exim, but did not say which command line option was involved. All I could
2931 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2932 running as the user.
2935 61. There was a buffer overflow vulnerability in the SPA authentication code
2936 (which came originally from the Samba project). I have added a test to the
2937 spa_base64_to_bits() function which I hope fixes it.
2940 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2941 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2943 63. The daemon start-up calls getloadavg() while still root for those OS that
2944 need the first call to be done as root, but it missed one case: when
2945 deliver_queue_load_max is set with deliver_drop_privilege. This is
2946 necessary for the benefit of the queue runner, because there is no re-exec
2947 when deliver_drop_privilege is set.
2949 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2950 This has been fixed.
2952 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2953 was in use, was not putting the data itself into the right store pool;
2954 consequently, it could be overwritten for a subsequent message in the same
2955 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2958 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2960 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2961 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2962 the function were treating the return as a boolean value, which happened to
2963 work because 0=false and not-0=true, but is not correct code.
2965 68. The host_aton() function was not handling scoped IPv6 addresses (those
2966 with, for example, "%eth0" on the end) correctly.
2968 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2969 negated items (that is, ~something) in unsigned ints. Some compilers
2970 apparently mutter when there is no cast.
2972 70. If an address verification called from an ACL failed, and did not produce a
2973 user-specific message (i.e. there was only a "system" message), nothing was
2974 put in $acl_verify_message. In this situation, it now puts the system
2977 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2978 session; change 4.31/43 added the unwanted input to the log line - except
2979 that it did not do this in the start of session case. It now does.
2981 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2982 This is wrong and can cause the other end to generate a synchronization
2983 error if it is another Exim or anything else that does the synchronization
2984 check. A QUIT command is no longer sent after a timeout.
2986 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2987 during host lookups.
2989 74. The defer_ok option of callout verification was not working if it was used
2990 when verifying addresses in header lines, that is, for this case:
2992 verify = header_sender/callout=defer_ok
2994 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2995 those file descriptors could be used for SMTP connections. If anything
2996 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2997 could be sent to the SMTP client, causing chaos. The daemon now opens
2998 stdin, stdout, and stderr to /dev/null when it puts itself into the
3001 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3002 log by default. The user can override this with suitable Perl magic.
3004 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3005 the log message when discard triggers, was not working for the DATA ACL or
3006 for the non-SMTP ACL.
3008 78. Error message wording change in sieve.c.
3010 79. If smtp_accept_max_per_host was set, the number of connections could be
3011 restricted to fewer than expected, because the daemon was trying to set up
3012 a new connection before checking whether the processes handling previous
3013 connections had finished. The check for completed processes is now done
3014 earlier. On busy systems, this bug wouldn't be noticed because something
3015 else would have woken the daemon, and it would have reaped the completed
3018 80. If a message was submitted locally by a user whose login name contained one
3019 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3020 It caused a spool format error. I have fixed the spool reading code. A
3021 related problem was that the "from" clause in the Received: line became
3022 illegal because of the space(s). It is now covered by ${quote_local_part.
3024 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3027 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3029 83. Patch from Sieve maintainer to fix unterminated string problem in
3030 "vacation" handling.
3032 84. Some minor changes to the Linux configuration files to help with other
3033 OS variants using glibc.
3035 85. One more patch for Sieve to update vacation handling to latest spec.
3038 ----------------------------------------------------
3039 See the note above about the 4.44 and 4.50 releases.
3040 ----------------------------------------------------
3046 1. Change 4.43/35 introduced a bug that caused file counts to be
3047 incorrectly computed when quota_filecount was set in an appendfile
3050 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3051 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3054 3. Give more explanation in the error message when the command for a transport
3055 filter fails to execute.
3057 4. There are several places where Exim runs a non-Exim command in a
3058 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3059 was being done only for the command run by the queryprogram router. It is
3060 now done for all such subprocesses. The other cases are: ${run, transport
3061 filters, and the commands run by the lmtp and pipe transports.
3063 5. Some older OS have a limit of 256 on the maximum number of file
3064 descriptors. Exim was using setrlimit() to set 1000 as a large value
3065 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3066 systems. I've change it so that if it can't get 1000, it tries for 256.
3068 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3069 was an oversight, and furthermore, ever since the addition of extra
3070 controls (e.g. 4.43/32), the checks on when to allow different forms of
3071 "control" were broken. There should now be diagnostics for all cases when a
3072 control that does not make sense is encountered.
3074 7. $recipients is now available in the predata ACL (oversight).
3076 8. Tidy the search cache before the fork to do a delivery from a message
3077 received from the command line. Otherwise the child will trigger a lookup
3078 failure and thereby defer the delivery if it tries to use (for example) a
3079 cached ldap connection that the parent has called unbind on.
3081 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3082 of $address_data from the recipient verification was clobbered by the
3083 sender verification.
3085 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3086 was its contents. (It was OK if the option was not defined at all.)
3088 11. A "Completed" log line is now written for messages that are removed from
3089 the spool by the -Mrm option.
3091 12. $host_address is now set to the target address during the checking of
3092 ignore_target_hosts.
3094 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3095 being passed; this would have caused $sender_host_name to have been used if
3096 matching the list had actually called for a host name (not very likely,
3097 since this list is usually IP addresses). A host name is now passed as
3100 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3101 code that specifies a non-restarting handler (typically sigaction() in
3102 modern systems) in an attempt to fix a rare and obscure crash bug.
3104 15. Narrowed the window for a race in the daemon that could cause it to ignore
3105 SIGCHLD signals. This is not a major problem, because they are used only to
3106 wake it up if nothing else does.
3108 16. A malformed maildirsize file could cause Exim to calculate negative values
3109 for the mailbox size or file count. Odd effects could occur as a result.
3110 The maildirsize information is now recalculated if the size or filecount
3113 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3114 support for a long time. Removed HAVE_SYS_VFS_H.
3116 18. Updated exipick to current release from John Jetmore.
3118 19. Allow an empty sender to be matched against a lookup in an address list.
3119 Previously the only cases considered were a regular expression, or an
3122 20. Exim went into a mad DNS lookup loop when doing a callout where the
3123 host was specified on the transport, if the DNS lookup yielded more than
3126 21. The RFC2047 encoding function was originally intended for short strings
3127 such as real names; it was not keeping to the 75-character limit for
3128 encoded words that the RFC imposes. It now respects the limit, and
3129 generates multiple encoded words if necessary. To be on the safe side, I
3130 have increased the buffer size for the ${rfc2047: expansion operator from
3133 22. Failure to deliver a bounce message always caused it to be frozen, even if
3134 there was an errors_to setting on the router. The errors_to setting is now
3137 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3138 canonical form (fully expanded) before being placed in
3139 $sender_host_address.
3141 24. Updated eximstats to version 1.33
3143 25. Include certificate and key file names in error message when GnuTLS fails
3144 to set them up, because the GnuTLS error message doesn't include the name
3145 of the failing file when there is a problem reading it.
3147 26. Expand error message when OpenSSL has problems setting up cert/key files.
3150 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3151 (this can affect the format of dates).
3153 28. exim_tidydb, when checking for the continued existence of a message for
3154 which it has found a message-specific retry record, was not finding
3155 messages that were in split spool directories. Consequently, it was
3156 deleting retry records that should have stayed in existence.
3158 29. eximstats updated to version 1.35
3159 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3160 1.35 - bugfix such that pie charts by volume are generated correctly
3162 30. The SPA authentication driver was not abandoning authentication and moving
3163 on to the next authenticator when an expansion was forced to fail,
3164 contradicting the general specification for all authenticators. Instead it
3165 was generating a temporary error. It now behaves as specified.
3167 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3168 (the order specifies the preference for clients). The order is now AES256,
3169 AES128, 3DES, ARCFOUR128.
3171 31. Small patch to Sieve code - explicitly set From: when generating an
3174 32. Exim crashed if a remote delivery caused a very long error message to be
3175 recorded - for instance if somebody sent an entire SpamAssassin report back
3176 as a large number of 550 error lines. This bug was coincidentally fixed by
3177 increasing the size of one of Exim's internal buffers (big_buffer) that
3178 happened as part of the Exiscan merge. However, to be on the safe side, I
3179 have made the code more robust (and fixed the comments that describe what
3182 33. Some experimental protocols are using DNS PTR records for new purposes. The
3183 keys for these records are domain names, not reversed IP addresses. The
3184 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3185 leaves it alone. Component reversal etc. now happens only for IP addresses.
3188 34. The host_aton() function is supposed to be passed a string that is known
3189 to be a valid IP address. However, in the case of IPv6 addresses, it was
3190 not checking this. This is a hostage to fortune. Exim now panics and dies
3191 if the condition is not met. A case was found where this could be provoked
3192 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3193 components; fortuitously, this particular loophole had already been fixed
3194 by change 4.50/55 or 4.44/33 above.
3196 If there are any other similar loopholes, the new check in host_aton()
3197 itself should stop them being exploited. The report I received stated that
3198 data on the command line could provoke the exploit when Exim was running as
3199 exim, but did not say which command line option was involved. All I could
3200 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3201 running as the user.
3204 35. There was a buffer overflow vulnerability in the SPA authentication code
3205 (which came originally from the Samba project). I have added a test to the
3206 spa_base64_to_bits() function which I hope fixes it.
3209 36. The daemon start-up calls getloadavg() while still root for those OS that
3210 need the first call to be done as root, but it missed one case: when
3211 deliver_queue_load_max is set with deliver_drop_privilege. This is
3212 necessary for the benefit of the queue runner, because there is no re-exec
3213 when deliver_drop_privilege is set.
3215 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3216 was in use, was not putting the data itself into the right store pool;
3217 consequently, it could be overwritten for a subsequent message in the same
3218 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3221 38. Sometimes the final signoff response after QUIT could fail to get
3222 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3223 < 0 before doing a fflush(). This bug looks as though it goes back to the
3224 introduction of TLS in release 3.20, but "sometimes" must have been rare
3225 because the tests only now provoked it.
3231 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3232 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3233 Now it returns an integer. A number of calls were still expecting a T/F
3234 return. Fortuitously, in all cases, the tests worked in OK situations,
3235 which is the norm. However, things would have gone wrong on any write
3236 failures on the smtp file descriptor. This function is used when sending
3237 messages over SMTP and also when doing verify callouts.
