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 dependency in version reporting
12 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
13 This fixes portability to compilers other than gcc, notably
14 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
16 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
17 makefiles for portability to HP-UX and POSIX correctness.
19 PP/01 Permit LOOKUP_foo enabling on the make command-line.
20 Also via indented variable definition in the Makefile.
21 (Debugging by Oliver Heesakkers).
23 PP/02 Restore caching of spamd results with expanded spamd_address.
24 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
26 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
27 Improves build reliability. Fix from: Frank Elsner
29 NM/02 Fix wide character breakage in the rfc2047 coding
30 Fixes bug 1064. Patch from Andrey N. Oktyabrski
32 NM/03 Allow underscore in dnslist lookups
33 Fixes bug 1026. Patch from Graeme Fowler
35 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
36 Code patches from Adam Ciarcinski of NetBSD.
38 NM/04 Fixed exiqgrep to cope with mailq missing size issue
41 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
42 is logged, to avoid truncation. Patch from John Horne.
44 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
45 Patch from Jakob Hirsch.
47 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
48 SQL string expansion failure details.
49 Patch from Andrey Oktyabrski.
51 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
52 Patch from Simon Arlott.
54 PP/09 New lookups functionality failed to compile on old gcc which rejects
55 extern declarations in function scope.
56 Patch from Oliver Fleischmann
58 PP/10 Use sig_atomic_t for flags set from signal handlers.
59 Check getgroups() return and improve debugging.
60 Fixed developed for diagnosis in bug 927 (which turned out to be
63 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
64 Patch from Mark Zealey.
66 PP/12 Bugzilla 1056: Improved spamd server selection.
67 Patch from Mark Zealey.
69 PP/13 Bugzilla 1086: Deal with maildir quota file races.
70 Based on patch from Heiko Schlittermann.
72 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
73 Patch from Uwe Doering, sign-off by Michael Haardt.
75 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
76 variable declaration deep within a block. Bug and patch from
79 PP/15 lookups-Makefile IRIX compatibilty coercion.
81 PP/16 Make DISABLE_DKIM build knob functional.
87 TF/01 Failure to get a lock on a hints database can have serious
88 consequences so log it to the panic log.
90 TF/02 Log LMTP confirmation messages in the same way as SMTP,
91 controlled using the smtp_confirmation log selector.
93 TF/03 Include the error message when we fail to unlink a spool file.
95 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
96 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
97 for maintaining out-of-tree patches for some time.
99 PP/01 Bugzilla 139: Documentation and portability issues.
100 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
101 Handle per-OS dynamic-module compilation flags.
103 PP/02 Let /dev/null have normal permissions.
104 The 4.73 fixes were a little too stringent and complained about the
105 permissions on /dev/null. Exempt it from some checks.
106 Reported by Andreas M. Kirchwitz.
108 PP/03 Report version information for many libraries, including
109 Exim version information for dynamically loaded libraries. Created
110 version.h, now support a version extension string for distributors
111 who patch heavily. Dynamic module ABI change.
113 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
114 privilege escalation vulnerability whereby the Exim run-time user
115 can cause root to append content of the attacker's choosing to
118 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
121 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
122 If dropping privileges for untrusted macros, we disabled normal logging
123 on the basis that it would fail; for the Exim run-time user, this is not
124 the case, and it resulted in successful deliveries going unlogged.
125 Fixed. Reported by Andreas Metzler.
131 PP/01 Date: & Message-Id: revert to normally being appended to a message,
132 only prepend for the Resent-* case. Fixes regression introduced in
133 Exim 4.70 by NM/22 for Bugzilla 607.
135 PP/02 Include check_rfc2047_length in configure.default because we're seeing
136 increasing numbers of administrators be bitten by this.
138 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
140 PP/03 Bugzilla 994: added openssl_options main configuration option.
142 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
144 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
146 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
148 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
149 without a peer certificate, leading to a segfault because of an
150 assumption that peers always have certificates. Be a little more
151 paranoid. Problem reported by Martin Tscholak.
153 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
154 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
155 NB: ClamAV planning to remove STREAM in "middle of 2010".
156 CL also introduces -bmalware, various -d+acl logging additions and
157 more caution in buffer sizes.
159 PP/09 Implemented reverse_ip expansion operator.
161 PP/10 Bugzilla 937: provide a "debug" ACL control.
163 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
165 PP/12 Bugzilla 973: Implement --version.
167 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
169 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
171 PP/15 Bugzilla 816: support multiple condition rules on Routers.
173 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
174 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
175 ignore trailing whitespace.
177 JJ/02 prevent non-panic DKIM error from being sent to paniclog
179 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
182 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
183 Notification from Dr Andrew Aitchison.
185 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
186 ExtendedDetectionInfo response format.
187 Notification from John Horne.
189 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
192 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
193 XSL and documented dependency on system catalogs, with examples of how
196 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
199 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
200 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
201 configuration file which is writeable by the Exim user or group.
203 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
204 of configuration files to cover files specified with the -C option if
205 they are going to be used with root privileges, not just the default
208 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
209 option (effectively making it always true).
211 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
212 files to be used while preserving root privileges.
214 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
215 that rogue child processes cannot use them.
217 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
218 run-time user, instead of root.
220 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
221 Exim run-time user without dropping privileges.
223 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
224 result string, instead of calling string_vformat() twice with the same
227 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
228 for other users. Others should always drop root privileges if they use
229 -C on the command line, even for a whitelisted configure file.
231 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
233 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
239 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
240 $data_path, and $header_path variables; fixed documentation bugs and
243 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
244 exipick to access non-standard spools, including the "frozen" queue
247 NM/01 Bugzilla 965: Support mysql stored procedures.
248 Patch from Alain Williams
250 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
252 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
253 Patch from Andreas Metzler
255 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
256 Patch from Kirill Miazine
258 NM/05 Bugzilla 671: Added umask to procmail example.
260 JJ/03 installed exipick 20100323.0, fixing doc bug
262 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
263 directory. Notification and patch from Dan Rosenberg.
265 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
267 TK/02 Improve log output when DKIM signing operation fails.
269 MH/01 Treat the transport option dkim_domain as a colon separated
270 list, not as a single string, and sign the message with each element,
271 omitting multiple occurences of the same signer.
273 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
274 Bugzilla 985, 986. Patch by Simon Arlott
276 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
277 Patch by Simon Arlott
279 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
280 MBX locking. Notification from Dan Rosenberg.
286 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
288 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
290 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
292 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
294 NM/04 Bugzilla 915: Flag broken perl installation during build.
300 TK/01 Added patch by Johannes Berg that expands the main option
301 "spamd_address" if it starts with a dollar sign.
303 TK/02 Write list of recipients to X-Envelope-Sender header when building
304 the mbox-format spool file for content scanning (suggested by Jakob
307 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
308 (http://www.dcc-servers.net/) support via dccifd. Activated by
309 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
311 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
312 by Mark Daniel Reidel <mr@df.eu>.
314 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
315 When building exim an external PCRE library is now needed -
316 PCRE is a system library on the majority of modern systems.
317 See entry on PCRE_LIBS in EDITME file.
319 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
320 conversation. Added nologin parameter to request.
321 Patch contributed by Kirill Miazine.
323 TF/01 Do not log submission mode rewrites if they do not change the address.
325 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
327 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
328 log files in place. Contributed by Roberto Lima.
330 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
332 TF/03 Bugzilla 615: When checking the local_parts router precondition
333 after a local_part_suffix or local_part_prefix option, Exim now
334 does not use the address's named list lookup cache, since this
335 contains cached lookups for the whole local part.
337 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
338 Robert Millan. Documentation is in experimental-spec.txt.
340 TF/04 Bugzilla 668: Fix parallel build (make -j).
342 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
344 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
345 Patch provided by Jan Srzednicki.
347 TF/05 Leading white space used to be stripped from $spam_report which
348 wrecked the formatting. Now it is preserved.
350 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
351 that they are available at delivery time.
353 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
355 TF/08 TLS error reporting now respects the incoming_interface and
356 incoming_port log selectors.
358 TF/09 Produce a more useful error message if an SMTP transport's hosts
359 setting expands to an empty string.
361 NM/06 Bugzilla 744: EXPN did not work under TLS.
362 Patch provided by Phil Pennock.
364 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
365 Patch provided by Richard Godbee.
367 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
368 acl_smtp_notquit, added index entry.
370 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
371 Patch provided by Eugene Bujak.
373 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
374 accept(). Patch provided by Maxim Dounin.
376 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
377 Patch provided by Phil Pennock.
379 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
381 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
382 Patch provided by Brad "anomie" Jorsch.
384 NM/14 Bugzilla 622: Added timeout setting to transport filter.
385 Patch provided by Dean Brooks.
387 TK/05 Add native DKIM support (does not depend on external libraries).
389 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
390 Patch provided by Graeme Fowler.
392 NM/16 Bugzilla 851: Documentation example syntax fix.
394 NM/17 Changed NOTICE file to remove references to embedded PCRE.
396 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
399 NM/19 Bugzilla 745: TLS version reporting.
400 Patch provided by Phil Pennock.
402 NM/20 Bugzilla 167: bool: condition support.
403 Patch provided by Phil Pennock.
405 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
406 clients. Patch provided by Phil Pennock.
408 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
409 Patch provided by Brad "anomie" Jorsch.
411 NM/23 Bugzilla 687: Fix misparses in eximstats.
412 Patch provided by Heiko Schlittermann.
414 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
415 Patch provided by Heiko Schlittermann.
417 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
418 plus update to original patch.
420 NM/26 Bugzilla 799: Documentation correction for ratelimit.
422 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
423 Patch provided by David Brownlee.
425 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
427 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
429 NM/30 Bugzilla 888: TLS documentation bugfixes.
431 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
433 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
434 Unlike the original bugzilla I have changed all shell scripts in src tree.
436 NM/33 Bugzilla 898: Transport filter timeout fix.
437 Patch by Todd Rinaldo.
439 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
440 Patch by Serge Demonchaux.
442 NM/35 Bugzilla 39: Base64 decode bug fixes.
443 Patch by Jakob Hirsch.
445 NM/36 Bugzilla 909: Correct connect() call in dcc code.
447 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
449 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
451 NM/39 Bugzilla 911: Fixed MakeLinks build script.
457 TK/01 Add preliminary DKIM support. Currently requires a forked version of
458 ALT-N's libdkim that I have put here:
459 http://duncanthrax.net/exim-experimental/
461 Note to Michael Haardt: I had to rename some vars in sieve.c. They
462 were called 'true' and it seems that C99 defines that as a reserved
463 keyword to be used with 'bool' variable types. That means you could
464 not include C99-style headers which use bools without triggering
465 build errors in sieve.c.
467 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
468 as mailq or other aliases. Changed the --help handling significantly
469 to do whats expected. exim_usage() emits usage/help information.
471 SC/01 Added the -bylocaldomain option to eximstats.
473 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
475 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
477 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
483 PH/01 Another patch from the Sieve maintainer.
485 PH/02 When an IPv6 address is converted to a string for single-key lookup
486 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
487 dots are used instead of colons so that keys in lsearch files need not
488 contain colons. This was done some time before quoting was made available
489 in lsearch files. However, iplsearch files do require colons in IPv6 keys
490 (notated using the quote facility) so as to distinguish them from IPv4
491 keys. This meant that lookups for IP addresses in host lists did not work
492 for iplsearch lookups.
494 This has been fixed by arranging for IPv6 addresses to be expressed with
495 colons if the lookup type is iplsearch. This is not incompatible, because
496 previously such lookups could never work.
498 The situation is now rather anomolous, since one *can* have colons in
499 ordinary lsearch keys. However, making the change in all cases is
500 incompatible and would probably break a number of configurations.
502 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
505 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
506 conversion specification without a maximum field width, thereby enabling
507 a rogue spamd server to cause a buffer overflow. While nobody in their
508 right mind would setup Exim to query an untrusted spamd server, an
509 attacker that gains access to a server running spamd could potentially
510 exploit this vulnerability to run arbitrary code as the Exim user.
512 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
513 $primary_hostname instead of what libspf2 thinks the hosts name is.
515 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
516 a directory entry by the name of the lookup key. Previously, if a
517 symlink pointed to a non-existing file or a file in a directory that
518 Exim lacked permissions to read, a lookup for a key matching that
519 symlink would fail. Now it is enough that a matching directory entry
520 exists, symlink or not. (Bugzilla 503.)
522 PH/03 The body_linecount and body_zerocount variables are now exported in the
525 PH/04 Added the $dnslist_matched variable.
527 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
528 This means they are set thereafter only if the connection becomes
531 PH/06 Added the client_condition to authenticators so that some can be skipped
532 by clients under certain conditions.
534 PH/07 The error message for a badly-placed control=no_multiline_responses left
535 "_responses" off the end of the name.
537 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
539 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
540 (without spaces) instead of just copying the configuration text.
542 PH/10 Added the /noupdate option to the ratelimit ACL condition.
544 PH/11 Added $max_received_linelength.
546 PH/12 Added +ignore_defer and +include_defer to host lists.
548 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
549 way in which PCRE > 7.0 is built.
551 PH/14 Implemented queue_only_load_latch.
553 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
554 MAIL command. The effect was to mangle the value on 64-bit systems.
556 PH/16 Another patch from the Sieve maintainer.
558 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
560 PH/18 If a system quota error occurred while trying to create the file for
561 a maildir delivery, the message "Mailbox is full" was not appended to the
562 bounce if the delivery eventually timed out. Change 4.67/27 below applied
563 only to a quota excession during the actual writing of the file.
565 PH/19 It seems that peer DN values may contain newlines (and other non-printing
566 characters?) which causes problems in log lines. The DN values are now
567 passed through string_printing() before being added to log lines.
569 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
570 and InterBase are left for another time.)
572 PH/21 Added message_body_newlines option.
574 PH/22 Guard against possible overflow in moan_check_errorcopy().
576 PH/23 POSIX allows open() to be a macro; guard against that.
578 PH/24 If the recipient of an error message contained an @ in the local part
579 (suitably quoted, of course), incorrect values were put in $domain and
580 $local_part during the evaluation of errors_copy.
586 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
587 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
590 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
591 issue a MAIL command.
593 PH/02 In an ACL statement such as
595 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
597 if a client was not listed at all, or was listed with a value other than
598 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
599 the condition was not true (as it should be), so access was not denied.
600 The bug was that the ! inversion was incorrectly passed on to the second
601 item. This has been fixed.
603 PH/03 Added additional dnslists conditions == and =& which are different from
604 = and & when the dns lookup returns more than one IP address.
606 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
607 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
609 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
610 FSYNC, which compiles an option called disable_fsync that allows for
611 bypassing fsync(). The documentation is heavily laced with warnings.
613 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
615 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
616 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
617 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
618 including adding "make clean"; (3) Added -fPIC when compiling the test
619 dynamically loaded module, to get rid of a warning.
621 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
622 message fails, move_frozen_messages = true and ignore_bounce_errors_after
623 = 0s. The bug is otherwise harmless.
625 PH/07 There was a bug in the dovecot authenticator such that the value of
626 $auth1 could be overwritten, and so not correctly preserved, after a
627 successful authentication. This usually meant that the value preserved by
628 the server_setid option was incorrect.
630 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
632 PH/09 Installed PCRE release 7.0.
634 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
635 run for batched SMTP input. It is now run at the start of every message
636 in the batch. While fixing this I discovered that the process information
637 (output by running exiwhat) was not always getting set for -bs and -bS
638 input. This is fixed, and it now also says "batched" for BSMTP.