3239 2. When Exim is called to do synchronous delivery of a locally submitted
3240 message (the -odf or -odi options), it no longer closes stderr before doing
3243 3. Implemented the mua_wrapper option.
3245 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3247 5. Implemented the functions header_remove(), header_testname(),
3248 header_add_at_position(), and receive_remove_recipient(), and exported them
3251 6. If an ACL "warn" statement specified the addition of headers, Exim already
3252 inserted X-ACL-Warn: at the start if there was no header name. However, it
3253 was not making this test for the second and subsequent header lines if
3254 there were newlines in the string. This meant that an invalid header could
3255 be inserted if Exim was badly configured.
3257 7. Allow an ACL "warn" statement to add header lines at the start or after all
3258 the Received: headers, as well as at the end.
3260 8. Added the rcpt_4xx retry error code.
3262 9. Added postmaster_mailfrom=xxx to callout verification option.
3264 10. Added mailfrom=xxxx to the callout verification option, for verify=
3267 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3268 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3270 12. Inserted some casts to stop certain compilers warning when using pointer
3271 differences as field lengths or precisions in printf-type calls (mostly
3272 affecting debugging statements).
3274 13. Added optional readline() support for -be (dynamically loaded).
3276 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3277 same clock tick as a message's arrival, so that its received time was the
3278 same as the "first fail" time on the retry record, and that message
3279 remained on the queue past the ultimate address timeout, every queue runner
3280 would try a delivery (because it was past the ultimate address timeout) but
3281 after another failure, the ultimate address timeout, which should have then
3282 bounced the address, did not kick in. This was a "< instead of <=" error;
3283 in most cases the first failure would have been in the next clock tick
3284 after the received time, and all would be well.
3286 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3287 being recognized when the domain list was tested by the match_domain
3288 condition in an expansion string.
3290 16. Added the ${str2b64: operator.
3292 17. Exim was always calling setrlimit() to set a large limit for the number of
3293 processes, without checking whether the existing limit was already
3294 adequate. (It did check for the limit on file descriptors.) Furthermore,
3295 errors from getrlimit() and setrlimit() were being ignored. Now they are
3296 logged to the main and panic logs, but Exim does carry on, to try to do its
3297 job under whatever limits there are.
3299 18. Imported PCRE 5.0.
3301 19. Trivial typo in log message " temporarily refused connection" (the leading
3304 20. If the log selector return_path_on_delivery was set and an address was
3305 redirected to /dev/null, the delivery process crashed because it assumed
3306 that a return path would always be set for a "successful" delivery. In this
3307 case, the whole delivery is bypassed as an optimization, and therefore no
3310 21. Internal re-arrangement: the function for sending a challenge and reading
3311 a response while authentication was assuming a zero-terminated challenge
3312 string. It's now changed to take a pointer and a length, to allow for
3313 binary data in such strings.
3315 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3317 23. Exim was not respecting finduser_retries when seeking the login of the
3318 uid under which it was called; it was always trying 10 times. (The default
3319 setting of finduser_retries is zero.) Also, it was sleeping after the final
3320 failure, which is pointless.
3322 24. Implemented tls_on_connect_ports.
3324 25. Implemented acl_smtp_predata.
3326 26. If the domain in control=submission is set empty, Exim assumes that the
3327 authenticated id is a complete email address when it generates From: or
3328 Sender: header lines.
3330 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3331 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3332 chown and chgrp in /bin and hostname in /usr/bin.
3334 28. Exim was keeping the "process log" file open after each use, just as it
3335 does for the main log. This opens the possibility of it remaining open for
3336 long periods when the USR1 signal hits a daemon. Occasional processlog
3337 errors were reported, that could have been caused by this. Anyway, it seems
3338 much more sensible not to leave this file open at all, so that is what now
3341 29. The long-running daemon process does not normally write to the log once it
3342 has entered its main loop, and it closes the log before doing so. This is
3343 so that log files can straightforwardly be renamed and moved. However,
3344 there are a couple of unusual error situations where the daemon does write
3345 log entries, and I had neglected to close the log afterwards.
3347 30. The text of an SMTP error response that was received during a remote
3348 delivery was being truncated at 512 bytes. This is too short for some of
3349 the long messages that one sometimes sees. I've increased the limit to
3352 31. It is now possible to make retry rules that apply only when a message has a
3353 specific sender, in particular, an empty sender.
3355 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3356 it possible to be selective about when SMTP synchronization is enforced.
3358 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3360 32. Implemented hosts_connection_nolog.
3362 33. Added an ACL for QUIT.
3364 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3367 35. Added mailbox_size and mailbox_filecount to appendfile.
3369 36. Added control = no_multiline_responses to ACLs.
3371 37. There was a bug in the logic of the code that waits for the clock to tick
3372 in the case where the clock went backwards by a substantial amount such
3373 that the microsecond fraction of "now" was more than the microsecond
3374 fraction of "then" (but the whole seconds number was less).
3376 38. Added support for the libradius Radius client library this is found on
3377 FreeBSD (previously only the radiusclient library was supported).
3383 1. When certain lookups returned multiple values in the form name=value, the
3384 quoting of the values was not always being done properly. Specifically:
3385 (a) If the value started with a double quote, but contained no whitespace,
3387 (b) If the value contained whitespace other than a space character (i.e.
3388 tabs or newlines or carriage returns) it was not quoted.
3389 This fix has been applied to the mysql and pgsql lookups by writing a
3390 separate quoting function and calling it from the lookup code. The fix
3391 should probably also be applied to nisplus, ibase and oracle lookups, but
3392 since I cannot test any of those, I have not disturbed their existing code.
3394 2. A hit in the callout cache for a specific address caused a log line with no
3395 reason for rejecting RCPT. Now it says "Previous (cached) callout
3396 verification failure".
3398 3. There was an off-by-one bug in the queryprogram router. An over-long
3399 return line was truncated at 256 instead of 255 characters, thereby
3400 overflowing its buffer with the terminating zero. As well as fixing this, I
3401 have increased the buffer size to 1024 (and made a note to document this).
3403 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3404 when Exim is waiting for an SMTP response from a remote server, Exim
3405 restarts its select() call on the socket, thereby resetting its timeout.
3406 This is not a problem when such interrupts are rare. Somebody set up a cron
3407 job to run exiwhat every 2 minutes, which is less than the normal select()
3408 timeout (5 or 10 minutes). This meant that the select() timeout never
3409 kicked in because it was always reset. I have fixed this by comparing the
3410 time when an interrupt arrives with the time at the start of the first call
3411 to select(). If more time than the timeout has elapsed, the interrupt is
3412 treated as a timeout.
3414 5. Some internal re-factoring in preparation for the addition of Sieve
3415 extensions (by MH). In particular, the "personal" test is moved to a
3416 separate function, and given an option for scanning Cc: and Bcc: (which is
3417 not set for Exim filters).
3419 6. When Exim created an email address using the login of the caller as the
3420 local part (e.g. when creating a From: or Sender: header line), it was not
3421 quoting the local part when it contained special characters such as @.
3423 7. Installed new OpenBSD configuration files.
3425 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3426 try to make them clearer.
3428 9. Callout options, other than the timeout value, were being ignored when
3429 verifying sender addresses in header lines. For example, when using
3431 verify = header_sender/callout=no_cache
3433 the cache was (incorrectly) being used.
3435 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3436 only the Cygwin environment.
3438 11. When return_path_on_delivery was set as a log selector, if different remote
3439 addresses in the same message used different return paths and parallel
3440 remote delivery occurred, the wrong values would sometimes be logged.
3441 (Whenever a remote delivery process finished, the return path value from
3442 the most recently started remote delivery process was logged.)
3444 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3445 lines when AUTH and/or TLS are in use. This is the "received protocol"
3446 field. Exim used to use "asmtp" for authenticated SMTP, without any
3447 indication (in the protocol name) for TLS use. Now it follows the RFC and
3448 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3449 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3450 names appear in log lines as well as in Received: header lines.
3452 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3453 extensions, and comparison tests, and to fix some bugs.
3455 14. Changes to the "personal" filter test:
3457 (1) The test was buggy in that it was just doing the equivalent of
3458 "contains" tests on header lines. For example, if a user's address was
3459 anne@some.where, the "personal" test would incorrectly be true for
3461 To: susanne@some.where
3463 This test is now done by extracting each address from the header in turn,
3464 and checking the entire address. Other tests that are part of "personal"
3465 are now done using regular expressions (for example, to check local parts
3466 of addresses in From: header lines).
3468 (2) The list of non-personal local parts in From: addresses has been
3469 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3470 taken from the Sieve specification recommendations.
3472 (3) If the message contains any header line starting with "List-" it is
3473 treated as non-personal.
3475 (4) The test for "circular" in the Subject: header line has been removed
3476 because it now seems ill-conceived.
3478 15. Minor typos in src/EDITME comments corrected.
3480 16. Installed latest exipick from John Jetmore.
3482 17. If headers_add on a router specified a text string that was too long for
3483 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3484 of string_sprintf() is now avoided.
3486 18. $message_body_size was not set (it was always zero) when running the DATA
3487 ACL and the local_scan() function.
3489 19. For the "mail" command in an Exim filter, no default was being set for
3490 the once_repeat time, causing a random time value to be used if "once" was
3491 specified. (If the value happened to be <= 0, no repeat happened.) The
3492 default is now 0s, meaning "never repeat". The "vacation" command was OK
3493 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3494 (I found it when inspecting the code).
3496 20. There is now an overall timeout for performing a callout verification. It
3497 defaults to 4 times the callout timeout, which applies to individual SMTP
3498 commands during the callout. The overall timeout applies when there is more
3499 than one host that can be tried. The timeout is checked before trying the
3500 next host. This prevents very long delays if there are a large number of
3501 hosts and all are timing out (e.g. when the network connections are timing
3502 out). The value of the overall timeout can be changed by specifying an
3503 additional sub-option for "callout", called "maxwait". For example:
3505 verify = sender/callout=5s,maxwait=20s
3507 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3508 to the end before writing, but this should make it even safer).
3510 22. Exim was forgetting that it had advertised PIPELINING for the second and
3511 subsequent messages on an SMTP connection. It was also not resetting its
3512 memory on STARTTLS and an internal HELO.
3514 23. When Exim logs an SMTP synchronization error within a session, it now
3515 records whether PIPELINING has been advertised or not.
3517 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3518 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3519 rather than long int.
3521 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3523 26. Added the never_mail option to autoreply.
3529 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3530 crash if the getsockname() call failed; this can happen if a connection is
3531 closed very soon after it is established. The problem was simply in the
3532 order in which certain operations were done, causing Exim to try to write
3533 to the SMTP stream before it had set up the file descriptor. The bug has
3534 been fixed by making things happen in the correct order.
3540 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3541 before the connection was closed, thus losing the rejection response.
3543 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3544 some early Solaris releases, but causes trouble in current releases where
3545 socklen_t is defined.
3547 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3550 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3553 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3554 that releasing the top of it at the end releases what was used for sub-
3555 expansions (unless the block got too big). However, discard this block if
3556 the first thing is a variable or header, so that we can use its block when
3557 it is dynamic (useful for very large $message_headers, for example).
3559 6. Lookups now cache *every* query, not just the most recent. A new, separate
3560 store pool is used for this. It can be recovered when all lookup caches are
3561 flushed. Lookups now release memory at the end of their result strings.
3562 This has involved some general refactoring of the lookup sources.
3564 7. Some code has been added to the store_xxx() functions to reduce the amount
3565 of flapping under certain conditions.
3567 8. log_incoming_interface used to affect only the <= reception log lines. Now
3568 it causes the local interface and port to be added to several more SMTP log
3569 lines, for example "SMTP connection from", and rejection lines.
3571 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3573 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3575 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3576 could be overwritten at the end of the current message (or the start of a
3577 new message if it was set in a HELO ACL). The value is now preserved for
3578 the duration of the SMTP connection.
3580 12. If a transport had a headers_rewrite setting, and a matching header line
3581 contained an unqualified address, that address was qualified, even if it
3582 did not match any rewriting rules. The underlying bug was that the values
3583 of the flags that permit the existence of unqualified sender and recipient
3584 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3585 non-local messages, and by -bnq for local messages) were not being
3586 preserved with the message after it was received.
3588 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3589 "next input=" as part of the text comprising the host identity instead of
3590 the correct text. The code was using the same buffer for two different
3591 strings. However, depending on which order the printing function evaluated
3592 its arguments, the bug did not always show up. Under Linux, for example, my
3593 test suite worked just fine.
3595 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3596 doesn't work with some older versions of Perl. It has been changed to "my",
3597 which in any case is probably the better facility to use.
3599 15. A really picky compiler found some instances of statements for creating
3600 error messages that either had too many or two few arguments for the format
3603 16. The size of the buffer for calls to the DNS resolver has been increased
3604 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3605 for addresses that have a lot of PTR records. This alleviates a problem; it
3606 does not fully solve it.
3608 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3609 buffer now truncates the list and logs the incident, which is the same
3610 action as happens when Exim is looking up a host name and its aliases.
3611 Previously in this situation something unpredictable would happen;
3612 sometimes it was "internal error: store_reset failed".
3614 18. If a server dropped the connection unexpectedly when an Exim client was
3615 using GnuTLS and trying to read a response, the client delivery process
3616 crashed while trying to generate an error log message.
3618 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3619 string, for example:
3621 warn message = H1: something\nH2: something
3623 the text was added as a single header line from Exim's point of view
3624 though it ended up OK in the delivered message. However, searching for the
3625 second and subsequent header lines using $h_h2: did not work. This has been
3626 fixed. Similarly, if a system filter added multiple headers in this way,
3627 the routers could not see them.
3629 20. Expanded the error message when iplsearch is called with an invalid key to
3630 suggest using net-iplsearch in a host list.
3632 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3633 ACLs are no longer actually imposed (and a message to that effect is
3636 22. If a "gecos" field in a passwd entry contained escaped characters, in
3637 particular, if it contained a \" sequence, Exim got it wrong when building
3638 a From: or a Sender: header from that name. A second bug also caused
3639 incorrect handling when an unquoted " was present following a character
3640 that needed quoting.
3642 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3643 was not being matched caselessly.
3645 24. Arranged for all hyphens in the exim.8 source to be escaped with
3648 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3649 properly. Recipient callout cache records were still being keyed to include
3650 the sender, even when use_sender was set false. This led to far more
3651 callouts that were necessary. The sender is no longer included in the key
3652 when use_sender is false.
3654 26. Added "control = submission" modifier to ACLs.
3656 27. Added the ${base62d: operator to decode base 62 numbers.
3658 28. dnsdb lookups can now access SRV records.
3660 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3661 the configuration file.
3663 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3664 (-v) mode. This makes the output for a verbose queue run more intelligible.
3666 31. Added a use_postmaster feature to recipient callouts.
3668 32. Added the $body_zerocount variable, containing the number of binary zero
3669 bytes in the message body.
3671 33. The time of last modification of the "new" subdirectory is now used as the
3672 "mailbox time last read" when there is a quota error for a maildir
3675 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3677 35. Added +ignore_unknown as a special item in host lists.
3679 36. Code for decoding IPv6 addresses in host lists is now included, even if
3680 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3681 address was recognized as an IP address, but was then not correctly decoded
3682 into binary, causing unexpected and incorrect effects when compared with
3689 1. Very minor rewording of debugging text in manualroute to say "list of
3690 hosts" instead of "hostlist".
3692 2. If verify=header_syntax was set, and a header line with an unqualified
3693 address (no domain) and a large number of spaces between the end of the
3694 name and the colon was received, the reception process suffered a buffer
3695 overflow, and (when I tested it) crashed. This was caused by some obsolete
3696 code that should have been removed. The fix is to remove it!
3698 3. When running in the test harness, delay a bit after writing a bounce
3699 message to get a bit more predictability in the log output.
3701 4. Added a call to search_tidyup() just before forking a reception process. In
3702 theory, someone could use a lookup in the expansion of smtp_accept_max_
3703 per_host which, without the tidyup, could leave open a database connection.
3705 5. Added the variables $recipient_data and $sender_data which get set from a
3706 lookup success in an ACL "recipients" or "senders" condition, or a router
3707 "senders" option, similar to $domain_data and $local_part_data.
3709 6. Moved the writing of debug_print from before to after the "senders" test
3712 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3713 problems for message scanning, either using a data ACL, or using
3714 local_scan() because the Received: header was not generated till after they
3715 were called (in order to set the time as the time of reception completion).
3716 I have revised the way this works. The header is now generated after the
3717 body is received, but before the ACL or local_scan() are called. After they
3718 are run, the timestamp in the header is updated.
3724 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3725 before starting a queue runner without re-exec. This happened only when
3726 deliver_drop_privilege was set or when the Exim user was set to root. The
3727 effect of the bug was that timeouts during subsequent deliveries caused
3728 crashes instead of being properly handled. The handler is now left at its
3729 default (and expected) setting.
3731 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3732 message, again when deliver_drop_privilege is set or Exim is run as root.
3733 The bug described in (1) was not present in this case, but the tidying up
3734 of the other signals was missing. I have made the two cases consistent.
3736 3. The ignore_target_hosts setting on a manualroute router was being ignored
3737 for hosts that were looked up using the /MX notation.
3739 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3742 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3743 operated on the sender address. After changing the $sender_address to <>
3744 for the sender address verify, Exim was re-instated it as the original
3745 (before rewriting) address, but remembering that it had rewritten it, so it
3746 wasn't rewriting it again. This bug also had the effect of breaking the
3747 sender address verification caching when the sender address was rewritten.
3749 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3750 This has been changed so that if the ip literal address matches
3751 ignore_target_hosts, the router declines.
3753 7. Added expansion conditions match_domain, match_address, and match_local_
3754 part (NOT match_host).
3756 8. The placeholder for the Received: header didn't have a length field set.
3758 9. Added code to Exim itself and to exim_lock to test for a specific race
3759 condition that could lead to file corruption when using MBX delivery. The
3760 issue is with the lockfile that is created in /tmp. If this file is removed
3761 after a process has opened it but before that process has acquired a lock,
3762 there is the potential for a second process to recreate the file and also
3763 acquire a lock. This could lead to two Exim processes writing to the file
3764 at the same time. The added code performs the same test as UW imapd; it
3765 checks after acquiring the lock that its file descriptor still refers to
3766 the same named file.
3768 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3769 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3772 11. Added the smtp_active_hostname option. If used, this will typically be made
3773 to depend on the incoming interface address. Because $interface_address is
3774 not set up until the daemon has forked a reception process, error responses
3775 that can happen earlier (such as "too many connections") no longer contain
3778 12. If an expansion in a condition on a "warn" statement fails because a lookup
3779 defers, the "warn" statement is abandoned, and the next ACL statement is
3780 processed. Previously this caused the whole ACL to be aborted.
3782 13. Added the iplsearch lookup type.
3784 14. Added ident_timeout as a log selector.
3786 15. Added tls_certificate_verified as a log selector.
3788 16. Added a global option tls_require_ciphers (compare the smtp transport
3789 option of the same name). This controls incoming TLS connections.
3791 17. I finally figured out how to make tls_require_ciphers do a similar thing
3792 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3793 before starting the TLS session.
3795 18. Tabs are now shown as \t in -bP output.
3797 19. If the log selector return_path_on_delivery was set, Exim crashed when
3798 bouncing a message because it had too many Received: header lines.
3800 20. If two routers both had headers_remove settings, and the first one included
3801 a superfluous trailing colon, the final name in the first list and the
3802 first name in the second list were incorrectly joined into one item (with a
3803 colon in the middle).
3809 1. Added -C and -D options to the exinext utility, mainly to make it easier
3810 to include in the automated testing, but these could be helpful when
3811 multiple configurations are in use.
3813 2. The exinext utility was not formatting the output nicely when there was
3814 an alternate port involved in the retry record key, nor when there was a
3815 message id as well (for retries that were specific to a specific message
3816 and a specific host). It was also confused by IPv6 addresses, because of
3817 the additional colons they contain. I have fixed the IPv4 problem, and
3818 patched it up to do a reasonable job for IPv6.