640 PH/11 Added control=no_pipelining.
642 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
643 patch, slightly modified), and move the expansion of helo_data till after
644 the connection is made in the smtp transport (so it can use these
647 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
649 PH/14 Added log_selector = +pid.
651 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
653 PH/16 Add ${if forany and ${if forall.
655 PH/17 Added dsn_from option to vary the From: line in DSNs.
657 PH/18 Flush SMTP output before performing a callout, unless control =
658 no_callout_flush is set.
660 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
661 was true (the default) a successful delivery failed to delete the retry
662 item, thus causing premature timeout of the address. The bug is now
665 PH/20 Added hosts_avoid_pipelining to the smtp transport.
667 PH/21 Long custom messages for fakedefer and fakereject are now split up
668 into multiline reponses in the same way that messages for "deny" and
669 other ACL rejections are.
671 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
672 with slight modification.
674 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
675 draft, changing the syntax and factoring some duplicate code".
677 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
678 for deliveries of the second and subsequent messages over the same SMTP
681 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
682 ${reduce, with only minor "tidies".
684 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
686 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
687 expansion side effects.
689 PH/27 When a message times out after an over-quota error from an Exim-imposed
690 quota, the bounce message says "mailbox is full". This message was not
691 being given when it was a system quota that was exceeded. It now should
694 MH/03 Made $recipients available in local_scan(). local_scan() already has
695 better access to the recipient list through recipients_list[], but
696 $recipients can be useful in postmaster-provided expansion strings.
698 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
699 in the case of a MAIL command with additional options following the
700 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
701 were accidentally chopped off.
703 PH/29 SMTP synchronization checks are implemented when a command is read -
704 there is a check that no more input is waiting when there shouldn't be
705 any. However, for some commands, a delay in an ACL can mean that it is
706 some time before the response is written. In this time, more input might
707 arrive, invalidly. So now there are extra checks after an ACL has run for
708 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
709 pipelining has not been advertised.
711 PH/30 MH's patch to allow iscntrl() characters to be list separators.
713 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
714 returned in the SMTP response when smtp_return_error_details was false.
717 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
718 instead of the C I/O that was originally supplied, because problems were
721 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
722 Exim which did not show up earlier: it was assuming that a call to
723 SSL_CTX_set_info_callback() might give an error value. In fact, there is
724 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
725 was a macro that became an assignment, so it seemed to work. This has
726 changed to a proper function call with a void return, hence the compile
727 error. Exim's code has been fixed.
729 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
732 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
734 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
736 PH/37 If a message is not accepted after it has had an id assigned (e.g.
737 because it turns out to be too big or there is a timeout) there is no
738 "Completed" line in the log. When some messages of this type were
739 selected by exigrep, they were listed as "not completed". Others were
740 picked up by some special patterns. I have improved the selection
741 criteria to be more general.
743 PH/38 The host_find_failed option in the manualroute router can now be set
744 to "ignore", to completely ignore a host whose IP address cannot be
745 found. If all hosts are ignored, the behaviour is controlled by the new
746 host_all_ignored option.
748 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
749 homing or because of multiple MX records with /mx) generated more than
750 one IP address, and the following item turned out to be the local host,
751 all the secondary addresses of the first item were incorrectly removed
752 from the list, along with the local host and any following hosts (which
753 is what is supposed to happen).
755 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
756 whoever called Exim into the -H file. In the case of the daemon it was
757 behaving confusingly. When first started, it used values for whoever
758 started the daemon, but after a SIGHUP it used the Exim user (because it
759 calls itself on a restart). I have changed the code so that it now always
762 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
763 message are rejected with the same error (e.g. no authentication or bad
764 sender address), and a DATA command is nevertheless sent (as can happen
765 with PIPELINING or a stupid MUA), the error message that was given to the
766 RCPT commands is included in the rejection of the DATA command. This is
767 intended to be helpful for MUAs that show only the final error to their
770 PH/42 Another patch from the Sieve maintainer.
772 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
773 Eximstats - Fixed some broken HTML links and added missing column headers
775 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
776 columns for Rejects, Temp Rejects, Ham, and Spam rows.
778 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
780 PH/43 Yet another patch from the Sieve maintainer.
782 PH/44 I found a way to check for a TCP/IP connection going away before sending
783 the response to the final '.' that terminates a message, but only in the
784 case where the client has not sent further data following the '.'
785 (unfortunately, this is allowed). However, in many cases there won't be
786 any further data because there won't be any more messages to send. A call
787 to select() can be used: if it shows that the input is "ready", there is
788 either input waiting, or the socket has been closed. An attempt to read
789 the next input character can distinguish the two cases. Previously, Exim
790 would have sent an OK response which the client would never have see.
791 This could lead to message repetition. This fix should cure that, at
792 least in a lot of common cases.
794 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
795 advertised in response to EHLO.
801 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
802 fixed by 4.65/MH/01 (is this a record?) are fixed:
804 (i) An empty string was always treated as zero by the numeric comparison
805 operators. This behaviour has been restored.
807 (ii) It is documented that the numeric comparison operators always treat
808 their arguments as decimal numbers. This was broken in that numbers
809 starting with 0 were being interpreted as octal.
811 While fixing these problems I realized that there was another issue that
812 hadn't been noticed. Values of message_size_limit (both the global option
813 and the transport option) were treated as octal if they started with 0.
814 The documentation was vague. These values are now always treated as
815 decimal, and I will make that clear in the documentation.
821 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
822 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
825 MH/01 Don't check that the operands of numeric comparison operators are
826 integers when their expansion is in "skipping" mode (fixes bug
827 introduced by 4.64-PH/07).
829 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
830 child addresses, Exim now panics and dies. Previously, because the count
831 is held in a short int, deliveries were likely to be lost. As such a
832 large number of recipients for a single message is ridiculous
833 (performance will be very, very poor), I have chosen to impose a limit
834 rather than extend the field.
840 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
841 leftover -K file (the existence of which was triggered by #402).
842 While we were at it, introduced process PID as part of the -K
843 filename. This should rule out race conditions when creating
846 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
847 processing considerably. Previous code took too long for large mails,
848 triggering a timeout which in turn triggers #401.
850 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
851 in the DK code in transports.c. sendfile() is not really portable,
852 hence the _LINUX specificness.
854 TF/01 In the add_headers option to the mail command in an Exim filter,
855 there was a bug that Exim would claim a syntax error in any
856 header after the first one which had an odd number of characters
859 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
860 callout verification, Exim cached a "reject" for the entire domain. This
861 is correct for most verifications, but it is not correct for a recipient
862 verification with use_sender or use_postmaster set, because in that case
863 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
864 case of MAIL FROM:<> rejection from other early rejections (e.g.
865 rejection of HELO). When verifying a recipient using a non-null MAIL
866 address, the cache is ignored if it shows MAIL FROM:<> rejection.
867 Whatever the result of the callout, the value of the domain cache is
868 left unchanged (for any other kind of callout, getting as far as trying
869 RCPT means that the domain itself is ok).
871 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
874 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
875 manifest itself as EPIPE rather than ECONNECT. When tidying away a
876 session, the daemon ignores ECONNECT errors and logs others; it now
877 ignores EPIPE as well.
879 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
880 (quoted-printable decoding).
882 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
883 later the small subsequent patch to fix an introduced bug.
885 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
887 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
889 PH/08 An error is now given if message_size_limit is specified negative.
891 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
892 to be given (somewhat) arbitrary names.
894 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
897 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
898 miscellaneous code fixes
900 PH/10 Added the log_reject_target ACL modifier to specify where to log
903 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
904 hostname. This is wrong, because it relates to the incoming message (and
905 probably the interface on which it is arriving) and not to the outgoing
906 callout (which could be using a different interface). This has been
907 changed to use the value of the helo_data option from the smtp transport
908 instead - this is what is used when a message is actually being sent. If
909 there is no remote transport (possible with a router that sets up host
910 addresses), $smtp_active_hostname is used.
912 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
913 tweaks were necessary in order to get it to work (see also 21 below):
914 (a) The code assumed that strncpy() returns a negative number on buffer
915 overflow, which isn't the case. Replaced with Exim's string_format()
917 (b) There were several signed/unsigned issues. I just did the minimum
918 hacking in of casts. There is scope for a larger refactoring.
919 (c) The code used strcasecmp() which is not a standard C function.
920 Replaced with Exim's strcmpic() function.
921 (d) The code set only $1; it now sets $auth1 as well.
922 (e) A simple test gave the error "authentication client didn't specify
923 service in request". It would seem that Dovecot has changed its
924 interface. Fortunately there's a specification; I followed it and
925 changed what the client sends and it appears to be working now.
927 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
930 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
931 address A is aliased to B and C, where B exists and C does not. Without
932 -v the output is "A verified" because verification stops after a
933 successful redirection if more than one address is generated. However,
934 with -v the child addresses are also verified. Exim was outputting "A
935 failed to verify" and then showing the successful verification for C,
936 with its parentage. It now outputs "B failed to verify", showing B's
937 parentage before showing the successful verification of C.
939 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
940 look up a TXT record in a specific list after matching in a combined
943 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
944 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
945 they consult the DNS. I had assumed they would set it the way they
946 wanted; and indeed my experiments on Linux seem to show that in some
947 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
948 To be on the safe side, however, I have now made the interface to
949 host_find_byname() similar to host_find_bydns(), with an argument
950 containing the DNS resolver options. The host_find_byname() function now
951 sets these options at its start, just as host_find_bydns() does. The smtp
952 transport options dns_qualify_single and dns_search_parents are passed to
953 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
954 of host_find_byname() use the default settings of RES_DEFNAMES
955 (qualify_single) but not RES_DNSRCH (search_parents).
957 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
958 spool_read_header() do less string testing, by means of a preliminary
959 switch on the second character of optional "-foo" lines. (This is
960 overdue, caused by the large number of possibilities that now exist.
961 Originally there were few.) While I was there, I also converted the
962 str(n)cmp tests so they don't re-test the leading "-" and the first
963 character, in the hope this might squeeze out yet more improvement.
965 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
966 flag allowing group syntax was set by the header_syntax check but not
967 turned off, possible causing trouble later; (2) The flag was not being
968 set at all for the header_verify test, causing "group"-style headers to
969 be rejected. I have now set it in this case, and also caused header_
970 verify to ignore an empty address taken from a group. While doing this, I
971 came across some other cases where the code for allowing group syntax
972 while scanning a header line wasn't quite right (mostly, not resetting
973 the flag correctly in the right place). These bugs could have caused
974 trouble for malformed header lines. I hope it is now all correct.
976 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
977 with the "reply" argument non-NULL. The code, however (which originally
978 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
979 but it didn't always do it. This confused somebody who was copying the
980 code for some other use. I have removed all the tests.
982 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
983 feature that was used to support insecure browsers during the U.S. crypto
984 embargo. It requires special client support, and Exim is probably the
985 only MTA that supported it -- and would never use it because real RSA is
986 always available. This code has been removed, because it had the bad
987 effect of slowing Exim down by computing (never used) parameters for the
988 RSA_EXPORT functionality.
990 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
991 authenticator to fail if there's a tab character in the incoming data
992 (there should never be unless someone is messing about, as it's supposed
993 to be base64-encoded). Also added, on Timo's advice, the "secured" option
994 if the connection is using TLS or if the remote IP is the same as the
995 local IP, and the "valid-client-cert option" if a client certificate has
998 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
999 authenticators. This can be used for authorization after authentication
1000 succeeds. (In the case of plaintext, it servers for both authentication
1003 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1004 if any retry times were supplied.
1006 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1007 connection, where there is no client IP address to check. In this
1008 situation, the verify now always succeeds.
1010 PH/25 Applied John Jetmore's -Mset patch.
1012 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1014 PH/27 In a string expansion for a processed (not raw) header when multiple
1015 headers of the same name were present, leading whitespace was being
1016 removed from all of them, but trailing whitespace was being removed only
1017 from the last one. Now trailing whitespace is removed from each header
1018 before concatenation. Completely empty headers in a concatenation (as
1019 before) are ignored.
1021 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1022 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1024 PH/29 [Removed. This was a change that I later backed out, and forgot to
1025 correct the ChangeLog entry (that I had efficiently created) before
1026 committing the later change.]
1028 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1029 address errors (4xx response to RCPT) over the same connection as other
1030 messages routed to the same hosts. Such deliveries are always "forced",
1031 so retry times are not inspected. This resulted in far too many retries
1032 for the affected addresses. The effect occurred only when there were more
1033 hosts than the hosts_max_try setting in the smtp transport when it had
1034 the 4xx errors. Those hosts that it had tried were not added to the list
1035 of hosts for which the message was waiting, so if all were tried, there
1036 was no problem. Two fixes have been applied:
1038 (i) If there are any address or message errors in an SMTP delivery, none
1039 of the hosts (tried or untried) are now added to the list of hosts
1040 for which the message is waiting, so the message should not be a
1041 candidate for sending over the same connection that was used for a
1042 successful delivery of some other message. This seems entirely
1043 reasonable: after all the message is NOT "waiting for some host".
1044 This is so "obvious" that I'm not sure why it wasn't done
1045 previously. Hope I haven't missed anything, but it can't do any
1046 harm, as the worst effect is to miss an optimization.
1048 (ii) If, despite (i), such a delivery is accidentally attempted, the
1049 routing retry time is respected, so at least it doesn't keep
1050 hammering the server.
1052 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1053 in ${readsocket because some servers need this prod.
1055 PH/32 Added some extra debug output when updating a wait-xxx database.
1057 PH/33 The hint "could be header name not terminated by colon", which has been
1058 given for certain expansion errors for a long time, was not being given
1059 for the ${if def:h_colon_omitted{... case.
1061 PH/34 The spec says: "With one important exception, whenever a domain list is
1062 being scanned, $domain contains the subject domain." There was at least
1063 one case where this was not true.
1065 PH/35 The error "getsockname() failed: connection reset by peer" was being
1066 written to the panic log as well as the main log, but it isn't really
1067 panic-worthy as it just means the connection died rather early on. I have
1068 removed the panic log writing for the ECONNRESET error when getsockname()
1071 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1072 runs only) independently of the message's sender address. This meant
1073 that, if the 4xx error was in fact related to the sender, a different
1074 message to the same recipient with a different sender could confuse
1075 things. In particualar, this can happen when sending to a greylisting
1076 server, but other circumstances could also provoke similar problems.
1077 I have changed the default so that the retry time for these errors is now
1078 based a combination of the sender and recipient addresses. This change
1079 can be overridden by setting address_retry_include_sender=false in the
1082 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1083 remote server are returned as part of bounce messages. This was not
1084 happening for LMTP over a pipe (the lmtp transport), but now it is the
1085 same for both kinds of LMTP.
1087 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1088 in header lines that were in fact CNAMEs. This is no longer the case.
1090 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1091 and queue runs started by the daemon processed all messages. This has
1092 been fixed so that -R and -S can now usefully be given with -q<time>.
1094 PH/40 Import PCRE release 6.7 (fixes some bugs).
1096 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1098 PH/42 Give an error if -q is specified more than once.
1100 PH/43 Renamed the variables $interface_address and $interface_port as
1101 $received_ip_address and $received_port, to make it clear that these
1102 values apply to message reception, and not to the outgoing interface when
1103 a message is delivered. (The old names remain recognized, of course.)
1105 PH/44 There was no timeout on the connect() call when using a Unix domain
1106 socket in the ${readsocket expansion. There now is.
1108 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1109 be meaningful with "accept".