3820 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3821 delivery, the log line now contains "pipelined" if PIPELINING was used.
3823 4. An SMTP transport process used to panic and die if the bind() call to set
3824 an explicit outgoing interface failed. This has been changed; it is now
3825 treated in the same way as a connect() failure.
3827 5. A reference to $sender_host_name in the part of a conditional expansion
3828 that was being skipped was still causing a DNS lookup. This no longer
3831 6. The def: expansion condition was not recognizing references to header lines
3832 that used bh_ and bheader_.
3834 7. Added the _cache feature to named lists.
3836 8. The code for checking quota_filecount in the appendfile transport was
3837 allowing one more file than it should have been.
3839 9. For compatibility with Sendmail, the command line option
3847 and sets the incoming protocol and host name (for trusted callers). The
3848 host name and its colon can be omitted when only the protocol is to be set.
3849 Note the Exim already has two private options, -pd and -ps, that refer to
3850 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3851 "s", but I don't think that's a major issue.
3853 10. A number of refactoring changes to the code, none of which should affect
3856 (a) The number of logging options was getting close to filling up the
3857 32-bit word that was used as a bit map. I have split them into two classes:
3858 those that are passed in the argument to log_write(), and those that are
3859 only ever tested independently outside of that function. These are now in
3860 separate 32-bit words, so there is plenty of room for expansion again.
3861 There is no change in the user interface or the logging behaviour.
3863 (b) When building, for example, log lines, the code previously used a
3864 macro that called string_cat() twice, in order to add two strings. This is
3865 not really sufficiently general. Furthermore, there was one instance where
3866 it was actually wrong because one of the argument was used twice, and in
3867 one call a function was used. (As it happened, calling the function twice
3868 did not affect the overall behaviour.) The macro has been replaced by a
3869 function that can join an arbitrary number of extra strings onto a growing
3872 (c) The code for expansion conditions now uses a table and a binary chop
3873 instead of a serial search (which was left over from when there were very
3874 few conditions). Also, it now recognizes conditions like "pam" even when
3875 the relevant support is not compiled in: a suitably worded error message is
3876 given if an attempt is made to use such a condition.
3878 11. Added ${time_interval:xxxxx}.
3880 12. A bug was causing one of the ddress fields not to be passed back correctly
3881 from remote delivery subprocesses. The field in question was not being
3882 subsequently used, so this caused to problems in practice.
3884 13. Added new log selectors queue_time and deliver_time.
3886 14. Might have fixed a bug in maildirsizefile handling that threw up
3887 "unexpected character" debug warnings, and recalculated the data
3888 unnecessarily. In any case, I expanded the warning message to give more
3891 15. Added the message "Restricted characters in address" to the statements in
3892 the default ACL that block characters like @ and % in local parts.
3894 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3895 Three changes have been made:
3897 (a) There was a serious bug; a negative response to MAIL caused the whole
3898 recipient domain to be cached as invalid, thereby blocking all messages
3899 to all local parts at the same domain, from all senders. This bug has
3900 been fixed. The domain is no longer cached after a negative response to
3901 MAIL if the sender used is not empty.
3903 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3906 (c) A new callout option, "use_sender" has been added for people who want
3907 the modified behaviour.
3913 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3916 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3917 indeed breaks things for older releases.
3919 3. Added additional logging to the case where there is a problem reading data
3920 from a filter that is running in a subprocess using a pipe, in order to
3921 try to track down a specific problem.
3923 4. Testing facility fudge: when running in the test harness and attempting
3924 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3925 getting "No route to host". Convert this to a timeout.
3927 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3930 6. Some OS don't have socklen_t but use size_t instead. This affects the
3931 fifth argument of getsockopt() amongst other things. This is now
3932 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3933 can be set for individual OS. I have set it for SunOS5, OSF1, and
3934 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3935 some earlier ones do not.
3937 7. Change 4.30/15 was not doing the test caselessly.
3939 8. The standard form for an IPv6 address literal was being rejected by address
3940 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3941 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3942 this, as well as the form without the "IPv6" on the front (but only when
3943 address literals are enabled, of course).
3945 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3947 10. Exim crashed if a message with an empty sender address specified by -f
3948 encountered a router with an errors_to setting. This could be provoked only
3949 by a command such as
3953 where an empty string was supplied; "<>" did not hit this bug.
3955 11. Installed PCRE release 4.5.
3957 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3958 remained set. It is now erased.
3960 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3961 times from message ids (which are base 36 rather than the normal 62).
3963 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3964 were being counted as actual protocol errors, and logged if the log
3965 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3966 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3967 and DATA following a set of rejected RCPTs do not count as protocol errors.
3968 In other words, Exim assumes they were pipelined, though this may not
3969 actually be the case. Of course, in all cases the client gets an
3970 appropriate error code.
3972 15. If a lookup fails in an ACL condition, a message about the failure may
3973 be available; it is used if testing the ACL cannot continue, because most
3974 such messages specify what the cause of the deferral is. However, some
3975 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3976 that caused an old message to be retained and used if a later statement
3977 caused a defer, replacing the real cause of the deferral.
3979 16. If an IP address had so many PTR records that the DNS lookup buffer
3980 was not large enough to hold them, Exim could crash while trying to process
3981 the truncated data. It now detects and logs this case.
3983 17. Further to 4.21/58, another change has been made: if (and only if) the
3984 first line of a message (the first header line) ends with CRLF, a bare LF
3985 in a subsequent header line has a space inserted after it, so as not to
3986 terminate the header.
3988 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3989 unnecessarily, used atoi() instead of strtol(), and didn't check the
3990 termination when getting file sizes from file names by regex.
3992 19. Completely re-implemented the support for maildirsize files, in the light
3993 of a number of problems with the previous contributed implementation
3994 (4.30/29). In particular:
3996 . If the quota is zero, the maildirsize file is maintained, but no quota is
3999 . If the maildir directory does not exist, it is created before any attempt
4000 to write a maildirsize file.
4002 . The quota value in the file is just a cache; if the quota is changed in
4003 the transport, the new value overrides.
4005 . A regular expression is available for excluding directories from the
4008 20. The autoreply transport checks the characters in options that define the
4009 message's headers; it allows continued headers, but it was checking with
4010 isspace() after an embedded newline instead of explicitly looking for a
4013 21. If all the "regular" hosts to which an address was routed had passed their
4014 expiry times, and had not reached their retry times, the address was
4015 bounced, even if fallback hosts were defined. Now Exim should go on to try
4018 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4019 equivalent code in the SMTP transport. Some hosts send humungous responses
4020 to HELO/EHLO, more than 1024 it seems.
4022 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4023 gives compiler warnings in some environments. I've now done it "properly",
4026 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4027 (because of problems with the built-in one) was declared to return uschar *
4028 instead of char *, causing compiler failure.
4030 25. Fixed a file descriptor leak when processing alias/forward files.
4032 26. Fixed a minor format string issue in dbfn.c.
4034 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4036 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4037 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4040 29. When a sender address is verified, it is cached, to save repeating the test
4041 when there is more than one recipient in a message. However, when the
4042 verification involves a callout, it is possible for different callout
4043 options to be set for different recipients. It is too complicated to keep
4044 track of this in the cache, so now Exim always runs a verification when a
4045 callout is required, relying on the callout cache for the optimization.
4046 The overhead is duplication of the address routing, but this should not be
4049 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4050 to be verified with callout=postmaster, and the main callout worked but the
4051 postmaster check failed, the verification correctly failed. However, if a
4052 subsequent RCPT command asked for sender verification *without* the
4053 postmaster check, incorrect caching caused this verification also to fail,
4056 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4057 it was not caching the DNS options (qualify_single, search_parents) that
4058 were used when the lookup failed. A subsequent lookup with different
4059 options therefore always gave the same answer, though there were cases
4060 where it should not have. (Example: a "domains = !$mx_any" option on a
4061 dnslookup router: the "domains" option is always processed without any
4062 widening, but the router might have qualify_single set.) Now Exim uses the
4063 cached value only when the same options are set.
4065 32. Added John Jetmore's "exipick" utility to the distribution.
4067 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4068 than a timeout (e.g. a certificate is required, and is not provided), an
4069 Exim server now closes the connection immediately. Previously it waited for
4070 the client to close - but if the client is SSL, it seems that they each
4071 wait for each other, leading to a delay before one of them times out.
4073 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4074 maintained 0.8.x compatibility because I don't think many are using it, and
4075 it is clearly obsolete.
4077 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4080 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4081 client certificate was expired. A simple patch fixes this, though I don't
4082 understand the full logic of why the verify callback is called multiple
4085 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4086 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4087 which causes problems with some clients (such as the Certicom SSL Plus
4088 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4089 disables the coutermeasure allowing Eudora to connect."
4091 38. Exim was not checking that a write() to a log file succeeded. This could
4092 lead to Bad Things if a log got too big, in particular if it hit a file
4093 size limit. Exim now panics and dies if it cannot write to a log file, just
4094 as it does if it cannot open a log file.
4096 39. Modified OS/Makefile-Linux so that it now contains
4098 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4100 The two -D definitions ensure that Exim is compiled with large file
4101 support, which makes it possible to handle log files that are bigger than
4104 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4105 instance) a domain was checked against a named list that involved a lookup,
4106 causing $domain_data to be set, then another domain was checked against the
4107 same list, then the first domain was re-checked, the value of $domain_data
4108 after the final check could be wrong. In particular, if the second check
4109 failed, it could be set empty. This bug probably also applied to
4112 41. The strip_trailing_dot option was not being applied to the address given
4113 with the -f command-line option.
4115 42. The code for reading a message's header from the spool was incrementing
4116 $received_count, but never initializing it. This meant that the value was
4117 incorrect (doubled) while delivering a message in the same process in which
4118 it was received. In the most common configuration of Exim, this never
4119 happens - a fresh exec is done - but it can happen when
4120 deliver_drop_privilege is set.
4122 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4123 it now includes up to 150 characters of the unexpected data in the log
4126 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4127 and building data strings. The size of both of these buffers was 10 000
4128 bytes - far larger than anybody would *ever* want, thought I. Needless to
4129 say, somebody hit the limit. I have increased the maximum line length to
4130 20 000 and the maximum data length of concatenated lines to 100 000. I have
4131 also fixed two bugs, because there was no checking on these buffers. Tsk,
4132 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4133 buffer is too small.
4135 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4136 lsearch lookups. Now it does.
4138 46. When parsing a route_list item in a manualroute router, a fixed-length
4139 buffer was used for the list of hosts. I made this 1024 bytes long,
4140 thinking that nobody would ever have a list of hosts that long. Wrong.
4141 Somebody had a whole pile of complicated expansion conditions, and the
4142 string was silently truncated, leading to an expansion error. It turns out
4143 that it is easier to change to an unlimited length (owing to other changes
4144 that have happened since this code was originally written) than to build
4145 structure for giving a limitation error. The length of the item that
4146 expands into the list of hosts is now unlimited.
4148 47. The lsearch lookup could not handle data where the length of text line was
4149 more than 4095 characters. Such lines were truncated, leading to shortened
4150 data being returned. It should now handle lines of any length.
4152 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4153 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4156 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4157 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4158 no longer happens in comment lines. A list of replacements is now placed
4159 at the head of all of the source files, except those whose only change is
4160 to replace PERL_COMMAND in the very first #! line.
4162 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4163 messages on the queue, with a bottom-up merge sort, using code contributed
4164 by Michael Haardt. This should make operations like -bp somewhat faster on
4165 large queues. It won't affect queue runners, except when queue_run_in_order
4168 51. Installed eximstats 1.31 in the distribution.
4170 52. Added support for SRV lookups to the dnslookup router.
4172 53. If an ACL referred to $message_body or $message_body_end, the value was not
4173 reset for any messages that followed in the same SMTP session.
4175 54. The store-handling optimization for building very long strings was not
4176 differentiating between the different store pools. I don't think this
4177 actually made any difference in practice, but I've tidied it.
4179 55. While running the routers to verify a sender address, $sender_address
4180 was still set to the sender address. This is wrong, because when routing to
4181 send a bounce to the sender, it would be empty. Therefore, I have changed
4182 it so that, while verifying a sender address, $sender_address is set to <>.
4183 (There is no change to what happens when verifying a recipient address.)
4185 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4186 target A or AAAA records (if not already returned) without resetting the
4187 qualify_single or search_parents options of the DNS resolver. These are
4188 inappropriate in this case because the targets of MX and SRV records must
4189 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4190 target that, when qualified, matched something in the local domain. These
4191 two options are now turned off when doing these lookups.
4193 57. It seems that at least some releases of Reiserfs (which does not have the
4194 concept of a fixed number of inodes) returns zero and not -1 for the
4195 number of available inodes. This interacted badly with check_spool_inodes,
4196 which assumed that -1 was the "no such thing" setting. What I have done is
4197 to check that the total number of inodes is greater than zero before doing
4198 the test of how many are available.
4200 58. When a "warn" ACL statement has a log_message modifier, the message is
4201 remembered, and not repeated. This is to avoid a lot of repetition when a
4202 message has many recipients that cause the same warning to be written.
4203 Howewer, Exim was preserving the list of already written lines for an
4204 entire SMTP session, which doesn't seem right. The memory is now reset if a
4205 new message is started.
4207 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4208 debugging output unless log_rewrite was also set.
4210 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4211 of (int)(handle) when we know that handle contains (void *)(-1).
4213 61. The Exim daemon panic-logs an error return when it closes the incoming
4214 connection. However "connection reset by peer" seems to be common, and
4215 isn't really an error worthy of noting specially, so that particular error
4218 62. When Exim is trying to find all the local interfaces, it used to panic and
4219 die if the ioctl to get the interface flags failed. However, it seems that
4220 on at least one OS (Solaris 9) it is possible to have an interface that is
4221 included in the list of interfaces, but for which you get a failure error
4222 for this call. This happens when the interface is not "plumbed" into a
4223 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4224 failure of the "get flags" call assumes that the interface is down.
4226 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4227 makes life easier for people who are doing arithmetic on fields extracted
4228 from dates, where you often get leading zeros that should not be
4229 interpreted as octal.
4231 64. Added qualify_domain to the redirect router, to override the global
4234 65. If a pathologically long header line contained very many addresses (the
4235 report of this problem mentioned 10 000) and each of them was rewritten,
4236 Exim could use up a very large amount of memory. (It kept on making new
4237 copies of the header line as it rewrote, and never released the old ones.)
4238 At the expense of a bit more processing, the header rewriting function has
4239 been changed so that it no longer eats memory in this way.
4241 66. The generation of the Received: header has been moved from the time that a
4242 message starts to be received, to the time that it finishes. The timestamp
4243 in the Received: header should now be very close to that of the <= log
4244 line. There are two side-effects of this change:
4246 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4247 logged header lines no longer include the local Received: line, because
4248 it has not yet been created. The same applies to a copy of the message
4249 that is returned to a non-SMTP sender when a message is rejected.
4251 (b) When a filter file is tested using -bf, no additional Received: header
4252 is added to the test message. After some thought, I decided that this
4255 This change does not affect the value of $received_for. It is still set
4256 after address rewriting, but before local_scan() is called.
4258 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4260 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4261 gave an unhelpful panic error message, and a defer error. I have managed to
4262 change this behaviour so that it now rejects any supplied certificate,
4263 which seems right, as the list of acceptable certificates is empty.
4265 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4266 gave an unhelpful defer error. I have not managed to make this reject any
4267 supplied certificates, but the error message it gives is "no certificate
4268 supplied", which is not helpful.
4270 70. exigrep's output now also includes lines that are not associated with any
4271 message, but which match the given pattern. Implemented by a patch from
4272 Martin Sluka, which also tidied up the Perl a bit.
4274 71. Recipient callout verification, like sender verification, was using <> in
4275 the MAIL FROM command. This isn't really the right thing, since the actual
4276 sender may affect whether the remote host accepts the recipient or not. I
4277 have changed it to use the actual sender in the callout; this means that
4278 the cache record is now keyed on a recipient/sender pair, not just the
4279 recipient address. There doesn't seem to be a real danger of callout loops,
4280 since a callout by the remote host to check the sender would use <>.
4281 [SEE ABOVE: changed after hitting problems.]
4283 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4284 temporary errors. However, in the case of such a code being given after
4285 the end of a data transmission (i.e. after ".") Exim was failing to write
4286 a retry record for the message. (Yes, there was some broken host that was
4287 actually sending 8xx at this point.)
4289 73. An unknown lookup type in a host list could cause Exim to panic-die when
4290 the list was checked. (An example that provoked this was putting <; in the
4291 middle of a list instead of at the start.) If this happened during a DATA
4292 ACL check, a -D file could be left lying around. This kind of configuration
4293 error no longer causes Exim to die; instead it causes a defer errror. The
4294 incident is still logged to the main and panic logs.
4296 74. Buglet left over from Exim 3 conversion. The message "too many messages
4297 in one connection" was written to the rejectlog but not the mainlog, except
4298 when address rewriting (yes!) was being logged.
4300 75. Added write_rejectlog option.
4302 76. When a system filter was run not as root (that is, when system_filter_user
4303 was set), the values of the $n variables were not being returned to the
4304 main process; thus, they were not subsequently available in the $sn
4307 77. Added +return_path_on_delivery log selector.
4309 78. A connection timeout was being treated differently from recipients deferred
4310 when testing hosts_max_try with a message that was older than the host's
4311 retry timeout. (The host should not be counted, thus allowing all hosts to
4312 be tried at least once before bouncing.) This may have been the cause of an
4313 occasionally reported bug whereby a message would remain on the queue
4314 longer than the retry timeout, but would be bounced if a delivery was
4315 forced. I say "may" because I never totally pinned down the problem;
4316 setting up timeout/retry tests is difficult. See also the next item.
4318 79. The ultimate address timeout was not being applied to errors that involved
4319 a combination of host plus message (for example, a timeout on a MAIL
4320 command). When an address resolved to a number of possible hosts, and they
4321 were not all tried for each delivery (e.g. because of hosts_max_try), a
4322 message could remain on the queue longer than the retry timeout.
4324 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4327 81. Fixed an obscure SMTP outgoing bug which required at least the following
4328 conditions: (a) there was another message waiting for the same server;
4329 (b) the server returned 5xx to all RCPT commands in the first message so
4330 that the message was not completed; (c) the server dropped the connection
4331 or gave a negative response to the RSET that Exim sends to abort the
4332 transaction. The observed case was a dropped connection after DATA that had
4333 been sent in pipelining mode. That is, the server had advertised PIPELINING
4334 but was not implementing it correctly. The effect of the bug was incorrect
4335 behaviour, such as trying another host, and this could lead to a crash.
4341 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4342 and daemon.c were passed as pointers to ints; they should have been
4343 pointers to socklen_t variables (which are typically unsigned ints).
4345 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4348 3. Fixed a really odd bug that affected only the testing scheme; patching a
4349 certain fixed string in the binary changed the value of another string that
4350 happened to be identical to the end of the original first string.
4352 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4353 name", it returns that address as the IP address. On some operating
4354 systems (e.g. Solaris), it also passes back the IP address string as the
4355 "host name". However, on others (e.g. Linux), it passes back an empty
4356 string. Exim wasn't checking for this, and was changing the host name to an
4357 empty string, assuming it had been canonicized.
4359 5. Although rare, it is permitted to have more than one PTR record for a given
4360 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4361 all the names associated with an address, because they do in Solaris.
4362 However, it seems that they do not in Linux for data that comes from the
4363 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4364 I found this out when I moved to a new Linux workstation and tried to run
4365 the Exim test suite.
4367 To get round this problem I have changed the code so that it now does its
4368 own call to the DNS to look up PTR records when searching for a host name.
4369 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4370 addresses that are only in /etc/hosts are still found.
4372 This behaviour is, however, controlled by an option called host_lookup_
4373 order, which defaults to "bydns:byaddr". If people want to use the other
4374 order, or indeed, just use one or the other means of lookup, they can
4375 specify it in this variable.