1111 SC/01 Eximstats V1.43
1112 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1114 SC/02 Eximstats V1.44
1115 Use a glob alias rather than an array ref in the generated
1116 parser. This improves both readability and performance.
1118 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1119 Collect SpamAssassin and rejection statistics.
1120 Don't display local sender or destination tables unless
1121 there is data to show.
1122 Added average volumes into the top table text output.
1124 SC/04 Eximstats V1.46
1125 Collect data on the number of addresses (recipients)
1126 as well as the number of messages.
1128 SC/05 Eximstats V1.47
1129 Added 'Message too big' to the list of mail rejection
1130 reasons (thanks to Marco Gaiarin).
1132 SC/06 Eximstats V1.48
1133 Mainlog lines which have GMT offsets and are too short to
1134 have a flag are now skipped.
1136 SC/07 Eximstats V1.49 (Alain Williams)
1137 Added the -emptyok flag.
1139 SC/08 Eximstats V1.50
1140 Fixes for obtaining the IP address from reject messages.
1142 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1143 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1144 whitesspace changes from 4.64-PH/27
1146 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1149 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1150 are found, allow negative numbers in numeric criteria)
1152 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1154 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1155 to match changes made in 4.64-PH/43
1157 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1159 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1160 "domain" is actually an IP address, and force a failure. This locks out
1161 those revolvers/nameservers that support "A-for-A" lookups, in
1162 contravention of the specifications.
1164 PH/48 When a host name was looked up from an IP address, and the subsequent
1165 forward lookup of the name timed out, the host name was left in
1166 $sender_host_name, contrary to the specification.
1168 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1169 restricted to single-key lookups, Exim was not diagnosing an error if
1170 * or *@ was used with a query-style lookup.
1172 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1174 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1175 long ago, but noone interested enough thought of it. Let's just say that
1176 the "1.1" means that there are some new functions that weren't there at
1177 some point in the past.
1179 PH/51 Error processing for expansion failure of helo_data from an smtp
1180 transport during callout processing was broken.
1182 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1183 tested/used via the -bh/-bhc/-bs options.
1185 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1186 bug, fixed in subsequent PCRE releases).
1188 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1189 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1191 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1197 SC/01 Use a glob alias rather than an array ref in eximstats generated
1198 parser. This improves both readability and performance.
1200 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1201 Don't display local sender or destination tables in eximstats unless
1202 there is data to show.
1203 Added average volumes into the eximstats top table text output.
1205 SC/03 Collect data on the number of addresses (recipients) as well
1206 as the number of messages in eximstats.
1208 TF/01 Correct an error in the documentation for the redirect router. Exim
1209 does (usually) call initgroups() when daemonizing.
1211 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1212 with consistent privilege compared to when running as a daemon.
1214 TF/03 Note in the spec that $authenticated_id is not set for local
1215 submissions from trusted users.
1217 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1218 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1220 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1221 by adding some example configuration directives to the default
1222 configuration file. A little bit of work is required to uncomment the
1223 directives and define how usernames and passwords are checked, but
1224 there is now a framework to start from.
1226 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1227 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1228 without this. I don't know how relevant this is to other LDAP libraries.
1230 PH/02 Add the verb name to the "unknown ACL verb" error.
1232 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1234 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1236 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1237 directory not expanded when it should be if an expanded home directory
1238 was set for the address (which is overridden by the transport).
1240 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1243 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1244 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1245 because it is too late at that time, and has no effect.
1247 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1248 security issue with \' (bugzilla #107). I could not use the
1249 PQescapeStringConn() function, because it needs a PGconn value as one of
1252 PH/08 When testing addresses using -bt, indicate those final addresses that
1253 are duplicates that would not cause an additional delivery. At least one
1254 person was confused, thinking that -bt output corresponded to deliveries.
1255 (Suppressing duplicates isn't a good idea as you lose the information
1256 about possibly different redirections that led to the duplicates.)
1258 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1259 systems where poll() doesn't work, in particular OS X.
1261 PH/10 Added more information to debugging output for retry time not reached.
1263 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1264 operations in malware.c.
1266 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1269 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1270 syslog_duplication set false, log lines that would normally be written
1271 both the the main log and to the reject log were not written to syslog at
1274 PH/14 In the default configuration, change the use of "message" in ACL warn
1275 statements to "add_header".
1277 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1278 not followed by a command (e.g. "seen endif").
1280 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1281 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1284 PH/17 Added extra conditions to the default value of delay_warning_condition
1288 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1289 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1290 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1293 The Auto-Submitted: and various List- headers are standardised, whereas I
1294 don't think Precedence: ever was.
1296 PH/18 Refactored debugging code in route_finduser() to show more information,
1297 in particular, the error code if getpwnam() issues one.
1299 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1300 This is apparently needed in addition to the PH/07 change above to avoid
1301 any possible encoding problems.
1303 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1304 but not after initializing Perl.
1306 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1307 output them only if debugging. By default they are written stderr,
1308 apparently, which is not desirable.
1310 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1313 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1316 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1318 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1319 authenticated or an ident call has been made. Suppress the default
1320 values for $authenticated_id and $authenticated_sender (but permit -oMai
1321 and -oMas) when testing with -bh.
1323 PH/24 Re-jigged the order of the tests in the default configuration so that the
1324 tests for valid domains and recipients precede the DNS black list and CSA
1325 tests, on the grounds that those ones are more expensive.
1327 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1328 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1329 This bug exists in every version of Exim that I still have, right back to
1332 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1333 However, an attempt to turn on case-sensitivity in a regex key by
1334 including (?-i) didn't work because the subject string was already
1335 lowercased, and the effects were non-intuitive. It turns out that a
1336 one-line patch can be used to allow (?-i) to work as expected.
1342 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1343 other effects) broke the use of negated acl sub-conditions.
1345 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1348 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1349 "Deny" causes Exim to reject the incoming connection with a 554 error.
1350 Unfortunately, if there is a major crisis, such as a disk failure,
1351 tcp-wrappers gives "deny", whereas what one would like would be some
1352 kind of temporary error. A kludge has been added to help with this.
1353 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1354 554 error is used if errno is still zero or contains ENOENT (which occurs
1355 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1358 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1360 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1361 errors. Otherwise a message that provokes a temporary error (when other
1362 messages do not) can cause a whole host to time out.
1364 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1365 addresses were routed directly to files or pipes from a redirect router.
1366 File deliveries just didn't batch; pipe deliveries might have suffered
1369 PH/06 A failure to get a lock for a hints database would erroneously always say
1370 "Failed to get write lock", even when it was really a read lock.
1372 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1373 of 0600. This has been changed to use the value of the lockfile_mode
1374 option (which defaults to 0600).
1376 PH/08 Applied small patch from the Sieve maintainer.
1378 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1379 folder from quota calculations, a direct delivery into this folder messed
1380 up the contents of the maildirsize file. This was because the regex was
1381 used only to exclude .Trash (or whatever) when the size of the mailbox
1382 was calculated. There was no check that a delivery was happening into an
1383 excluded directory. This bug has been fixed by ignoring all quota
1384 processing for deliveries into excluded directories.
1386 PH/10 Added the maildirfolder_create_regex option to appendfile.
1392 PH/01 The code for finding all the local interface addresses on a FreeBSD
1393 system running IPv6 was broken. This may well have applied to all BSD
1394 systems, as well as to others that have similar system calls. The broken
1395 code found IPv4 interfaces correctly, but gave incorrect values for the
1396 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1397 that it would not match correctly against @[] and not recognize the IPv6
1400 PH/02 The ipliteral router was not recognizing addresses of the form user@
1401 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1403 PH/03 Added disable_ipv6.
1405 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1406 decoded form, because it is most often used to construct To: headers
1407 lines in autoreplies, and the decoded form may well be syntactically
1408 invalid. However, $reply_address has leading white space removed, and all
1409 newlines turned into spaces so that the autoreply transport does not
1412 PH/05 If group was specified without a user on a router, and no group or user
1413 was specified on a transport, the group from the router was ignored.
1415 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1416 for visible compile-time settings that can be used to change these
1417 numbers, for those that want even more. Backwards compatibility with old
1418 spool files has been maintained. However, going back to a previous Exim
1419 release will lost any variables that are in spool files.
1421 PH/07 Two small changes when running in the test harness: increase delay when
1422 passing a TCP/IP connection to a new process, in case the original
1423 process has to generate a bounce, and remove special handling of
1424 127.0.0.2 (sic), which is no longer necessary.
1426 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1427 be the same on different OS.
1429 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1432 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1433 whether --show-vars was specified or not
1435 JJ/02 exipick: Added support for new ACL variable spool format introduced
1438 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1439 syntactically invalid From: or Reply-to: line, and a filter used this to
1440 generate an autoreply, and therefore failed to obtain an address for the
1441 autoreply, Exim could try to deliver to a non-existent relative file
1442 name, causing unrelated and misleading errors. What now happens is that
1443 it logs this as a hard delivery error, but does not attempt to create a
1446 PH/11 The exinext utility has a -C option for testing purposes, but although
1447 the given file was scanned by exinext itself; it wasn't being passed on
1448 when Exim was called.
1450 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1451 an end-of-file indication when reading a command response.
1453 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1454 compiled. In many other places in Exim, IPv6 addresses are always
1455 recognized, so I have changed this. It also means that IPv4 domain
1456 literals of the form [IPV4:n.n.n.n] are now always recognized.
1458 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1459 used if the router is not running as root, for example, when verifying at
1460 ACL time, or when using -bh. The debugging output from this situation was
1461 non-existent - all you got was a failure to exec. I have made two
1464 (a) Failures to set uid/gid, the current directory, or a process leader
1465 in a subprocess such as that created by queryprogram now generate
1466 suitable debugging ouput when -d is set.
1468 (b) The queryprogram router detects when it is not running as root,
1469 outputs suitable debugging information if -d is set, and then runs
1470 the subprocess without attempting to change uid/gid.
1472 PH/15 Minor change to Makefile for building test_host (undocumented testing
1475 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1476 additional section of a DNS packet that returns MX or SRV records.
1477 Instead, it always explicitly searches for A/AAAA records. This avoids
1478 major problems that occur when a DNS server includes only records of one
1479 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1480 fixed another bug: if SRV records were looked up and the corresponding
1481 address records were *not* found in the additional section, the port
1482 values from the SRV records were lost.
1484 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1485 using the correct key (the original address) when searching the retry
1486 rules in order to find which one to use for generating the retry hint.
1488 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1489 adding the default one. Similarly, if it contains a Reply-To: header, the
1490 errors_reply_to option, if set, is not used.
1492 PH/19 When calculating a retry time, Exim used to measure the "time since
1493 failure" by looking at the "first failed" field in the retry record. Now
1494 it does not use this if it is later than than the arrival time of the
1495 message. Instead it uses the arrival time. This makes for better
1496 behaviour in cases where some deliveries succeed, thus re-setting the
1497 "first failed" field. An example is a quota failure for a huge message
1498 when small messages continue to be delivered. Without this change, the
1499 "time since failure" will always be short, possible causing more frequent
1500 delivery attempts for the huge message than are intended.
1501 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1503 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1504 $1, $2, $3) because the numerical variables can be reset during some
1505 expansion items (e.g. "match"), thereby losing the authentication data.
1507 PH/21 Make -bV show the size of off_t variables so that the test suite can
1508 decide whether to run tests for quotas > 2G.
1510 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1511 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1512 filecount value is greater than 2G or if a quota value is greater than 2G
1513 on a system where the size of off_t is not greater than 4, a panic error
1516 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1517 never match. The debug and -bh output now contains an explicit error
1518 message indicating a malformed IPv4 address or mask.
1520 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1521 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1522 PH/23 above applies.
1524 PH/25 Do not write to syslog when running in the test harness. The only
1525 occasion when this arises is a failure to open the main or panic logs
1526 (for which there is an explicit test).
1528 PH/26 Added the /no_tell option to "control=freeze".
1530 PH/27 If a host name lookup failed very early in a connection, for example, if
1531 the IP address matched host_lookup and the reverse lookup yielded a name
1532 that did not have a forward lookup, an error message of the form "no IP
1533 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1534 could be logged. Now it outputs the IP address instead of "NULL".
1536 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1537 allows the sender and the authenticated sender to be set when
1538 submitting a message from within Exim. Since child_open_exim() is
1539 documented for local_scan(), the new function should be too.
1541 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1542 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1543 results in an empty string is now treated as unset.
1545 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1547 PH/31 Added sender_verify_fail logging option.
1549 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1550 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1551 tidied the source and removed it altogether.
1553 PH/33 When a queue run was abandoned because the load average was too high, a
1554 log line was always written; now it is written only if the queue_run log
1555 selector is set. In addition, the log line for abandonment now contains
1556 information about the queue run such as the pid. This is always present
1557 in "start" and "stop" lines but was omitted from the "abandon" line.
1559 PH/34 Omit spaces between a header name and the colon in the error message that
1560 is given when verify = headers_syntax fails (if there are lots of them,
1561 the message gets confusing).
1563 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1564 names, as there are now some PTR records that contain slashes. This check
1565 is only to protect against broken name servers that fall over on strange
1566 characters, so the fact that it applies to all lookups doesn't matter.
1568 PH/36 Now that the new test suite is complete, we can remove some of the
1569 special code in Exim that was needed for the old test suite. For example,
1570 sorting DNS records because real resolvers return them in an arbitrary
1571 order. The new test suite's fake resolver always returns records in the
1574 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1575 bounces) except when queue_only is set, to avoid logging races between
1576 the different processes.
1578 PH/38 Panic-die if .include specifies a non-absolute path.
1580 PH/39 A tweak to the "H" retry rule from its user.
1582 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1583 a label. They prevented compilation on older perls.
1585 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1586 a warning to be raised on newish perls.
1588 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1589 on queue. Changes to match documented behaviour of showing count of
1590 messages matching specified criteria.
1592 PH/40 Changed the default ident timeout from 30s to 5s.
1594 PH/41 Added support for the use of login_cap features, on those BSD systems
1595 that have them, for controlling the resources used by pipe deliveries.
1597 PH/42 The content-scanning code uses fopen() to create files in which to put
1598 message data. Previously it was not paying any attention to the mode of
1599 the files. Exim runs with umask(0) because the rest of the code creates
1600 files with open(), and sets the required mode explicitly. Thus, these
1601 files were ending up world-writeable. This was not a big issue, because,
1602 being within the spool directory, they were not world-accessible. I have
1603 created a function called modefopen, which takes an additional mode
1604 argument. It sets umask(777), creates the file, chmods it to the required
1605 mode, then resets the umask. All the relevant calls to fopen() in the
1606 content scanning code have been changed to use this function.
1608 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1609 to 24 hours. This avoids potential overflow problems when processing G
1610 and H retry rules. I suspect nobody ever tinkers with this value.
1612 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1614 PH/45 When the plaintext authenticator is running as a client, the server's
1615 challenges are checked to ensure they are valid base64 strings. By
1616 default, the authentication attempt is cancelled if an invalid string is
1617 received. Setting client_ignore_invalid_base64 true ignores these errors.
1618 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1619 they are received. Thus, the responses can be made to depend on the
1620 challenges. If an invalid string is ignored, an empty string is placed in
1623 PH/46 Messages that are created by the autoreply transport now contains a
1624 References: header, in accordance with RFCs 2822 and 3834.
1626 PH/47 Added authenticated_sender_force to the smtp transport.
1628 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1630 PH/49 Installed latest patch from the Sieve maintainer.