4377 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4378 some operating systems, this comes back from gethostbyaddr() as an empty
4379 string, and this is what Exim used to test for. However, it seems that in
4380 other systems, "." is yielded. Exim now tests for this case too.
4382 7. The values of check_spool_space and check_log_space are now held internally
4383 as a number of kilobytes instead of an absolute number of bytes. If a
4384 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4385 kilobyte. This means that much larger values can be stored.
4387 8. Exim monitor: an attempt to get the action menu when not actually pointing
4388 at a message produces an empty menu entitled "No message selected". This
4389 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4390 no entries in it ("Shell widget menu has zero width and/or height"). So I
4391 have added a single, blank menu entry in this case.
4393 9. Added ${quote_local_part.
4395 10. MIME decoding is now applied to the contents of Subject: header lines when
4398 11. Now that a reference to $sender_host_address automatically causes a reverse
4399 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4400 host lookup before query-style lookups in lists that might use this
4401 variable. This has therefore been abolished, and the "net-" prefix is no
4402 longer necessary for query-style lookups.
4404 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4405 have been a typo for LFLAGS, so it has been changed.
4407 13. The install script calls Exim with "-C /dev/null" in order to find the
4408 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4409 to be output. Howeve, since Exim outputs its version number before the
4410 error, it didn't break the script. It just looked ugly. I fixed this by
4411 always allowing "-C /dev/null" if the caller is root.
4413 14. Ignore overlarge ACL variable number when reading spool file - insurance
4414 against a later release with more variables having written the file.
4416 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4417 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4418 this, as well as the form without the "IPv6" on the front.
4420 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4421 OS/Makefile-Darwin file.
4423 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4424 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4426 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4427 variable. However, some OS use ints for inodes. I've added cast to long int
4428 to get rid of the compiler warning.
4430 19. I had forgotten to lock out "/../" in configuration file names when
4431 ALT_CONFIG_PREFIX was set.
4433 20. Routers used for verification do not need to specify transports. However,
4434 if such a router generated a host list, and callout was configured, Exim
4435 crashed, because it could not find a port number from the (non-existent)
4436 transport. It now assumes port 25 in this circumstance.
4438 21. Added the -t option to exigrep.
4440 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4441 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4442 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4443 them is set, LOOKUP_LSEARCH is forced.
4445 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4447 24. Added sender and host information to the "rejected by local_scan()" log
4448 line; previously there was no indication of these.
4450 25. Added .include_if_exists.
4452 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4453 Linux. It turns out that not all file systems support this. Apparently some
4454 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4455 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4456 not supported on the file descriptor, is now ignored when Exim is trying to
4457 sync a directory. This applies only to Linux.
4459 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4461 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4462 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4465 29. Added support for maildirsize files from supplied patch (modified a bit).
4467 30. The use of :fail: followed by an empty string could lead Exim to respond to
4468 sender verification failures with (e.g.):
4470 550 Verification failed for <xxx>
4471 550 Sender verify failed
4473 where the first response line was missing the '-' that indicates it is not
4474 the final line of the response.
4476 31. The loop for finding the name of the user that called Exim had a hardwired
4477 limit of 10; it now uses the value of finduser_retries, which is used for
4478 all other user lookups.
4480 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4483 33. Exim was neglecting to zero errno before one call of strtol() when
4484 expanding a string and expecting an integer value. On some systems this
4485 resulted in spurious "integer overflow" errors. Also, it was casting the
4486 result into an int without checking.
4488 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4489 did not work. The code looks as if it has *never* worked, though it appears
4490 to have been documented since at least releast 1.62. I have made it work.
4492 35. The "timeout_DNS" error in retry rules, also documented since at least
4493 1.62, also never worked. As it isn't clear exactly what this means, and
4494 clearly it isn't a major issue, I have abolished the feature by treating it
4495 as "timeout", and writing a warning to the main and panic logs.
4497 36. The display of retry rules for -brt wasn't always showing the error code
4500 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4501 timeout_connect_A, timeout_connect_MX.
4503 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4504 to the empty sender.
4506 39. The daemon was not analysing the content of -oX till after it had closed
4507 stderr and disconnected from the controlling terminal. This meant that any
4508 syntax errors were only noted on the panic log, and the return code from
4509 the command was 0. By re-arranging the code a little, I've made the
4510 decoding happen first, so such errors now appear on stderr, and the return
4511 code is 1. However, the actual setting up of the sockets still happens in
4512 the disconnected process, so errors there are still only recorded on the
4515 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4516 connections (as happens on some IP stacks) was logged at start up time as
4517 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4518 it from "IPv6 and IPv4", which means that two separate sockets are being
4521 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4522 says whether AF_INET or AF_INET6 was passed as an argument.
4524 42. Exiwhat output was messed up when time zones were included in log
4527 43. Exiwhat now gives more information about the daemon's listening ports,
4528 and whether -tls-on-connect was used.
4530 44. The "port" option of the smtp transport is now expanded.
4532 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4533 silently ignored. Now an error message is written to the main and panic
4536 46. There's a new ACL modifier called "logwrite" which writes to a log file
4537 as soon as it is encountered.
4539 47. Added $local_user_uid and $local_user_gid at routing time.
4541 48. Exim crashed when trying to verify a sender address that was being
4544 49. Exim was recognizing only a space character after ".include". It now also
4545 recognizes a tab character.
4547 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4548 extracting the relevant information from the specification. The man page no
4549 longer contains scrambled data for the -d option, and I've added a section
4550 at the front about calling Exim under different names.
4552 51. Added "extra_headers" argument to the "mail" command in filter files.
4554 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4557 53. Installed eximstats 1.29.
4559 54. Added transport_filter_timeout as a generic transport option.
4561 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4562 Cc: header lines. This was required by RFC 822, but it not required by RFC
4565 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4566 incoming messages that did not have them. Now it does so only if the
4567 message originates locally, that is, if there is no associated remote host
4568 address. When Resent- header lines are present, this applies to the Resent-
4569 lines rather than the non-Resent- lines.
4571 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4572 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4574 58. Messages for configuration errors now include the name of the main
4575 configuration file - useful now that there may be more than one file in a
4576 list (.included file names were always shown).
4578 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4579 for those rare installations that do not start the daemon as root or run it
4580 setuid root. I've cut out the call to initgroups() if the daemon is not
4583 60. The Exim user and group can now be bound into the binary as text strings
4584 that are looked up at the start of Exim's processing.
4586 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4588 62. Added $mailstore_basename variable.
4590 63. Installed patch to sieve.c from Michael Haardt.
4592 64. When Exim failed to open the panic log after failing to open the main log,
4593 the original message it was trying to log was written to stderr and debug
4594 output, but if they were not available (the usual case in production), it
4595 was lost. Now it is written to syslog before the two lines that record the
4596 failures to open the logs.
4598 65. Users' Exim filters run in subprocesses under the user's uid. It is
4599 possible for a "deliver" command or an alias in a "personal" command to
4600 provoke an address rewrite. If logging of address rewriting is configured,
4601 this fails because the process is not running as root or exim. There may be
4602 a better way of dealing with this, but for the moment (because 4.30 needs
4603 to be released), I have disabled address rewrite logging when running a
4604 filter in a non-root, non-exim process.
4610 1. The buildconfig auxiliary program wasn't quoting the value set for
4611 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4612 not defined. This bug was masked in 4.22 by the effect that was fixed in
4615 2. Some messages that were rejected after a message id was allocated were
4616 shown as "incomplete" by exigrep. It no longer does this for messages that
4617 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4619 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4620 have allow_domain_literals set, the ID did not get logged in the <= line.
4621 Domain literals are now always recognized in Message-ID: header lines.
4623 4. The first argument for a ${extract expansion item is the key name or field
4624 number. Leading and trailing spaces in this item were not being ignored,
4625 causing some misleading effects.
4627 5. When deliver_drop_privilege was set, single queue runner processes started
4628 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4629 same command in the process it spins off) were not dropping privilege.
4631 6. When the daemon running as "exim" started a queue runner, it always
4632 re-executed Exim in the spun-off process. This is a waste of effort when
4633 deliver_drop_privilege is set. The new process now just calls the
4634 queue-runner function directly.
4640 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4643 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4644 ignored. Though the use of -oP was forcing the writing of a pid file, it
4645 was always written to the default place.
4647 3. If the message "no IP address found for host xxxx" is generated during
4648 incoming verification, it is now followed by identification of the incoming
4649 connection (so you can more easily find what provoked it).
4651 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4653 5. Added some features to "harden" Exim a bit more against certain attacks:
4655 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4656 be put in Local/Makefile. This is like the never_users runtime option,
4657 but it cannot be overridden. The default setting is "root".
4659 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4660 prefix string with which any file named in a -C command line option
4663 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4664 is retained for -C and -D only if the caller of Exim is root. Without
4665 it, the exim user may also use -C and -D and retain privilege.
4667 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4668 command line option is disabled.
4670 6. Macro names set by the -D option must start with an upper case letter, just
4671 like macro names defined in the configuration file.
4673 7. Added "dereference=" facility to LDAP.
4675 8. Two instances of the typo "uknown" in the source files are fixed.
4677 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4678 the Configure-Makefile script screwed up while processing it.
4680 10. Incorporated PCRE 4.4.
4682 11. The SMTP synchronization check was not operating right at the start of an
4683 SMTP session. For example, it could not catch a HELO sent before the client
4684 waited for the greeting. There is now a check for outstanding input at the
4685 point when the greeting is written. Because of the duplex, asynchronous
4686 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4687 way, but not yet received, when the check is performed.
4689 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4690 on TCP/IP sockets, because this apparently causes some broken clients to
4693 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4694 unchanged) from the Cygwin maintainer.
4696 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4697 was supported instead of testing for mbx. Effectively a typo.
4699 15. The spa authenticator server code was not checking that the input it
4700 received was valid base64.
4702 16. The debug output line for the "set" modifier in ACLs was not showing the
4703 name of the variable that was being set.
4705 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4707 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4708 lookup on its own. Something else was needed to trigger the lookup. For
4709 example, a match in host_lookup or the need for a host name in a host list.
4710 Now, if $sender_host_name is referenced and the host name has not yet been
4711 looked up, a lookup is performed. If the lookup fails, the variable remains
4712 empty, and $host_lookup_failed is set to "1".
4714 19. Added "eqi" as a case-independent comparison operator.
4716 20. The saslauthd authentication condition could segfault if neither service
4717 nor realm was specified.
4719 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4720 error message that was logged was misleading, and incoming SMTP
4721 connections were dropped. The message is now more accurate, and temporary
4722 errors are given to SMTP connections.
4724 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4725 (except RSET) with a 421 error, until QUIT is received. However, it was
4726 failing to send a response to QUIT.
4728 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4729 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4730 ACL is now run after the helo_try_verify_hosts code.
4732 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4733 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4734 case-independent, so other case variants are also recognized). Apparently
4735 some systems use these upper case variants.
4737 25. If more than two messages were waiting for the same host, and a transport
4738 filter was specified for the transport, Exim sent two messages over the
4739 same TCP/IP connection, and then failed with "socket operation on non-
4740 socket" when it tried to send the third.
4742 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4744 27. The extern definition of crypt16() in expand.c was not being excluded when
4745 the OS had its own crypt16() function.
4747 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4748 as a preferred synonym for return_size_limit, both as an option and as an
4751 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4753 30. Changed the default configuration ACL to relax the local part checking rule
4754 for addresses that are not in any local domains. For these addresses,
4755 slashes and pipe symbols are allowed within local parts, but the sequence
4756 /../ is explicitly forbidden.
4758 31. SPA server authentication was not clearing the challenge buffer before
4761 32. log_message in a "warn" ACL statement was writing to the reject log as
4762 well as to the main log, which contradicts the documentation and doesn't
4763 seem right (because no rejection is happening). So I have stopped it.
4765 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4766 However, I am unable to do any testing of this.
4768 34. Fixed an infelicity in the appendfile transport. When checking directories
4769 for a mailbox, to see if any needed to be created, it was accidentally
4770 using path names with one or more superfluous leading slashes; tracing
4771 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4773 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4774 contents are added to the log line that is written for every discarded
4775 recipient. (Previously a log_message setting was ignored.)
4777 36. The ${quote: operator now quotes the string if it is empty.
4779 37. The install script runs exim in order to find its version number. If for
4780 some reason other than non-existence or emptiness, which it checks, it
4781 could not run './exim', it was installing it with an empty version number,
4782 i.e. as "exim-". This error state is now caught, and the installation is
4785 38. An argument was missing from the function that creates an error message
4786 when Exim fails to connect to the socket for saslauthd authentication.
4787 This could cause Exim to crash, or give a corrupted message.
4789 39. Added isip, isip4, and isip6 to ${if conditions.
4791 40. The ACL variables $acl_xx are now saved with the message, and can be
4792 accessed later in routers, transports, and filters.
4794 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4795 strings in the file are not string-expanded.
4797 42. If a MAIL command specified a SIZE value that was too large to fit into an
4798 int variable, the check against message_size_limit failed. Such values are
4799 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4800 day this will have to be increased, but I don't think I want to be around
4801 when emails are that large.
4808 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4809 iconv() is not standard in FreeBSD.
4811 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4812 IPv6 enabled. The observed symptom was a segmentation fault on return from
4813 the function os_common_find_running_interfaces() in src/os.c.
4815 3. In the check_special_case() function in daemon.c I had used "errno" as an
4816 argument name, which causes warnings on some systems. This was basically a
4817 typo, since it was named "eno" in the comments!
4819 4. The code that waits for the clock to tick (at a resolution of some fraction
4820 of a second) so as to ensure message-id uniqueness was always waiting for
4821 at least one whole tick, when it could have waited for less. [This is
4822 almost certainly not relevant at current processor speeds, where it is
4823 unlikely to ever wait at all. But we try to future-proof.]
4825 5. The function that sleeps for a time interval that includes fractions of a
4826 second contained a race. It did not block SIGALRM between setting the
4827 timer, and suspending (a couple of lines later). If the interval was short
4828 and the sigsuspend() was delayed until after it had expired, the suspension
4829 never ended. On busy systems this could lead to processes getting stuck for
4832 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4833 process, before it forks any delivery processes. The open lookup caching
4834 mechanism meant that the open file or database connection was passed into
4835 the delivery process. The problem was that delivery processes always tidy
4836 up cached lookup data. This could cause a problem for the next delivery
4837 process started by the queue runner, because the external queue runner
4838 process does not know about the closure. So the next delivery process
4839 still has data in the lookup cache. In the case of a file lookup, there was
4840 no problem because closing a file descriptor in a subprocess doesn't affect
4841 the parent. However, if the lookup was caching a connection to a database,
4842 the connection was closed, and the second delivery process was likely to
4843 see errors such as "PGSQL: query failed: server closed the connection
4844 unexpectedly". The problem has been fixed by closing all cached lookups
4845 in a queue runner before running a delivery process.
4847 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4848 seem to have the "const" qualifier which it has on other OS. I've
4851 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4852 *of the same type* (client or server) with the same public name that an
4853 error should be diagnosed.
4855 9. When Exim looked up a host name for an IP address, but failed to find the
4856 original IP address when looking up the host name (a safety check), it
4857 output the message "<ip address> does not match any IP for NULL", which was
4858 confusing, to say the least. The bug was that the host name should have
4859 appeared instead of "NULL".
4861 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4862 user that is built into the binary, and the -C or -D options is used, root
4863 privilege is dropped before the configuration file is read. In addition,
4864 logging is switched to stderr instead of the normal log files. If the
4865 configuration then re-defines the Exim user, the unprivileged environment
4866 is probably not what is expected, so Exim logs a panic warning message (but
4869 However, if deliver_drop_privilege is set, the unprivileged state may well
4870 be exactly what is intended, so the warning has been cut out in that case,
4871 and Exim is allowed to try to write to its normal log files.
4877 1. smtp_return_error_details was not giving details for temporary sender
4878 or receiver verification errors.
4880 2. Diagnose a configuration error if two authenticators have the same public
4883 3. Exim used not to create the message log file for a message until the first
4884 delivery attempt. This could be confusing when incoming messages were held
4885 for policy or load reasons. The message log file is now created at the time
4886 the message is received, and an initial "Received" line is written to it.
4888 4. The automatically generated man page for command line options had a minor
4889 bug that caused no ill effects; however, a more serious problem was that
4890 the procedure for building the man page automatically didn't always
4891 operate. Consequently, release 4.20 contains an out-of-date version. This
4892 shouldn't happen again.
4894 5. When building Exim with embedded Perl support, the script that builds the
4895 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4896 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4898 6. The freeze_tell option was not being used for messages that were frozen on
4899 arrival, either by an ACL or by local_scan().
4901 7. Added the smtp_incomplete_transaction log selector.
4903 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4904 was accepting AUTH without a new EHLO.
4906 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4907 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4910 10. Logging of TCP/IP connections (when configured) now happens in the main
4911 daemon process instead of the child process, so that the TCP/IP connection
4912 count is more accurate (but it can never be perfect).
4914 11. The use of "drop" in a nested ACL was not being handled correctly in the
4915 outer ACL. Now, if condition failure induced by the nested "drop" causes
4916 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4917 or "require"), the connection is dropped.
4919 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4920 that yield "discard" can now be used with an "accept" or a "discard" verb,
4921 but an error is generated for any others (because I can't see a useful way
4922 to define what should happen).
4924 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4925 are now processed in the same way as lines in the Exim configuration file.
4926 In particular, continuation lines are supported.
4928 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4930 15. Added -ti meaning -t -i.
4932 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4933 domains, and warn by logging if others are found.
4935 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4936 returned from GIFCONF when Exim is trying to find the list of interfaces on
4937 a host. The code in os.c has been modified to copy each ifreq to an aligned
4938 structure in all cases.
4940 Also, in some cases, the returned ifreq's were being copied to a 'struct
4941 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4942 means the last couple of bytes of an IPv6 address could be chopped if the
4943 ifreq contained only a normal sockaddr (14 bytes storage).
4945 18. Named domain lists were not supported in the hosts_treat_as_local option.
4946 An entry such as +xxxx was not recognized, and was treated as a literal
4949 19. Ensure that header lines added by a DATA ACL are included in the reject log
4950 if the ACL subsequently rejects the message.
4952 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4953 MD5 (which is deprecated).
4955 21. When testing a filter file using -bf, Exim was writing a message when it
4956 took the sender from a "From " line in the message, but it was not doing so
4957 when it took $return_path from a Return-Path: header line. It now does.
4959 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4960 with a valid header line field name (a series of printing characters
4961 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4963 23. Changed "disc" in the source to "disk" to conform to the documentation and
4964 the book and for uniformity.
4966 24. Ignore Sendmail's -Ooption=value command line item.
4968 25. When execve() failed while trying to run a command in a pipe transport,
4969 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4970 could be confused with a return value of 69 from the command itself. This
4971 has been changed to 127, the value the shell returns if it is asked to run
4972 a non-existent command. The wording for the related log line suggests a
4973 non-existent command as the problem.
4975 26. If received_header_text expands to an empty string, do not add a Received:
4976 header line to the message. (Well, it adds a token one on the spool, but
4977 marks it "old" so that it doesn't get used or transmitted.)
4979 27. Installed eximstats 1.28 (addition of -nt option).
4981 28. There was no check for failure on the call to getsockname() in the daemon
4982 code. This can fail if there is a shortage of resources on the system, with
4983 ENOMEM, for example. A temporary error is now given on failure.
4985 29. Contrary to the C standard, it seems that in some environments, the
4986 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4987 program. Exim now does this explicitly; it affects the formatting of
4988 timestamps using strftime().
4990 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4991 complaints. I've now initialized all the variables, to avoid this.
4993 32. Header lines added by a system filter were not being "seen" during
4994 transport-time rewrites.
4996 33. The info_callback() function passed to OpenSSL is set up with type void
4997 (*)(SSL *, int, int), as described somewhere. However, when calling the
4998 function (actually a macro) that sets it up, the type void(*)() is
4999 expected. I've put in a cast to prevent warnings from picky compilers.
5001 34. If a DNS black list lookup found a CNAME record, but there were no A
5002 records associated with the domain it pointed at, Exim crashed.