1632 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1633 was also set, the appendfile transport was unnecessarily scanning a
1634 directory of message files (e.g. for maildir delivery) to find the count
1635 of files (along with the size), even though it did not need this
1636 information. It now does the scan only if it needs to find either the
1637 size of the count of files.
1639 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1641 PH/52 Two bugs concerned with error handling when the smtp transport is
1644 (i) Exim was not creating retry information for temporary errors given
1645 for individual recipients after the DATA command when the smtp transport
1646 was used in LMTP mode. This meant that they could be retried too
1647 frequently, and not timed out correctly.
1649 (ii) Exim was setting the flag that allows error details to be returned
1650 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1651 recipients that were returned after the DATA command.
1653 PH/53 This is related to PH/52, but is more general: for any failing address,
1654 when detailed error information was permitted to be returned to the
1655 sender, but the error was temporary, then after the final timeout, only
1656 "retry timeout exceeded" was returned. Now it returns the full error as
1657 well as "retry timeout exceeded".
1659 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1660 do this, and (what is worse) MTAs that accept it.
1662 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1663 will now be deprecated.
1665 PH/56 New os.c-cygwin from the Cygwin maintainer.
1667 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1668 formats (previously only available in exim formats via -bpr, -bpru,
1669 and -bpra. Now also available in native and exiqgrep formats)
1671 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1672 with very large, slow to parse queues
1674 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1676 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1678 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1679 responses to authentication challenges, though it was showing the
1680 challenges; (ii) I've removed the CR characters from the debug output for
1683 PH/58 Allow for the insertion of a newline as well as a space when a string
1684 is turned into more than one encoded-word during RFC 2047 encoding. The
1685 Sieve code now uses this.
1687 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1688 data_4xx, lost_connection, tls_required.
1690 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1691 message was being sent as an SMTP response.
1693 PH/61 Add -l and -k options to exicyclog.
1695 PH/62 When verifying, if an address was redirected to one new address, so that
1696 verification continued, and the new address failed or deferred after
1697 having set something in $address_data, the value of $address_data was not
1698 passed back to the ACL. This was different to the case when no
1699 redirection occurred. The value is now passed back in both cases.
1701 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1702 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1703 use login_cap.h, so on its own it isn't the distinguishing feature. The
1704 new name refers directly to the setclassresources() function.
1706 PH/65 Added configuration files for NetBSD3.
1708 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1710 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1711 is preferred over IPv4.
1713 PH/68 The bounce_return_message and bounce_return_body options were not being
1714 honoured for bounces generated during the reception of non-SMTP messages.
1715 In particular, this applied to messages rejected by the ACL. This bug has
1716 been fixed. However, if bounce_return_message is true and bounce_return_
1717 body is false, the headers that are returned for a non-SMTP message
1718 include only those that have been read before the error was detected.
1719 (In the case of an ACL rejection, they have all been read.)
1721 PH/69 The HTML version of the specification is now built in a directory called
1722 spec_html instead of spec.html, because the latter looks like a path with
1723 a MIME-type, and this confuses some software.
1725 PH/70 Catch two compiler warnings in sieve.c.
1727 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1728 function verify_get_ident() calls ip_connect() to connect a socket, but
1729 if the "connect()" function timed out, ip_connect() used to close the
1730 socket. However, verify_get_ident() also closes the socket later, and in
1731 between Exim writes to the log, which may get opened at this point. When
1732 the socket was closed in ip_connect(), the log could get the same file
1733 descriptor number as the socket. This naturally causes chaos. The fix is
1734 not to close the socket in ip_connect(); the socket should be closed by
1735 the function that creates it. There was only one place in the code where
1736 this was missing, in the iplookup router, which I don't think anybody now
1737 uses, but I've fixed it anyway.
1739 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1740 well as to direct DNS lookups. Otherwise the handling of names in host
1741 lists is inconsistent and therefore confusing.
1747 PH/01 Two changes to the default runtime configuration:
1749 (1) Move the checks for relay_from_hosts and authenticated clients from
1750 after to before the (commented out) DNS black list checks.
1752 (2) Add control=submission to the relay_from_hosts and authenticated
1753 clients checks, on the grounds that messages accepted by these
1754 statements are most likely to be submissions.
1756 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1758 (1) Generate an error if the third argument for the ${prvs expansion is
1761 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1764 (3) Reset the variables that are obtained from the first argument of
1765 ${prvscheck and used in the second argument before leaving the code,
1766 because their memory is reclaimed, so using them afterwards may do
1769 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1770 one (it's much easier than Tom thought :-).
1772 (5) Because of (4), we can now allow for the use of $prvscheck_result
1773 inside the third argument.
1775 PH/03 For some reason, the default setting of PATH when running a command from
1776 a pipe transport was just "/usr/bin". I have changed it to
1779 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1780 anything to be listed in the output from -bV.
1782 PH/05 When a filter generated an autoreply, the entire To: header line was
1783 quoted in the delivery log line, like this:
1785 => >A.N.Other <ano@some.domain> <original@ddress> ...
1787 This has been changed so that it extracts the operative address. There
1788 may be more than one such address. If so, they are comma-separated, like
1791 => >ano@some.domain,ona@other.domain <original@ddress> ...
1793 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1794 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1795 not being looked up in the rDNS to get a host name, Exim was showing the
1796 IP address twice in Received: lines, even though the IP addresses were
1797 identical. For example:
1799 Received: from [1.2.3.4] (helo=[1.2.3.4])
1801 However, if the real host name was known, it was omitting the HELO data
1802 if it matched the actual IP address. This has been tidied up so that it
1803 doesn't show the same IP address twice.
1805 PH/07 When both +timestamp and +memory debugging was on, the value given by
1806 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1807 called by the debug printing, thereby overwriting the timestamp buffer.
1808 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1810 PH/08 When the original message was included in an autoreply transport, it
1811 always said "this is a copy of the message, including all the headers",
1812 even if body_only or headers_only was set. It now gives an appropriate
1815 PH/09 Applied a patch from the Sieve maintainer which:
1817 o fixes some comments
1818 o adds the (disabled) notify extension core
1819 o adds some debug output for the result of if/elsif tests
1820 o points to the current vacation draft in the documentation
1821 and documents the missing references header update
1825 o fixes a bug in processing the envelope test (when testing
1826 multiple envelope elements, the last element determinted the
1829 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1830 Electronic Mail") by including:
1832 Auto-submitted: auto-generated
1834 in the messages that it generates (bounce messages and others, such as
1835 warnings). In the case of bounce messages for non-SMTP mesages, there was
1836 also a typo: it was using "Auto_submitted" (underscore instead of
1837 hyphen). Since every message generated by Exim is necessarily in response
1838 to another message, thes have all been changed to:
1840 Auto-Submitted: auto-replied
1842 in accordance with these statements in the RFC:
1844 The auto-replied keyword:
1846 - SHOULD be used on messages sent in direct response to another
1847 message by an automatic process,
1849 - MUST NOT be used on manually-generated messages,
1851 - MAY be used on Delivery Status Notifications (DSNs) and Message
1852 Disposition Notifications (MDNs),
1854 - MUST NOT be used on messages generated by automatic or periodic
1855 processes, except for messages which are automatic responses to
1858 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1859 to the default Received: header definition.
1861 PH/12 Added log selector acl_warn_skipped (default on).
1863 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1864 variables because (a) they are in the wrong storage pool and (b) even if
1865 they were copied, it wouldn't work properly because of the caching.
1867 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1868 checking when decoding. Apparently there are clients that generate
1869 overlong encoded strings. Why am I not surprised?
1871 PH/15 If the first argument of "${if match_address" was not empty, but did not
1872 contain an "@" character, Exim crashed. Now it writes a panic log message
1873 and treats the condition as false.
1875 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1877 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1878 extension "envelope-auth". The code is finished and in agreement with
1879 other implementations, but there is no documentation so far and in fact,
1880 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1881 not changing the active code.
1883 Print executed "if" and "elsif" statements when debugging is used. This
1884 helps a great deal to understand what a filter does.
1886 Document more things not specified clearly in RFC3028. I had all this
1887 sorted out, when out of a sudden new issues came to my mind. Oops."
1889 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1892 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1893 canonical form (as documented). However, after a host name lookup from
1894 the IP address, check_host() was doing a simple string comparison with
1895 addresses acquired from the DNS when checking that the found name did
1896 have the original IP as one of its addresses. Since any found IPv6
1897 addresses are likely to be in abbreviated form, the comparison could
1898 fail. Luckily, there already exists a function for doing the comparison
1899 by converting both addresses to binary, so now that is used instead of
1900 the text comparison.
1902 PH/20 There was another similar case to PH/19, when a complete host name was
1903 given in a host list; looking up its IP address could give an abbreviated
1904 form, whereas the current host's name might or might not be abbreviated.
1905 The same fix has been applied.
1911 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1912 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1915 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1916 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1918 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1920 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1921 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1922 Also, for "Auto-Submitted", treat anything other than "no" as
1923 non-personal, in accordance with RFC 3834. (Previously it treated
1924 anything starting "auto-" as non-personal.)
1926 TF/01 The control=submission/name=... option had a problem with syntax
1927 errors if the name included a slash character. The /name= option
1928 now slurps the rest of the string, so it can include any characters
1929 but it must come last in the list of options (after /sender_retain
1932 PH/05 Some modifications to the interface to the fake nameserver for the new
1940 TK/01 Added the "success_on_redirect" address verification option. See
1941 NewStuff for rationale and an example.
1943 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1945 PH/02 Patch to exigrep to allow it to work on syslog lines.
1947 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1948 fread() to skip over the body file's header line, because in Cygwin the
1949 header line is locked and is inaccessible.
1951 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1952 co-exist for some time) to make it clear that it is the Exim ID that is
1953 referenced, not the Message-ID: header line.
1955 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1956 string_format() function, because snprintf() does not exist on all
1959 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1960 ${stat: expansion item.
1962 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1963 protocol synchronization error", to keep the pedants happy.
1965 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1966 well as for IRIX systems, when gcc is being used. See the host.c source
1969 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1971 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1974 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1975 transport and to the smtp transport in LMTP mode.
1977 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1979 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1980 run a filter in a subprocess. This could lead to confusion in subsequent
1981 lookups in the parent process. There should also be a search_tidyup() at
1982 the end of the subprocess.
1984 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1985 only if the host matched helo_try_verify_hosts, which caused the
1986 verification to occur when the EHLO/HELO command was issued. The ACL just
1987 tested the remembered result. Now, if a previous verification attempt has
1988 not happened, "verify = helo" does it there and then.
1990 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1992 TK/03 Fix log output including CR from clamd.
1994 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1995 exist provoked a memory error which could cause a segfault.
1997 PH/15 Installed PCRE 6.2
1999 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2001 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2002 of the problem. Specifically, suggested +O2 rather than +O1 for the
2005 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2007 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2008 record, the port was not correctly logged when the outgoing_port log
2009 selector was set (it logged the transort's default port).
2011 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2012 fallback_hosts, and "hosts" in the smtp transport.
2014 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2015 host errors such as "Connection refused".
2017 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2018 authentication with radiusclient 0.4.9:
2020 - Error returned from rc_read_config was caught wrongly
2021 - Username/password not passed on to radius server due to wrong length.
2023 The presumption is that some radiusclient API changes for 4.51/PH/17
2024 were not taken care of correctly. The code is still untested by me (my
2025 Linux distribution still has 0.3.2 of radiusclient), but it was
2026 contributed by a Radius user.
2028 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2029 expanding the "port" option of the smtp transport.
2031 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2032 while reading a MIME header. Thanks to Tom Hughes for a patch.
2034 PH/24 Include config.h inside local_scan.h so that configuration settings are
2037 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2038 that in an ACL for RCPT (for example), you can examine exactly what was
2041 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2042 commands, but it was not correctly comparing the address with the actual
2043 client host address. Thus, it would show the EHLO address in Received:
2044 header lines when this was not necessary.
2046 PH/27 Added the % operator to ${eval:}.
2048 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2049 it should be ignoring failures (because with -C, for example, it has lost
2050 privilege). It wasn't ignoring creation failures other than "already
2053 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2056 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2057 because an input line was too long, either on its own, or by virtue of
2058 too many continuations, the temporary file was not being removed, and the
2059 return code was incorrect.
2061 PH/31 Missing "BOOL" in function definition in filtertest.c.
2063 PH/32 Applied Sieve patches from the maintainer.
2065 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2067 PH/33 Added "verify = not_blind".
2069 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2070 Local/Makefile (with some defaults set). These are used in built scripts
2071 such as exicyclog, but they have never been used in the exim_install
2072 script (though there are many overriding facilities there). I have
2073 arranged that the exim_install script now takes note of these two
2076 PH/35 Installed configuration files for Dragonfly.
2078 PH/36 When a locally submitted message by a trusted user did not contain a
2079 From: header, and the sender address was obtained from -f or from an SMTP
2080 MAIL command, and the trusted user did not use -F to supply a sender
2081 name, $originator_name was incorrectly used when constructing a From:
2082 header. Furthermore, $originator_name was used for submission mode
2083 messages from external hosts without From: headers in a similar way,
2084 which is clearly wrong.
2086 PH/37 Added control=suppress_local_fixups.
2088 PH/38 When log_selector = +received_sender was set, and the addition of the
2089 sender made the log line's construction buffer exactly full, or one byte
2090 less than full, an overflow happened when the terminating "\n" was
2093 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2094 when the result of a list match is failure because a DNS lookup failed.
2096 PH/40 RM_COMMAND is now used in the building process.
2098 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2099 the "build-* directories that it finds.
2101 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2102 address in a domain literal was a prefix of an interface address.
2104 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2105 when verifying a sender address, unless rewrite_headers is false.
2107 PH/44 Wrote a long comment about why errors_to addresses are verified as
2108 recipients, not senders.
2110 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2111 the ratelimit ACL was added.
2113 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2115 PH/46 Added extra information about PostgreSQL errors to the error string.
2117 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2118 avoiding the need to install special zones in a real server. This is
2119 backwards compatible; if it can't find the fake resolver, it drops back.
2120 Thus, both old and new test suites can be run.
2122 TF/02 Added util/ratelimit.pl
2124 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2125 clock is set back in time.
2127 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2128 Candler <B.Candler@pobox.com>.
2130 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2131 OK for addresses that are the result of redirections.
2133 PH/48 A number of further additions for the benefit of the new test suite,
2134 including a fake gethostbyname() that interfaces to the fake DNS resolver
2137 TF/06 The fix for widen_domains has also been applied to qualify_single and
2138 search_parents which are the other dnslookup options that can cause
2141 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2144 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2146 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2147 service" if it encountered an MX record with an empty target hostname.
2148 The message is now "an MX or SRV record indicated no SMTP service".
2150 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2151 if the DNS of the sending site is misconfigured. This is quite a
2152 common situation. This change restores the behaviour of treating a
2153 helo verification defer as a failure.
2155 PH/51 If self=fail was set on a router, the bounce message did not include the
2156 actual error message.
2162 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2164 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2165 command itself ended in error, the underlying message about the transport
2166 filter timeout was being overwritten with the pipe command error. Now the
2167 underlying error message should be appended to the second error message.
2169 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2171 PH/02 Reduce the amount of output that "make" produces by default. Full output
2172 can still be requested.
2174 PH/03 The warning log line about a condition test deferring for a "warn" verb
2175 was being output only once per connection, rather than after each
2176 occurrence (because it was using the same function as for successful
2177 "warn" verbs). This seems wrong, so I have changed it.
2179 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2180 it should not have, which might have caused a crash in the right
2181 circumstances, but probably never did.
2183 PH/04 Installed a modified version of Tony Finch's patch to make submission
2184 mode fix the return path as well as the Sender: header line, and to
2185 add a /name= option so that you can make the user's friendly name appear
2188 TF/03 Added the control = fakedefer ACL modifier.
2190 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2191 Mark Lowes for thorough testing.
2193 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2195 TK/03 Merged latest SRS patch from Miles Wilton.
2197 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2198 with the definition in sysexits.h (which is #included earlier).
2199 Fortunately, Exim does not actually use EX_OK. The code used to try to
2200 preserve the sysexits.h value, by assumimg that macro definitions were
2201 scanned for macro replacements. I have been disabused of this notion,
2202 so now the code just undefines EX_OK before #including unistd.h.
2204 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2205 in the smtp transport. When a block could not be written in a single
2206 write() function, the timeout was being re-applied to each part-write.
2207 This seems wrong - if the receiver was accepting one byte at a time it
2208 would take for ever. The timeout is now adjusted when this happens. It
2209 doesn't have to be particularly precise.
2211 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2212 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2214 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2215 without a domain if the check to <postmaster@domain> fails.
2217 SC/01 Eximstats: added -xls and the ability to specify output files
2218 (patch written by Frank Heydlauf).
2220 SC/02 Eximstats: use FileHandles for outputing results.
2222 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2224 SC/04 Eximstats: fixed display of large numbers with -nvr option
2226 SC/05 Eximstats: fixed merging of reports with empty tables.
2228 SC/06 Eximstats: added the -include_original_destination flag
2230 SC/07 Eximstats: removed tabs and trailing whitespace.
2232 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2234 TK/06 MBOX spool code: Add real "From " MBOX separator line
2235 so the .eml file is really in mbox format (even though
2236 most programs do not really care). Patch from Alex Miller.
2238 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2239 The latter is generated from $received_to and is only set if the
2240 message has one envelope recipient. SA can use these headers,
2241 obviously out-of-the-box. Patch from Alex Miller.
2243 PH/08 The ${def test on a variable was returning false if the variable's
2244 value was "0", contrary to what the specification has always said!
2245 The result should be true unless the variable is empty.
2247 PH/09 The syntax error of a character other than { following "${if
2248 def:variable_name" (after optional whitespace) was not being diagnosed.
2249 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2250 accidental colon was present, for example, could give incorrect results.
2252 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2253 result is used (not including appendfile, where other changes are about
2256 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2257 This involved changing a lot of size variables from int to off_t. It
2258 should work with maildirs and everything.
2260 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2261 spamd dying while we are connected to it.
2263 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2266 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2267 function for BDB 4.3.
2269 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2271 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2272 the sources for PCRE have been re-arranged and more files are now
2275 PH/15 The code I had for printing potentially long long variables in PH/11
2276 above was not the best (it lost precision). The length of off_t variables
2277 is now inspected at build time, and an appropriate printing format (%ld
2278 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2279 to be "long long int" or "long int". This is needed for the internal
2280 formatting function string_vformat().
2282 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2283 the configuration file to be ":syslog", then the script "guesses" where
2284 the logs files are, rather than using the compiled in default. In our
2285 case the guess is not the same as the compiled default, so the script
2286 suddenly stopped working when I started to use syslog. The patch checks
2287 to see if log_file_path is "". If so, it attempts to read it from exim
2288 with no configuration file to get the compiled in version, before it
2289 falls back to the previous guessing code."
2291 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2292 implementing BATV in an Exim configuration. See NewStuff for the gory
2295 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2296 Makefile that are specific to HP-UX.
2298 PH/18 If the "use_postmaster" option was set for a recipient callout together
2299 with the "random" option, the postmaster address was used as the MAIL
2300 FROM address for the random test, but not for the subsequent recipient
2301 test. It is now used for both.
2303 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2304 patch removes a few documentation additions to RFC 3028, because the
2305 latest draft now contains them. It adds the new en;ascii-case comparator
2306 and a new error check for 8bit text in MIME parts. Comparator and
2307 require names are now matched exactly. I enabled the subaddress
2308 extension, but it is not well tested yet (read: it works for me)."
2310 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2311 rework some of the code of TK/09 above to avoid the hardwired use of
2312 "%lld" and "long long". Replaced the call to snprintf() with a call to
2315 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2316 records point to non-existent hosts", "retry timeout exceeded", and
2317 "retry time not reached for any host after a long failure period".
2319 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2320 experimental DomainKeys support:
2322 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2323 (2) On an error such as an illegally used "control", the wrong name for
2324 the control was given.
2326 These problems did NOT occur unless DomainKeys support was compiled.
2328 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2330 PH/24 Added ${if match_ip condition.
2332 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2333 databases so that it will be absolutely obvious if a crash occurs in the
2334 DB library. This is a regular occurrence (often caused by mis-matched
2337 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2338 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2339 picked up on a user's system that detects such things. There doesn't seem
2340 to be a gcc warning option for this - only an attribute that has to be
2341 put on the function's prototype. It seems that in Fedora Core 4 they have
2342 set this on a number of new functions. No doubt there will be more in due
2345 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2346 specify a transport. However, if an address that was verified by such a
2347 router was the subject of a callout, Exim crashed because it tried to
2348 read the rcpt_include_affixes from the non-existent transport. Now it
2349 just assumes that the setting of that option is false. This bug was
2350 introduced by 4.51/PH/31.
2352 PH/28 Changed -d+all to exclude +memory, because that information is very
2353 rarely of interest, but it makes the output a lot bigger. People tend to
2354 do -d+all out of habit.
2356 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2357 code in os-type was giving problems when libc.so lives in lib64, like on
2360 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2361 aren't the modern standard, and it seems that some systems' include files
2362 don't always have them. Exim was already checking for some of the newer
2363 ones like T_AAAA, and defining it itself. I've added checks for all the
2364 record types that Exim uses.
2366 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2367 not automatically generating a new one, as it is supposed to. This
2368 prevented TLS from working. If the file did exist, but contained invalid
2369 data, a new version was generated, as expected. It was only the case of a
2370 non-existent file that was broken.
2372 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2373 with a change in libdomainkeys > 0.64.
2375 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2376 from DNS. If the selector record carries the flag, it now has
2377 precedence over the domain-wide flag.
2379 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2381 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2382 the use of a port name that isn't defined in /etc/services) occurred, the
2383 message was deferred as in a normal delivery, and thus remained on the
2384 spool, instead of being failed because of the mua_wrapper setting. This
2385 is now fixed, and I tidied up some of the mua_wrapper messages at the
2388 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2389 the messages in a hash of arrays rather than using individual hashes.
2390 This is a bit cleaner and results in dramatic memory savings, albeit
2391 at a slight CPU cost.
2393 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2394 as requested by Marc Sherman.
2396 SC/10 Eximstats: added histograms for user specified patterns as requested
2399 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2401 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2402 fopen() in the content-scanning modules that did not already have it.
2408 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2409 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2411 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2413 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2415 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2416 to test Sieve filters that use "vacation".
2418 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2419 that changes the way the GnuTLS parameters are stored in the cache file.
2420 The new format can be generated externally. For backward compatibility,
2421 if the data in the cache doesn't make sense, Exim assumes it has read an
2422 old-format file, and it generates new data and writes a new file. This
2423 means that you can't go back to an older release without removing the
2426 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2427 work if there are any delivery delays because "one_time" forces the
2428 parent to be marked "delivered", so its unseen clone is never tried
2429 again. For this reason, Exim now forbids the simultaneous setting of
2432 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2433 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2434 entry if you want to know the details. The fix, however, neglected to
2435 consider the case when local delivery batching is involved. The test for
2436 "previously delivered" was not happening when checking to see if an
2437 address could be batched with a previous (undelivered) one; under
2438 certain circumstances this could lead to multiple deliveries to the same
2441 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2442 in its include files, and this causes problems building Exim.
2444 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2445 header_syntax) but Exim was just ignoring anything given after a slash.
2446 In particular, this caused confusion with an attempt to use "verify =
2447 reverse_host_lookup/defer_ok". An error is now given when options are
2448 supplied for verify items that do not have them. (Maybe reverse_host_
2449 lookup should have a defer_ok option, but that's a different point.)
2451 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2452 defined by RFC 821) to 2048, because there were problems with some AUTH
2453 commands, and RFC 1869 says the size should be increased for extended
2454 SMTP commands that take arguments.
2456 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2459 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2460 "unknown" error; now it says that the functionality isn't in the binary.
2462 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2463 an address' error message when a string expansion fails (syntax or
2464 whatever). Otherwise the password may appear in the log. Following change
2465 PH/42 below, there is no longer a chance of it appearing in a bounce
2468 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2470 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2471 of its addresses was ever tried. (Bugzilla bug #2.)
2473 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2474 the result incorrectly in the debug output. (It correctly added a newline
2475 to what was transported.)
2477 TF/01 Added $received_time.
2479 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2480 commented out examples of how to interface to a virus scanner and to
2481 SpamAssassin. Also added commented examples of av_scanner and
2482 spamd_address settings.
2484 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2485 and controls are allowed in which ACLs. There were a couple of minor
2486 errors. Some of the entries in the conditions table (which is a table of
2487 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2488 negation of where the condition IS allowed.
2490 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2492 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2493 header file does not have a version number, so I've had to invent a new
2494 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2495 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2496 radiusclient), but it was contributed by a Radius user.
2498 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2499 files or directories, for OpenSSL.
2501 PH/19 When an Exim process that is running as root has to create an Exim log
2502 file, it does so in a subprocess that runs as exim:exim so as to get the
2503 ownership right at creation (otherwise, other Exim processes might see
2504 the file with the wrong ownership). There was no test for failure of this
2505 fork() call, which would lead to the process getting stuck as it waited
2506 for a non-existent subprocess. Forks do occasionally fail when resources
2507 run out. I reviewed all the other calls to fork(); they all seem to check
2510 PH/20 When checking for unexpected SMTP input at connect time (before writing
2511 the banner), Exim was not dealing correctly with a non-positive return
2512 from the read() function. If the client had disconnected by this time,
2513 the result was a log entry for a synchronization error with an empty
2514 string after "input=" when read() returned zero. If read() returned -1
2515 (an event I could not check), uninitialized data bytes were printed.
2516 There were reports of junk text (parts of files, etc) appearing after
2519 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2521 PH/22 Added support for macro redefinition, and (re)definition in between
2522 driver and ACL definitions.
2524 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2525 forgetting to use the resulting value; it was using the unexpanded value.
2527 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2528 hadn't been configured. The fix is from Juergen Kreileder, who
2529 understands it better than I do:
2531 "Here's what I see happening with three configured cyrus_sasl
2532 authenticators configured (plain, login, cram-md5):
2534 On startup auth_cyrus_sasl_init() gets called for each of these.
2535 This means three calls to sasl_listmech() without a specified mech_list.
2536 => SASL tests which mechs of all available mechs actually work
2537 => three warnings about OTP not working
2538 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2540 With the patch, sasl_listmech() also gets called three times. But now
2541 SASL's mech_list option is set to the server_mech specified in the the
2542 authenticator. Or in other words, the answer from sasl_listmech()
2543 gets limited to just the mech you're testing for (which is different
2545 => the return list contains just 'plain' or 'login', 'cram-md5' or
2546 nothing depending on the value of ob->server_mech.
2548 I've just tested the patch: Authentication still works fine,
2549 unavailable mechs specified in the exim configuration are still
2550 caught, and the auth.log warnings about OTP are gone."
2552 PH/25 When debugging is enabled, the contents of the command line are added
2553 to the debugging output, even when log_selector=+arguments is not
2556 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2557 answer is "GNU", and only if the return is "GNU/something" is the answer
2560 PH/27 $acl_verify_message is now set immediately after the failure of a
2561 verification in an ACL, and so is available in subsequent modifiers. In
2562 particular, the message can be preserved by coding like this:
2564 warn !verify = sender
2565 set acl_m0 = $acl_verify_message
2567 Previously, $acl_verify_message was set only while expanding "message"
2568 and "log_message" when a very denied access.
2570 PH/28 Modified OS/os.c-Linux with
2572 -#ifndef OS_LOAD_AVERAGE
2573 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2575 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2576 nomenclature these days.)
2578 PH/29 Installed patch from the Sieve maintainer that adds the options
2579 sieve_useraddress and sieve_subaddress to the redirect router.
2581 PH/30 In these circumstances:
2582 . Two addresses routed to the same list of hosts;
2583 . First host does not offer TLS;
2584 . First host accepts first address;
2585 . First host gives temporary error to second address;
2586 . Second host offers TLS and a TLS session is established;
2587 . Second host accepts second address.
2588 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2589 and peerdn, if requested) that were in fact used only for the second
2592 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2593 attention to any local part prefix or suffix that was matched by the
2594 router that accepted the address. It now behaves in the same way as it
2595 does for delivery: the affixes are removed from the local part unless
2596 rcpt_include_affixes is set on the transport.
2598 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2599 timeout during the DATA phase of an incoming message.
2601 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2602 applied a patch sanctioned by the Sieve maintainer.
2604 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2605 the uid or gid is negative. A case of a negative gid caused this to be
2606 noticed. The fix allows for either to be negative.
2608 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2609 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2612 PH/36 The change PH/12 above was broken. Fixed it.
2614 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2615 the grounds that routing the same address twice would always produce the
2616 same answer. This might have been true once, but it is certainly no
2617 longer true now. Routing a child address may depend on the previous
2618 routing that produced that child. Some complicated redirection strategies
2619 went wrong when messages had multiple recipients, and made Exim's
2620 behaviour dependent on the order in which the addresses were given.
2622 I have moved the duplicate checking until after the routing is complete.
2623 Exim scans the addresses that are assigned to local and remote
2624 transports, and removes any duplicates. This means that more work will be
2625 done, as duplicates will always all be routed, but duplicates are
2626 presumably rare, so I don't expect this is of any significance.
2628 For deliveries to pipes, files, and autoreplies, the duplicate checking
2629 still happens during the routing process, since they are not going to be
2632 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2633 It corrects a timeout issue with spamd. This is Ian's comment: "The
2634 background is that sometimes spamd either never reads data from a
2635 connection it has accepted, or it never writes response data. The exiscan
2636 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2637 blindly assumes that writes won't block so it may never time out."
2639 PH/39 Allow G after quota size as well as K and M.
2641 PH/40 The value set for $authenticated_id in an authenticator may not contain
2642 binary zeroes or newlines because the value is written to log lines and
2643 to spool files. There was no check on this. Now the value is run through
2644 the string_printing() function so that such characters are converted to
2645 printable escape sequences.
2647 PH/41 $message_linecount is a new variable that contains the total number of
2648 lines in the message. Compare $body_linecount, which is the count for the
2651 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2652 bounce and delay warning messages, except in certain special cases, which
2655 (a) An SMTP error message from a remote host;
2656 (b) A message specified in a :fail: redirection;
2657 (c) A message specified in a "fail" command in a system filter;
2658 (d) A message specified in a FAIL return from the queryprogram router;
2659 (e) A message specified by the cannot_route_message router option.
2661 In these cases only, Exim does include the error details in bounce and
2662 warning messages. There are also a few cases where bland messages such
2663 as "unrouteable address" or "local delivery error" are given.
2665 PH/43 $value is now also set for the "else" part of a ${run expansion.
2667 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2668 being worked on, but at least Exim now implements the latest version to
2671 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2672 process to complete was treated as a delivery failure, a timeout while
2673 writing the message to the pipe was logged, but erroneously treated as a
2674 successful delivery. Such timeouts include transport filter timeouts. For
2675 consistency with the overall process timeout, these timeouts are now
2676 treated as errors, giving rise to delivery failures by default. However,
2677 there is now a new Boolean option for the pipe transport called
2678 timeout_defer, which, if set TRUE, converts the failures into defers for
2679 both kinds of timeout. A transport filter timeout is now identified in
2682 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2683 systems where "make" and "gmake" are different, calling "gmake" at top
2684 level broke things. I've arranged for the value of $(MAKE) to be passed
2685 from the Makefile to this script so that it can call the same version of
2689 A note about Exim versions 4.44 and 4.50
2690 ----------------------------------------
2692 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2693 changes of various kinds. As a consequence, a big documentation update was
2694 needed. This delayed the release for rather longer than seemed good, especially
2695 in the light of a couple of (minor) security issues. Therefore, the changes
2696 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2697 release. So 4.44 and 4.50 are in effect two different branches that both start
2700 I have left the 4.50 change log unchanged; it contains all the changes since
2701 4.43. The change log for 4.44 is below; many of its items are identical to
2702 those for 4.50. This seems to be the most sensible way to preserve the
2703 historical information.
2709 1. Minor wording change to the doc/README.SIEVE file.
2711 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2712 computation of the current number of files was incorrect.
2714 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2715 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2718 4. Give more explanation in the error message when the command for a transport
2719 filter fails to execute.
2721 5. There are several places where Exim runs a non-Exim command in a
2722 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2723 was being done only for the command run by the queryprogram router. It is
2724 now done for all such subprocesses. The other cases are: ${run, transport
2725 filters, and the commands run by the lmtp and pipe transports.
2727 6. Added CONFIGURE_GROUP build-time option.
2729 7. Some older OS have a limit of 256 on the maximum number of file
2730 descriptors. Exim was using setrlimit() to set 1000 as a large value
2731 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2732 systems. I've change it so that if it can't get 1000, it tries for 256.
2734 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2735 was an oversight, and furthermore, ever since the addition of extra
2736 controls (e.g. 4.43/32), the checks on when to allow different forms of
2737 "control" were broken. There should now be diagnostics for all cases when a
2738 control that does not make sense is encountered.
2740 9. Added the /retain_sender option to "control=submission".
2742 10. $recipients is now available in the predata ACL (oversight).
2744 11. Tidy the search cache before the fork to do a delivery from a message
2745 received from the command line. Otherwise the child will trigger a lookup
2746 failure and thereby defer the delivery if it tries to use (for example) a
2747 cached ldap connection that the parent has called unbind on.
2749 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2750 of $address_data from the recipient verification was clobbered by the
2751 sender verification.
2753 13. The value of address_data from a sender verification is now available in
2754 $sender_address_data in subsequent conditions in the ACL statement.
2756 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2758 15. Added a new option "connect=<time>" to callout options, to set a different
2761 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2762 was its contents. (It was OK if the option was not defined at all.)
2764 17. A "Completed" log line is now written for messages that are removed from
2765 the spool by the -Mrm option.
2767 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2768 information about exactly what failed.
2770 19. Added -dd to debug only the daemon process.
2772 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2773 handles timeouts, both on the server side and network timeouts. Renamed the
2774 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2776 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2777 It is now set to "smtps".
2779 22. $host_address is now set to the target address during the checking of
2780 ignore_target_hosts.
2782 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2783 being passed; this would have caused $sender_host_name to have been used if
2784 matching the list had actually called for a host name (not very likely,
2785 since this list is usually IP addresses). A host name is now passed as
2788 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2789 code that specifies a non-restarting handler (typically sigaction() in
2790 modern systems) in an attempt to fix a rare and obscure crash bug.
2792 25. Narrowed the window for a race in the daemon that could cause it to ignore
2793 SIGCHLD signals. This is not a major problem, because they are used only to
2794 wake it up if nothing else does.
2796 26. A malformed maildirsize file could cause Exim to calculate negative values
2797 for the mailbox size or file count. Odd effects could occur as a result.
2798 The maildirsize information is now recalculated if the size or filecount
2801 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2802 support for a long time. Removed HAVE_SYS_VFS_H.
2804 28. Installed the latest version of exipick from John Jetmore.
2806 29. In an address list, if the pattern was not a regular expression, an empty
2807 subject address (from a bounce message) matched only if the pattern was an
2808 empty string. Non-empty patterns were not even tested. This was the wrong
2809 because it is perfectly reasonable to use an empty address as part of a
2810 database query. An empty address is now tested by patterns that are
2811 lookups. However, all the other forms of pattern expect the subject to
2812 contain a local part and a domain, and therefore, for them, an empty
2813 address still always fails if the pattern is not itself empty.
2815 30. Exim went into a mad DNS loop when attempting to do a callout where the
2816 host was specified on an smtp transport, and looking it up yielded more
2817 than one IP address.
2819 31. Re-factored the code for checking spool and log partition space into a
2820 function that finds that data and another that does the check. The former
2821 is then used to implement four new variables: $spool_space, $log_space,
2822 $spool_inodes, and $log_inodes.
2824 32. The RFC2047 encoding function was originally intended for short strings
2825 such as real names; it was not keeping to the 75-character limit for
2826 encoded words that the RFC imposes. It now respects the limit, and
2827 generates multiple encoded words if necessary. To be on the safe side, I
2828 have increased the buffer size for the ${rfc2047: expansion operator from
2831 33. It is now permitted to omit both strings after an "if" condition; if the
2832 condition is true, the result is "true". As before, when the second string
2833 is omitted, a false condition yields an empty string. This makes it less
2834 cumbersome to write custom ACL and router conditions.
2836 34. Failure to deliver a bounce message always caused it to be frozen, even if
2837 there was an errors_to setting on the router. The errors_to setting is now
2840 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2841 canonical form (fully expanded) before being placed in
2842 $sender_host_address.
2844 36. The table in the code that translates DNS record types into text (T_A to
2845 "A" for instance) was missing entries for NS and CNAME. It is just possible
2846 that this could have caused confusion if both these types were looked up
2847 for the same domain, because the text type is used as part of Exim's
2848 per-process caching. But the chance of anyone hitting this buglet seems
2851 37. The dnsdb lookup has been extended in a number of ways.
2853 (1) There is a new type, "zns", which walks up the domain tree until it
2854 finds some nameserver records. It should be used with care.
2856 (2) There is a new type, "mxh", which is like "mx" except that it returns
2857 just the host names, not the priorities.
2859 (3) It is now possible to give a list of domains (or IP addresses) to be
2860 looked up. The behaviour when one of the lookups defers can be
2861 controlled by a keyword.
2863 (4) It is now possible to specify the separator character for use when
2864 multiple records are returned.
2866 38. The dnslists ACL condition has been extended: it is now possible to supply
2867 a list of IP addresses and/or domains to be looked up in a particular DNS
2870 39. Added log_selector=+queue_time_overall.
2872 40. When running the queue in the test harness, wait just a tad after forking a
2873 delivery process, to get repeatability of debugging output.
2875 41. Include certificate and key file names in error message when GnuTLS fails
2876 to set them up, because the GnuTLS error message doesn't include the name
2877 of the failing file when there is a problem reading it.
2879 42. Allow both -bf and -bF in the same test run.
2881 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2883 44. The "Exiscan patch" is now merged into the mainline Exim source.
2885 45. Sometimes the final signoff response after QUIT could fail to get
2886 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2887 < 0 before doing a fflush(). This bug looks as though it goes back to the
2888 introduction of TLS in release 3.20, but "sometimes" must have been rare
2889 because the tests only now provoked it.
2891 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2892 (this can affect the format of dates).
2894 47. exim_tidydb, when checking for the continued existence of a message for
2895 which it has found a message-specific retry record, was not finding
2896 messages that were in split spool directories. Consequently, it was
2897 deleting retry records that should have stayed in existence.
2899 48. Steve fixed some bugs in eximstats.
2901 49. The SPA authentication driver was not abandoning authentication and moving
2902 on to the next authenticator when an expansion was forced to fail,
2903 contradicting the general specification for all authenticators. Instead it
2904 was generating a temporary error. It now behaves as specified.
2906 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2907 (the order specifies the preference for clients). The order is now AES256,
2908 AES128, 3DES, ARCFOUR128.
2910 51. Small patch to Sieve code - explicitly set From: when generating an
2913 52. Exim crashed if a remote delivery caused a very long error message to be
2914 recorded - for instance if somebody sent an entire SpamAssassin report back
2915 as a large number of 550 error lines. This bug was coincidentally fixed by
2916 increasing the size of one of Exim's internal buffers (big_buffer) that
2917 happened as part of the Exiscan merge. However, to be on the safe side, I
2918 have made the code more robust (and fixed the comments that describe what
2921 53. Now that there can be additional text after "Completed" in log lines (if
2922 the queue_time_overall log selector is set), a one-byte patch to exigrep
2923 was needed to allow it to recognize "Completed" as not the last thing in
2926 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2927 patch that reportedly fixes this has been added. I am not expert enough to
2928 create a test for it. This is what the patch creator wrote:
2930 "I found a little strange behaviour of ldap code when working with
2931 Windows 2003 AD Domain, where users was placed in more than one
2932 Organization Units. When I tried to give exim partial DN, the exit code
2933 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2934 But simultaneously result of request was absolutely normal ldap result,
2935 so I produce this patch..."
2937 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2938 REFERENCE, so I have modified the code to exclude the patch when that macro
2941 55. Some experimental protocols are using DNS PTR records for new purposes. The
2942 keys for these records are domain names, not reversed IP addresses. The
2943 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2944 leaves it alone. Component reversal etc. now happens only for IP addresses.
2947 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2949 57. Double the size of the debug message buffer (to 2048) so that more of very
2950 long debug lines gets shown.
2952 58. The exicyclog utility now does better if the number of log files to keep
2953 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2955 59. Two changes related to the smtp_active_hostname option:
2957 (1) $smtp_active_hostname is now available as a variable.
2958 (2) The default for smtp_banner uses $smtp_active_hostname instead
2959 of $primary_hostname.
2961 60. The host_aton() function is supposed to be passed a string that is known
2962 to be a valid IP address. However, in the case of IPv6 addresses, it was
2963 not checking this. This is a hostage to fortune. Exim now panics and dies
2964 if the condition is not met. A case was found where this could be provoked
2965 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2966 components; fortuitously, this particular loophole had already been fixed
2967 by change 4.50/55 above.
2969 If there are any other similar loopholes, the new check in host_aton()
2970 itself should stop them being exploited. The report I received stated that
2971 data on the command line could provoke the exploit when Exim was running as
2972 exim, but did not say which command line option was involved. All I could
2973 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2974 running as the user.
2977 61. There was a buffer overflow vulnerability in the SPA authentication code
2978 (which came originally from the Samba project). I have added a test to the
2979 spa_base64_to_bits() function which I hope fixes it.
2982 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2983 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2985 63. The daemon start-up calls getloadavg() while still root for those OS that
2986 need the first call to be done as root, but it missed one case: when
2987 deliver_queue_load_max is set with deliver_drop_privilege. This is
2988 necessary for the benefit of the queue runner, because there is no re-exec
2989 when deliver_drop_privilege is set.
2991 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2992 This has been fixed.
2994 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2995 was in use, was not putting the data itself into the right store pool;
2996 consequently, it could be overwritten for a subsequent message in the same
2997 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3000 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3002 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3003 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3004 the function were treating the return as a boolean value, which happened to
3005 work because 0=false and not-0=true, but is not correct code.
3007 68. The host_aton() function was not handling scoped IPv6 addresses (those
3008 with, for example, "%eth0" on the end) correctly.
3010 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3011 negated items (that is, ~something) in unsigned ints. Some compilers
3012 apparently mutter when there is no cast.
3014 70. If an address verification called from an ACL failed, and did not produce a
3015 user-specific message (i.e. there was only a "system" message), nothing was
3016 put in $acl_verify_message. In this situation, it now puts the system
3019 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3020 session; change 4.31/43 added the unwanted input to the log line - except
3021 that it did not do this in the start of session case. It now does.
3023 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3024 This is wrong and can cause the other end to generate a synchronization
3025 error if it is another Exim or anything else that does the synchronization
3026 check. A QUIT command is no longer sent after a timeout.
3028 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3029 during host lookups.
3031 74. The defer_ok option of callout verification was not working if it was used
3032 when verifying addresses in header lines, that is, for this case:
3034 verify = header_sender/callout=defer_ok
3036 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3037 those file descriptors could be used for SMTP connections. If anything
3038 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3039 could be sent to the SMTP client, causing chaos. The daemon now opens
3040 stdin, stdout, and stderr to /dev/null when it puts itself into the
3043 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3044 log by default. The user can override this with suitable Perl magic.
3046 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3047 the log message when discard triggers, was not working for the DATA ACL or
3048 for the non-SMTP ACL.
3050 78. Error message wording change in sieve.c.
3052 79. If smtp_accept_max_per_host was set, the number of connections could be
3053 restricted to fewer than expected, because the daemon was trying to set up
3054 a new connection before checking whether the processes handling previous
3055 connections had finished. The check for completed processes is now done
3056 earlier. On busy systems, this bug wouldn't be noticed because something
3057 else would have woken the daemon, and it would have reaped the completed
3060 80. If a message was submitted locally by a user whose login name contained one
3061 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3062 It caused a spool format error. I have fixed the spool reading code. A
3063 related problem was that the "from" clause in the Received: line became
3064 illegal because of the space(s). It is now covered by ${quote_local_part.
3066 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3069 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3071 83. Patch from Sieve maintainer to fix unterminated string problem in
3072 "vacation" handling.
3074 84. Some minor changes to the Linux configuration files to help with other
3075 OS variants using glibc.
3077 85. One more patch for Sieve to update vacation handling to latest spec.
3080 ----------------------------------------------------
3081 See the note above about the 4.44 and 4.50 releases.
3082 ----------------------------------------------------
3088 1. Change 4.43/35 introduced a bug that caused file counts to be
3089 incorrectly computed when quota_filecount was set in an appendfile
3092 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3093 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3096 3. Give more explanation in the error message when the command for a transport
3097 filter fails to execute.
3099 4. There are several places where Exim runs a non-Exim command in a
3100 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3101 was being done only for the command run by the queryprogram router. It is
3102 now done for all such subprocesses. The other cases are: ${run, transport
3103 filters, and the commands run by the lmtp and pipe transports.
3105 5. Some older OS have a limit of 256 on the maximum number of file
3106 descriptors. Exim was using setrlimit() to set 1000 as a large value
3107 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3108 systems. I've change it so that if it can't get 1000, it tries for 256.
3110 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3111 was an oversight, and furthermore, ever since the addition of extra
3112 controls (e.g. 4.43/32), the checks on when to allow different forms of
3113 "control" were broken. There should now be diagnostics for all cases when a
3114 control that does not make sense is encountered.
3116 7. $recipients is now available in the predata ACL (oversight).
3118 8. Tidy the search cache before the fork to do a delivery from a message
3119 received from the command line. Otherwise the child will trigger a lookup
3120 failure and thereby defer the delivery if it tries to use (for example) a
3121 cached ldap connection that the parent has called unbind on.
3123 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3124 of $address_data from the recipient verification was clobbered by the
3125 sender verification.
3127 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3128 was its contents. (It was OK if the option was not defined at all.)
3130 11. A "Completed" log line is now written for messages that are removed from
3131 the spool by the -Mrm option.
3133 12. $host_address is now set to the target address during the checking of
3134 ignore_target_hosts.
3136 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3137 being passed; this would have caused $sender_host_name to have been used if
3138 matching the list had actually called for a host name (not very likely,
3139 since this list is usually IP addresses). A host name is now passed as
3142 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3143 code that specifies a non-restarting handler (typically sigaction() in
3144 modern systems) in an attempt to fix a rare and obscure crash bug.
3146 15. Narrowed the window for a race in the daemon that could cause it to ignore
3147 SIGCHLD signals. This is not a major problem, because they are used only to
3148 wake it up if nothing else does.
3150 16. A malformed maildirsize file could cause Exim to calculate negative values
3151 for the mailbox size or file count. Odd effects could occur as a result.
3152 The maildirsize information is now recalculated if the size or filecount
3155 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3156 support for a long time. Removed HAVE_SYS_VFS_H.
3158 18. Updated exipick to current release from John Jetmore.
3160 19. Allow an empty sender to be matched against a lookup in an address list.
3161 Previously the only cases considered were a regular expression, or an
3164 20. Exim went into a mad DNS lookup loop when doing a callout where the
3165 host was specified on the transport, if the DNS lookup yielded more than
3168 21. The RFC2047 encoding function was originally intended for short strings
3169 such as real names; it was not keeping to the 75-character limit for
3170 encoded words that the RFC imposes. It now respects the limit, and
3171 generates multiple encoded words if necessary. To be on the safe side, I
3172 have increased the buffer size for the ${rfc2047: expansion operator from
3175 22. Failure to deliver a bounce message always caused it to be frozen, even if
3176 there was an errors_to setting on the router. The errors_to setting is now
3179 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3180 canonical form (fully expanded) before being placed in
3181 $sender_host_address.
3183 24. Updated eximstats to version 1.33
3185 25. Include certificate and key file names in error message when GnuTLS fails
3186 to set them up, because the GnuTLS error message doesn't include the name
3187 of the failing file when there is a problem reading it.
3189 26. Expand error message when OpenSSL has problems setting up cert/key files.
3192 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3193 (this can affect the format of dates).
3195 28. exim_tidydb, when checking for the continued existence of a message for
3196 which it has found a message-specific retry record, was not finding
3197 messages that were in split spool directories. Consequently, it was
3198 deleting retry records that should have stayed in existence.
3200 29. eximstats updated to version 1.35
3201 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3202 1.35 - bugfix such that pie charts by volume are generated correctly
3204 30. The SPA authentication driver was not abandoning authentication and moving
3205 on to the next authenticator when an expansion was forced to fail,
3206 contradicting the general specification for all authenticators. Instead it
3207 was generating a temporary error. It now behaves as specified.
3209 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3210 (the order specifies the preference for clients). The order is now AES256,
3211 AES128, 3DES, ARCFOUR128.
3213 31. Small patch to Sieve code - explicitly set From: when generating an
3216 32. Exim crashed if a remote delivery caused a very long error message to be
3217 recorded - for instance if somebody sent an entire SpamAssassin report back
3218 as a large number of 550 error lines. This bug was coincidentally fixed by
3219 increasing the size of one of Exim's internal buffers (big_buffer) that
3220 happened as part of the Exiscan merge. However, to be on the safe side, I
3221 have made the code more robust (and fixed the comments that describe what
3224 33. Some experimental protocols are using DNS PTR records for new purposes. The
3225 keys for these records are domain names, not reversed IP addresses. The
3226 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3227 leaves it alone. Component reversal etc. now happens only for IP addresses.
3230 34. The host_aton() function is supposed to be passed a string that is known
3231 to be a valid IP address. However, in the case of IPv6 addresses, it was
3232 not checking this. This is a hostage to fortune. Exim now panics and dies
3233 if the condition is not met. A case was found where this could be provoked
3234 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3235 components; fortuitously, this particular loophole had already been fixed
3236 by change 4.50/55 or 4.44/33 above.
3238 If there are any other similar loopholes, the new check in host_aton()
3239 itself should stop them being exploited. The report I received stated that
3240 data on the command line could provoke the exploit when Exim was running as
3241 exim, but did not say which command line option was involved. All I could
3242 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3243 running as the user.
3246 35. There was a buffer overflow vulnerability in the SPA authentication code
3247 (which came originally from the Samba project). I have added a test to the
3248 spa_base64_to_bits() function which I hope fixes it.
3251 36. The daemon start-up calls getloadavg() while still root for those OS that
3252 need the first call to be done as root, but it missed one case: when
3253 deliver_queue_load_max is set with deliver_drop_privilege. This is
3254 necessary for the benefit of the queue runner, because there is no re-exec
3255 when deliver_drop_privilege is set.
3257 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3258 was in use, was not putting the data itself into the right store pool;
3259 consequently, it could be overwritten for a subsequent message in the same
3260 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3263 38. Sometimes the final signoff response after QUIT could fail to get
3264 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3265 < 0 before doing a fflush(). This bug looks as though it goes back to the
3266 introduction of TLS in release 3.20, but "sometimes" must have been rare
3267 because the tests only now provoked it.
3273 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3274 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3275 Now it returns an integer. A number of calls were still expecting a T/F
3276 return. Fortuitously, in all cases, the tests worked in OK situations,
3277 which is the norm. However, things would have gone wrong on any write
3278 failures on the smtp file descriptor. This function is used when sending
3279 messages over SMTP and also when doing verify callouts.
3281 2. When Exim is called to do synchronous delivery of a locally submitted
3282 message (the -odf or -odi options), it no longer closes stderr before doing
3285 3. Implemented the mua_wrapper option.
3287 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3289 5. Implemented the functions header_remove(), header_testname(),
3290 header_add_at_position(), and receive_remove_recipient(), and exported them
3293 6. If an ACL "warn" statement specified the addition of headers, Exim already
3294 inserted X-ACL-Warn: at the start if there was no header name. However, it
3295 was not making this test for the second and subsequent header lines if
3296 there were newlines in the string. This meant that an invalid header could
3297 be inserted if Exim was badly configured.
3299 7. Allow an ACL "warn" statement to add header lines at the start or after all
3300 the Received: headers, as well as at the end.
3302 8. Added the rcpt_4xx retry error code.
3304 9. Added postmaster_mailfrom=xxx to callout verification option.
3306 10. Added mailfrom=xxxx to the callout verification option, for verify=
3309 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3310 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3312 12. Inserted some casts to stop certain compilers warning when using pointer
3313 differences as field lengths or precisions in printf-type calls (mostly
3314 affecting debugging statements).
3316 13. Added optional readline() support for -be (dynamically loaded).
3318 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3319 same clock tick as a message's arrival, so that its received time was the
3320 same as the "first fail" time on the retry record, and that message
3321 remained on the queue past the ultimate address timeout, every queue runner
3322 would try a delivery (because it was past the ultimate address timeout) but
3323 after another failure, the ultimate address timeout, which should have then
3324 bounced the address, did not kick in. This was a "< instead of <=" error;
3325 in most cases the first failure would have been in the next clock tick
3326 after the received time, and all would be well.
3328 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3329 being recognized when the domain list was tested by the match_domain
3330 condition in an expansion string.
3332 16. Added the ${str2b64: operator.
3334 17. Exim was always calling setrlimit() to set a large limit for the number of
3335 processes, without checking whether the existing limit was already
3336 adequate. (It did check for the limit on file descriptors.) Furthermore,
3337 errors from getrlimit() and setrlimit() were being ignored. Now they are
3338 logged to the main and panic logs, but Exim does carry on, to try to do its
3339 job under whatever limits there are.
3341 18. Imported PCRE 5.0.
3343 19. Trivial typo in log message " temporarily refused connection" (the leading
3346 20. If the log selector return_path_on_delivery was set and an address was
3347 redirected to /dev/null, the delivery process crashed because it assumed
3348 that a return path would always be set for a "successful" delivery. In this
3349 case, the whole delivery is bypassed as an optimization, and therefore no
3352 21. Internal re-arrangement: the function for sending a challenge and reading
3353 a response while authentication was assuming a zero-terminated challenge
3354 string. It's now changed to take a pointer and a length, to allow for
3355 binary data in such strings.
3357 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3359 23. Exim was not respecting finduser_retries when seeking the login of the
3360 uid under which it was called; it was always trying 10 times. (The default
3361 setting of finduser_retries is zero.) Also, it was sleeping after the final
3362 failure, which is pointless.
3364 24. Implemented tls_on_connect_ports.
3366 25. Implemented acl_smtp_predata.
3368 26. If the domain in control=submission is set empty, Exim assumes that the
3369 authenticated id is a complete email address when it generates From: or
3370 Sender: header lines.
3372 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3373 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3374 chown and chgrp in /bin and hostname in /usr/bin.
3376 28. Exim was keeping the "process log" file open after each use, just as it
3377 does for the main log. This opens the possibility of it remaining open for
3378 long periods when the USR1 signal hits a daemon. Occasional processlog
3379 errors were reported, that could have been caused by this. Anyway, it seems
3380 much more sensible not to leave this file open at all, so that is what now
3383 29. The long-running daemon process does not normally write to the log once it
3384 has entered its main loop, and it closes the log before doing so. This is
3385 so that log files can straightforwardly be renamed and moved. However,
3386 there are a couple of unusual error situations where the daemon does write
3387 log entries, and I had neglected to close the log afterwards.
3389 30. The text of an SMTP error response that was received during a remote
3390 delivery was being truncated at 512 bytes. This is too short for some of
3391 the long messages that one sometimes sees. I've increased the limit to
3394 31. It is now possible to make retry rules that apply only when a message has a
3395 specific sender, in particular, an empty sender.
3397 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3398 it possible to be selective about when SMTP synchronization is enforced.
3400 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3402 32. Implemented hosts_connection_nolog.
3404 33. Added an ACL for QUIT.
3406 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3409 35. Added mailbox_size and mailbox_filecount to appendfile.
3411 36. Added control = no_multiline_responses to ACLs.
3413 37. There was a bug in the logic of the code that waits for the clock to tick
3414 in the case where the clock went backwards by a substantial amount such
3415 that the microsecond fraction of "now" was more than the microsecond
3416 fraction of "then" (but the whole seconds number was less).
3418 38. Added support for the libradius Radius client library this is found on
3419 FreeBSD (previously only the radiusclient library was supported).
3425 1. When certain lookups returned multiple values in the form name=value, the
3426 quoting of the values was not always being done properly. Specifically:
3427 (a) If the value started with a double quote, but contained no whitespace,
3429 (b) If the value contained whitespace other than a space character (i.e.
3430 tabs or newlines or carriage returns) it was not quoted.
3431 This fix has been applied to the mysql and pgsql lookups by writing a
3432 separate quoting function and calling it from the lookup code. The fix
3433 should probably also be applied to nisplus, ibase and oracle lookups, but
3434 since I cannot test any of those, I have not disturbed their existing code.
3436 2. A hit in the callout cache for a specific address caused a log line with no
3437 reason for rejecting RCPT. Now it says "Previous (cached) callout
3438 verification failure".
3440 3. There was an off-by-one bug in the queryprogram router. An over-long
3441 return line was truncated at 256 instead of 255 characters, thereby
3442 overflowing its buffer with the terminating zero. As well as fixing this, I
3443 have increased the buffer size to 1024 (and made a note to document this).
3445 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3446 when Exim is waiting for an SMTP response from a remote server, Exim
3447 restarts its select() call on the socket, thereby resetting its timeout.
3448 This is not a problem when such interrupts are rare. Somebody set up a cron
3449 job to run exiwhat every 2 minutes, which is less than the normal select()
3450 timeout (5 or 10 minutes). This meant that the select() timeout never
3451 kicked in because it was always reset. I have fixed this by comparing the
3452 time when an interrupt arrives with the time at the start of the first call
3453 to select(). If more time than the timeout has elapsed, the interrupt is
3454 treated as a timeout.
3456 5. Some internal re-factoring in preparation for the addition of Sieve
3457 extensions (by MH). In particular, the "personal" test is moved to a
3458 separate function, and given an option for scanning Cc: and Bcc: (which is
3459 not set for Exim filters).
3461 6. When Exim created an email address using the login of the caller as the
3462 local part (e.g. when creating a From: or Sender: header line), it was not
3463 quoting the local part when it contained special characters such as @.
3465 7. Installed new OpenBSD configuration files.
3467 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3468 try to make them clearer.
3470 9. Callout options, other than the timeout value, were being ignored when
3471 verifying sender addresses in header lines. For example, when using
3473 verify = header_sender/callout=no_cache
3475 the cache was (incorrectly) being used.
3477 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3478 only the Cygwin environment.
3480 11. When return_path_on_delivery was set as a log selector, if different remote
3481 addresses in the same message used different return paths and parallel
3482 remote delivery occurred, the wrong values would sometimes be logged.
3483 (Whenever a remote delivery process finished, the return path value from
3484 the most recently started remote delivery process was logged.)
3486 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3487 lines when AUTH and/or TLS are in use. This is the "received protocol"
3488 field. Exim used to use "asmtp" for authenticated SMTP, without any
3489 indication (in the protocol name) for TLS use. Now it follows the RFC and
3490 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3491 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3492 names appear in log lines as well as in Received: header lines.
3494 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3495 extensions, and comparison tests, and to fix some bugs.
3497 14. Changes to the "personal" filter test:
3499 (1) The test was buggy in that it was just doing the equivalent of
3500 "contains" tests on header lines. For example, if a user's address was
3501 anne@some.where, the "personal" test would incorrectly be true for
3503 To: susanne@some.where
3505 This test is now done by extracting each address from the header in turn,
3506 and checking the entire address. Other tests that are part of "personal"
3507 are now done using regular expressions (for example, to check local parts
3508 of addresses in From: header lines).
3510 (2) The list of non-personal local parts in From: addresses has been
3511 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3512 taken from the Sieve specification recommendations.
3514 (3) If the message contains any header line starting with "List-" it is
3515 treated as non-personal.
3517 (4) The test for "circular" in the Subject: header line has been removed
3518 because it now seems ill-conceived.
3520 15. Minor typos in src/EDITME comments corrected.
3522 16. Installed latest exipick from John Jetmore.
3524 17. If headers_add on a router specified a text string that was too long for
3525 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3526 of string_sprintf() is now avoided.
3528 18. $message_body_size was not set (it was always zero) when running the DATA
3529 ACL and the local_scan() function.
3531 19. For the "mail" command in an Exim filter, no default was being set for
3532 the once_repeat time, causing a random time value to be used if "once" was
3533 specified. (If the value happened to be <= 0, no repeat happened.) The
3534 default is now 0s, meaning "never repeat". The "vacation" command was OK
3535 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3536 (I found it when inspecting the code).
3538 20. There is now an overall timeout for performing a callout verification. It
3539 defaults to 4 times the callout timeout, which applies to individual SMTP
3540 commands during the callout. The overall timeout applies when there is more
3541 than one host that can be tried. The timeout is checked before trying the
3542 next host. This prevents very long delays if there are a large number of
3543 hosts and all are timing out (e.g. when the network connections are timing
3544 out). The value of the overall timeout can be changed by specifying an
3545 additional sub-option for "callout", called "maxwait". For example:
3547 verify = sender/callout=5s,maxwait=20s
3549 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3550 to the end before writing, but this should make it even safer).
3552 22. Exim was forgetting that it had advertised PIPELINING for the second and
3553 subsequent messages on an SMTP connection. It was also not resetting its
3554 memory on STARTTLS and an internal HELO.
3556 23. When Exim logs an SMTP synchronization error within a session, it now
3557 records whether PIPELINING has been advertised or not.
3559 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3560 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3561 rather than long int.
3563 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3565 26. Added the never_mail option to autoreply.
3571 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3572 crash if the getsockname() call failed; this can happen if a connection is
3573 closed very soon after it is established. The problem was simply in the
3574 order in which certain operations were done, causing Exim to try to write
3575 to the SMTP stream before it had set up the file descriptor. The bug has
3576 been fixed by making things happen in the correct order.
3582 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3583 before the connection was closed, thus losing the rejection response.
3585 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3586 some early Solaris releases, but causes trouble in current releases where
3587 socklen_t is defined.
3589 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3592 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3595 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3596 that releasing the top of it at the end releases what was used for sub-
3597 expansions (unless the block got too big). However, discard this block if
3598 the first thing is a variable or header, so that we can use its block when
3599 it is dynamic (useful for very large $message_headers, for example).
3601 6. Lookups now cache *every* query, not just the most recent. A new, separate
3602 store pool is used for this. It can be recovered when all lookup caches are
3603 flushed. Lookups now release memory at the end of their result strings.
3604 This has involved some general refactoring of the lookup sources.
3606 7. Some code has been added to the store_xxx() functions to reduce the amount
3607 of flapping under certain conditions.
3609 8. log_incoming_interface used to affect only the <= reception log lines. Now
3610 it causes the local interface and port to be added to several more SMTP log
3611 lines, for example "SMTP connection from", and rejection lines.
3613 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3615 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3617 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3618 could be overwritten at the end of the current message (or the start of a
3619 new message if it was set in a HELO ACL). The value is now preserved for
3620 the duration of the SMTP connection.
3622 12. If a transport had a headers_rewrite setting, and a matching header line
3623 contained an unqualified address, that address was qualified, even if it
3624 did not match any rewriting rules. The underlying bug was that the values
3625 of the flags that permit the existence of unqualified sender and recipient
3626 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3627 non-local messages, and by -bnq for local messages) were not being
3628 preserved with the message after it was received.
3630 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3631 "next input=" as part of the text comprising the host identity instead of
3632 the correct text. The code was using the same buffer for two different
3633 strings. However, depending on which order the printing function evaluated
3634 its arguments, the bug did not always show up. Under Linux, for example, my
3635 test suite worked just fine.
3637 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3638 doesn't work with some older versions of Perl. It has been changed to "my",
3639 which in any case is probably the better facility to use.
3641 15. A really picky compiler found some instances of statements for creating
3642 error messages that either had too many or two few arguments for the format
3645 16. The size of the buffer for calls to the DNS resolver has been increased
3646 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3647 for addresses that have a lot of PTR records. This alleviates a problem; it
3648 does not fully solve it.
3650 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3651 buffer now truncates the list and logs the incident, which is the same
3652 action as happens when Exim is looking up a host name and its aliases.
3653 Previously in this situation something unpredictable would happen;
3654 sometimes it was "internal error: store_reset failed".
3656 18. If a server dropped the connection unexpectedly when an Exim client was
3657 using GnuTLS and trying to read a response, the client delivery process
3658 crashed while trying to generate an error log message.
3660 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3661 string, for example:
3663 warn message = H1: something\nH2: something
3665 the text was added as a single header line from Exim's point of view
3666 though it ended up OK in the delivered message. However, searching for the
3667 second and subsequent header lines using $h_h2: did not work. This has been
3668 fixed. Similarly, if a system filter added multiple headers in this way,
3669 the routers could not see them.
3671 20. Expanded the error message when iplsearch is called with an invalid key to
3672 suggest using net-iplsearch in a host list.
3674 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3675 ACLs are no longer actually imposed (and a message to that effect is
3678 22. If a "gecos" field in a passwd entry contained escaped characters, in
3679 particular, if it contained a \" sequence, Exim got it wrong when building
3680 a From: or a Sender: header from that name. A second bug also caused
3681 incorrect handling when an unquoted " was present following a character
3682 that needed quoting.
3684 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3685 was not being matched caselessly.
3687 24. Arranged for all hyphens in the exim.8 source to be escaped with
3690 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3691 properly. Recipient callout cache records were still being keyed to include
3692 the sender, even when use_sender was set false. This led to far more
3693 callouts that were necessary. The sender is no longer included in the key
3694 when use_sender is false.
3696 26. Added "control = submission" modifier to ACLs.
3698 27. Added the ${base62d: operator to decode base 62 numbers.
3700 28. dnsdb lookups can now access SRV records.
3702 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3703 the configuration file.
3705 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3706 (-v) mode. This makes the output for a verbose queue run more intelligible.
3708 31. Added a use_postmaster feature to recipient callouts.
3710 32. Added the $body_zerocount variable, containing the number of binary zero
3711 bytes in the message body.
3713 33. The time of last modification of the "new" subdirectory is now used as the
3714 "mailbox time last read" when there is a quota error for a maildir
3717 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3719 35. Added +ignore_unknown as a special item in host lists.
3721 36. Code for decoding IPv6 addresses in host lists is now included, even if
3722 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3723 address was recognized as an IP address, but was then not correctly decoded
3724 into binary, causing unexpected and incorrect effects when compared with
3731 1. Very minor rewording of debugging text in manualroute to say "list of
3732 hosts" instead of "hostlist".
3734 2. If verify=header_syntax was set, and a header line with an unqualified
3735 address (no domain) and a large number of spaces between the end of the
3736 name and the colon was received, the reception process suffered a buffer
3737 overflow, and (when I tested it) crashed. This was caused by some obsolete
3738 code that should have been removed. The fix is to remove it!
3740 3. When running in the test harness, delay a bit after writing a bounce
3741 message to get a bit more predictability in the log output.
3743 4. Added a call to search_tidyup() just before forking a reception process. In
3744 theory, someone could use a lookup in the expansion of smtp_accept_max_
3745 per_host which, without the tidyup, could leave open a database connection.
3747 5. Added the variables $recipient_data and $sender_data which get set from a
3748 lookup success in an ACL "recipients" or "senders" condition, or a router
3749 "senders" option, similar to $domain_data and $local_part_data.
3751 6. Moved the writing of debug_print from before to after the "senders" test
3754 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3755 problems for message scanning, either using a data ACL, or using
3756 local_scan() because the Received: header was not generated till after they
3757 were called (in order to set the time as the time of reception completion).
3758 I have revised the way this works. The header is now generated after the
3759 body is received, but before the ACL or local_scan() are called. After they
3760 are run, the timestamp in the header is updated.
3766 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3767 before starting a queue runner without re-exec. This happened only when
3768 deliver_drop_privilege was set or when the Exim user was set to root. The
3769 effect of the bug was that timeouts during subsequent deliveries caused
3770 crashes instead of being properly handled. The handler is now left at its
3771 default (and expected) setting.
3773 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3774 message, again when deliver_drop_privilege is set or Exim is run as root.
3775 The bug described in (1) was not present in this case, but the tidying up
3776 of the other signals was missing. I have made the two cases consistent.
3778 3. The ignore_target_hosts setting on a manualroute router was being ignored
3779 for hosts that were looked up using the /MX notation.
3781 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3784 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3785 operated on the sender address. After changing the $sender_address to <>
3786 for the sender address verify, Exim was re-instated it as the original
3787 (before rewriting) address, but remembering that it had rewritten it, so it
3788 wasn't rewriting it again. This bug also had the effect of breaking the
3789 sender address verification caching when the sender address was rewritten.
3791 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3792 This has been changed so that if the ip literal address matches
3793 ignore_target_hosts, the router declines.
3795 7. Added expansion conditions match_domain, match_address, and match_local_
3796 part (NOT match_host).
3798 8. The placeholder for the Received: header didn't have a length field set.
3800 9. Added code to Exim itself and to exim_lock to test for a specific race
3801 condition that could lead to file corruption when using MBX delivery. The
3802 issue is with the lockfile that is created in /tmp. If this file is removed
3803 after a process has opened it but before that process has acquired a lock,
3804 there is the potential for a second process to recreate the file and also
3805 acquire a lock. This could lead to two Exim processes writing to the file
3806 at the same time. The added code performs the same test as UW imapd; it
3807 checks after acquiring the lock that its file descriptor still refers to
3808 the same named file.
3810 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3811 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3814 11. Added the smtp_active_hostname option. If used, this will typically be made
3815 to depend on the incoming interface address. Because $interface_address is
3816 not set up until the daemon has forked a reception process, error responses
3817 that can happen earlier (such as "too many connections") no longer contain
3820 12. If an expansion in a condition on a "warn" statement fails because a lookup
3821 defers, the "warn" statement is abandoned, and the next ACL statement is
3822 processed. Previously this caused the whole ACL to be aborted.
3824 13. Added the iplsearch lookup type.
3826 14. Added ident_timeout as a log selector.
3828 15. Added tls_certificate_verified as a log selector.
3830 16. Added a global option tls_require_ciphers (compare the smtp transport
3831 option of the same name). This controls incoming TLS connections.
3833 17. I finally figured out how to make tls_require_ciphers do a similar thing
3834 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3835 before starting the TLS session.
3837 18. Tabs are now shown as \t in -bP output.
3839 19. If the log selector return_path_on_delivery was set, Exim crashed when
3840 bouncing a message because it had too many Received: header lines.
3842 20. If two routers both had headers_remove settings, and the first one included
3843 a superfluous trailing colon, the final name in the first list and the
3844 first name in the second list were incorrectly joined into one item (with a
3845 colon in the middle).
3851 1. Added -C and -D options to the exinext utility, mainly to make it easier
3852 to include in the automated testing, but these could be helpful when
3853 multiple configurations are in use.
3855 2. The exinext utility was not formatting the output nicely when there was
3856 an alternate port involved in the retry record key, nor when there was a
3857 message id as well (for retries that were specific to a specific message
3858 and a specific host). It was also confused by IPv6 addresses, because of
3859 the additional colons they contain. I have fixed the IPv4 problem, and
3860 patched it up to do a reasonable job for IPv6.
3862 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3863 delivery, the log line now contains "pipelined" if PIPELINING was used.
3865 4. An SMTP transport process used to panic and die if the bind() call to set
3866 an explicit outgoing interface failed. This has been changed; it is now
3867 treated in the same way as a connect() failure.
3869 5. A reference to $sender_host_name in the part of a conditional expansion
3870 that was being skipped was still causing a DNS lookup. This no longer
3873 6. The def: expansion condition was not recognizing references to header lines
3874 that used bh_ and bheader_.
3876 7. Added the _cache feature to named lists.
3878 8. The code for checking quota_filecount in the appendfile transport was
3879 allowing one more file than it should have been.
3881 9. For compatibility with Sendmail, the command line option
3889 and sets the incoming protocol and host name (for trusted callers). The
3890 host name and its colon can be omitted when only the protocol is to be set.
3891 Note the Exim already has two private options, -pd and -ps, that refer to
3892 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3893 "s", but I don't think that's a major issue.
3895 10. A number of refactoring changes to the code, none of which should affect
3898 (a) The number of logging options was getting close to filling up the
3899 32-bit word that was used as a bit map. I have split them into two classes:
3900 those that are passed in the argument to log_write(), and those that are
3901 only ever tested independently outside of that function. These are now in
3902 separate 32-bit words, so there is plenty of room for expansion again.
3903 There is no change in the user interface or the logging behaviour.
3905 (b) When building, for example, log lines, the code previously used a
3906 macro that called string_cat() twice, in order to add two strings. This is
3907 not really sufficiently general. Furthermore, there was one instance where
3908 it was actually wrong because one of the argument was used twice, and in
3909 one call a function was used. (As it happened, calling the function twice
3910 did not affect the overall behaviour.) The macro has been replaced by a
3911 function that can join an arbitrary number of extra strings onto a growing
3914 (c) The code for expansion conditions now uses a table and a binary chop
3915 instead of a serial search (which was left over from when there were very
3916 few conditions). Also, it now recognizes conditions like "pam" even when
3917 the relevant support is not compiled in: a suitably worded error message is
3918 given if an attempt is made to use such a condition.
3920 11. Added ${time_interval:xxxxx}.
3922 12. A bug was causing one of the ddress fields not to be passed back correctly
3923 from remote delivery subprocesses. The field in question was not being
3924 subsequently used, so this caused to problems in practice.
3926 13. Added new log selectors queue_time and deliver_time.
3928 14. Might have fixed a bug in maildirsizefile handling that threw up
3929 "unexpected character" debug warnings, and recalculated the data
3930 unnecessarily. In any case, I expanded the warning message to give more
3933 15. Added the message "Restricted characters in address" to the statements in
3934 the default ACL that block characters like @ and % in local parts.
3936 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3937 Three changes have been made:
3939 (a) There was a serious bug; a negative response to MAIL caused the whole
3940 recipient domain to be cached as invalid, thereby blocking all messages
3941 to all local parts at the same domain, from all senders. This bug has
3942 been fixed. The domain is no longer cached after a negative response to
3943 MAIL if the sender used is not empty.
3945 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3948 (c) A new callout option, "use_sender" has been added for people who want
3949 the modified behaviour.
3955 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3958 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3959 indeed breaks things for older releases.
3961 3. Added additional logging to the case where there is a problem reading data
3962 from a filter that is running in a subprocess using a pipe, in order to
3963 try to track down a specific problem.
3965 4. Testing facility fudge: when running in the test harness and attempting
3966 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3967 getting "No route to host". Convert this to a timeout.
3969 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3972 6. Some OS don't have socklen_t but use size_t instead. This affects the
3973 fifth argument of getsockopt() amongst other things. This is now
3974 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3975 can be set for individual OS. I have set it for SunOS5, OSF1, and
3976 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3977 some earlier ones do not.
3979 7. Change 4.30/15 was not doing the test caselessly.
3981 8. The standard form for an IPv6 address literal was being rejected by address
3982 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3983 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3984 this, as well as the form without the "IPv6" on the front (but only when
3985 address literals are enabled, of course).
3987 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3989 10. Exim crashed if a message with an empty sender address specified by -f
3990 encountered a router with an errors_to setting. This could be provoked only
3991 by a command such as
3995 where an empty string was supplied; "<>" did not hit this bug.
3997 11. Installed PCRE release 4.5.
3999 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4000 remained set. It is now erased.
4002 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4003 times from message ids (which are base 36 rather than the normal 62).
4005 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4006 were being counted as actual protocol errors, and logged if the log
4007 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4008 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4009 and DATA following a set of rejected RCPTs do not count as protocol errors.
4010 In other words, Exim assumes they were pipelined, though this may not
4011 actually be the case. Of course, in all cases the client gets an
4012 appropriate error code.
4014 15. If a lookup fails in an ACL condition, a message about the failure may
4015 be available; it is used if testing the ACL cannot continue, because most
4016 such messages specify what the cause of the deferral is. However, some
4017 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4018 that caused an old message to be retained and used if a later statement
4019 caused a defer, replacing the real cause of the deferral.
4021 16. If an IP address had so many PTR records that the DNS lookup buffer
4022 was not large enough to hold them, Exim could crash while trying to process
4023 the truncated data. It now detects and logs this case.
4025 17. Further to 4.21/58, another change has been made: if (and only if) the
4026 first line of a message (the first header line) ends with CRLF, a bare LF
4027 in a subsequent header line has a space inserted after it, so as not to
4028 terminate the header.
4030 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4031 unnecessarily, used