5004 35. If a DNS black list lookup returned more than one A record, Exim ignored
5005 all but the first. It now scans all returned addresses if a particular IP
5006 value is being sought. In this situation, the contents of the
5007 $dnslist_value variable are a list of all the addresses, separated by a
5010 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5011 to accept a host name and all its aliases if the forward lookup for any of
5012 them yielded the IP address of the incoming connection. Now it accepts only
5013 those names whose forward lookup yields the correct IP address. Any other
5014 names are discarded. This closes a loophole whereby a rogue DNS
5015 administrator could create reverse DNS records to break through a
5016 wildcarded host restriction in an ACL.
5018 37. If a user filter or a system filter that ran in a subprocess used any of
5019 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5020 the wrong values were passed to the pipe command ($thisaddress had the
5021 value of $0, $0 had the value of $1, etc). This bug was introduced by
5022 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5024 38. Improved the line breaking for long SMTP error messages from ACLs.
5025 Previously, if there was no break point between 40 and 75 characters, Exim
5026 left the rest of the message alone. Two changes have been made: (a) I've
5027 reduced the minimum length to 35 characters; (b) if it can't find a break
5028 point between 35 and 75 characters, it looks ahead and uses the first one
5029 that it finds. This may give the occasional overlong line, but at least the
5030 remaining text gets split now.
5032 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5033 file descriptors might be low, and that setting 1000 would always raise it.
5034 It turns out that in some environments, the limit is already over 1000 and
5035 that lowering it causes trouble. So now Exim takes care not to decrease it.
5037 40. When delivering a message, the value of $return_path is set to $sender_
5038 address at the start of routing (routers may change the value). By an
5039 oversight, this default was not being set up when an address was tested by
5040 -bt or -bv, which affected the outcome if any router or filter referred to
5043 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5044 log only when "message" or "log_message" are set. However, if one of the
5045 conditions was an address verification, or a call to a nested ACL, the
5046 messages generated by the underlying test were being passed through. This
5047 no longer happens. The underlying message is available in $acl_verify_
5048 message for both "message" and "log_message" expansions, so it can be
5049 passed through if needed.
5051 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5052 new expansion $bh_ to give the encoded byte string without charset
5053 translation. Translation happens only if iconv() is available; HAVE_ICONV
5054 indicates this at build time. HEADERS_CHARSET gives the charset to
5055 translate to; headers_charset can change it in the configuration, and
5056 "headers charset" can change it in an individual filter file.
5058 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5059 that creates RFC 2047 encoded "words" labels them as that charset instead
5060 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5061 expansion operator; (ii) when Exim creates a From: line for a local
5062 message; (iii) when a header line is rewritten to include a "phrase" part.
5064 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5065 buggy, causing it to skip the first lines of messages whose message ID
5066 ended in 'D'. This would not have bitten before Exim release 4.14, because
5067 message IDs were unlikely to end in 'D' before then. The effect was to have
5068 incorrect size information for certain domains.
5070 45. #include "config.h" was missing at the start of the crypt16.c module. This
5071 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5074 46. If there was a timeout during a "random" callout check, Exim treated it as
5075 a failure of the random address, and carried on sending RSET and the real
5076 address. If the delay was just some slowness somewhere, the response to the
5077 original RCPT would be taken as a response to RSET and so on, causing
5078 mayhem of various kinds.
5080 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5081 when I implemented it. It didn't allow for the fact that some option values
5082 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5083 the right test for positive values.
5085 48. Domain names in DNS records are case-independent. Exim always looks them up
5086 in lower case. Some resolvers return domain names in exactly the case they
5087 appear in the zone file, that is, they may contain uppercase letters. Not
5088 all resolvers do this - some return always lower case. Exim was treating a
5089 change of case by a resolver as a change of domain, similar to a widening
5090 of a domain abbreviation. This triggered its re-routing code and so it was
5091 trying to route what was effectively the same domain again. This normally
5092 caused routing to fail (because the router wouldn't handle the domain
5093 twice). Now Exim checks for this case specially, and just changes the
5094 casing of the domain that it ultimately uses when it transmits the message
5097 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5100 50. If a filter generated a file delivery with a non-absolute name (possible if
5101 no home directory exists for the router), the forbid_file option was not
5104 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5105 the existing equality matching.
5107 52. Exim was using ints instead of ino_t variables in some places where it was
5108 dealing with inode numbers.
5110 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5111 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5112 called TMPDIR, and if it finds it is different, it changes its value.
5114 54. The smtp_printf() function is now made available to local_scan() so
5115 additional output lines can be written before returning. There is also an
5116 smtp_fflush() function to enable the detection of a dropped connection.
5117 The variables smtp_input and smtp_batched_input are exported to
5120 55. Changed the default runtime configuration: the message "Unknown user"
5121 has been removed from the ACL, and instead placed on the localuser router,
5122 using the cannot_route_message feature. This means that any verification
5123 failures that generate their own messages won't get overridden. Similarly,
5124 the "Unrouteable address" message that was in the ACL for unverifiable
5125 relay addresses has also been removed.
5127 56. Added hosts_avoid_esmtp to the smtp transport.
5129 57. The exicyclog script was not checking for the esoteric option
5130 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5131 will work only if exicyclog is run under the appropriate euid.
5133 58. Following a discussion on the list, the rules by which Exim recognises line
5134 endings on incoming messages have been changed. The -dropcr and drop_cr
5135 options are now no-ops, retained only for backwards compatibility. The
5136 following line terminators are recognized: LF CRLF CR. However, special
5137 processing applies to CR:
5139 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5140 nor a local message in the state where . is a terminator.
5142 (ii) If a bare CR is encountered in a header line, an extra space is added
5143 after the line terminator so as not to end the header. The reasoning
5144 behind this is that bare CRs in header lines are most likely either
5145 to be mistakes, or people trying to play silly games.
5147 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5148 was being incorrectly given as 18 bytes larger than it should have been.
5149 This is a VOB (very old bug).
5151 60. This may never have affected anything current, but just in case it has:
5152 When the local host is found other than at the start of a list of hosts,
5153 the local host, those with the same MX, and any that follow, are discarded.
5154 When the list in question was part of a longer list of hosts, the following
5155 hosts (not currently being processed) were also being discarded. This no
5156 longer happens. I'm not sure if this situation could ever has previously
5159 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5162 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5163 header. This is something that is recommended in a new Internet Draft, and
5164 is something that is documented as being done by Sendmail. There are two
5165 possible values. For messages generated by the autoreply transport, Exim
5168 Auto-Submitted: auto-replied
5170 whereas for all other generated messages (e.g. bounces) it adds
5172 Auto-Submitted: auto-generated
5174 63. The "personal" condition in filters now includes a test for the
5175 Auto-Submitted: header. If it contains the string "auto-" the message it
5176 not considered personal.
5178 64. Added rcpt_include_affixes as a generic transport option.
5180 65. Added queue_only_override (default true).
5182 66. Added the syslog_duplication option.
5184 67. If what should have been the first header line of a message consisted of
5185 a space followed by a colon, Exim was mis-interpreting it as a header line.
5186 It isn't of course - it is syntactically invalid and should therefore be
5187 treated as the start of the message body. The misbehaviour could have
5188 caused a number of strange effects, including loss of data in subsequent
5189 header lines, and spool format errors.
5191 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5192 client host had authenticated. This control can now be exercised by an ACL
5193 for more flexibility.
5195 69. By default, callouts do not happen when testing with -bh. There is now a
5196 variant, -bhc, which does actually run the callout code, including
5197 consulting and updating the callout cache.
5199 70. Added support for saslauthd authentication, courtesy of Alexander
5202 71. If statvfs() failed on the spool or log directories while checking their
5203 size for availability, Exim confusingly gave the error "space shortage".
5204 Furthermore, in debugging mode it crashed with a floating point exception.
5205 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5206 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5207 serious problem, Exim now writes to the main and panic logs when this
5208 happens, with details of the failure. It then refuses to accept the
5209 incoming message, giving the message "spool directory problem" or "log
5210 directory problem" with a 421 code for SMTP messages.
5212 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5213 0, 1, and 2 exist, because some OS complain for execs without them (see
5214 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5215 descriptors. However, the code omitted to check that the open succeeded,
5216 causing mysterious errors if for some reason the permissions on /dev/null
5217 got screwed. Now Exim writes a message to the main and panic logs, and
5218 bombs out if it can't open /dev/null.
5220 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5221 interact so that it is all more flexible. It is supposed to remain
5222 backwards compatible. Also added extra_local_interfaces.
5224 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5225 to bomb out with an assertion failure - to the client this appears as a
5226 connection drop. This problem occurs in the part of the code that was taken
5227 from the Samba project. Fortunately, the assertion is in a very simple
5228 function, so I have fixed this by reproducing the function inline in the
5229 one place where it is called, and arranging for authentication to fail
5230 instead of killing the process with assert().
5232 75. The SPA client code was not working when the server requested OEM rather
5233 than Unicode encoding.
5235 76. Added code to make require_files with a specific uid setting more usable in
5236 the case where statting the file as root fails - usually a non-root-mounted
5237 NFS file system. When this happens and the failure is EACCES, Exim now
5238 forks a subprocess and does the per-uid checking as the relevant uid.
5240 77. Added process_log_path.
5242 78. If log_file_path was not explicitly set, a setting of check_log_space or
5243 check_log_inodes was ignored.
5245 79. If a space check for the spool or log partitions fails, the incident is now
5246 logged. Of course, in the latter case the data may get lost...
5248 80. Added the %p formatting code to string_format() so that it can be used to
5249 print addresses in debug_print(). Adjusted all the address printing in the
5250 debugging in store.c to use %p rather than %d.
5252 81. There was a concern that a line of code in smtp_in.c could overflow a
5253 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5254 initially expressed, the concern was not well-founded, because trailing
5255 spaces are removed early. However, if the trailing spaces were followed by
5256 a NULL, they did not get removed, so the overflow was possible. Two fixes
5259 (a) I re-wrote the offending code in a cleaner fashion.
5260 (b) If an incoming SMTP command contains a NULL character, it is rejected
5263 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5264 runs initgroups(), so that if the Exim user is in any additional groups,
5265 they will be used during message reception.
5271 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob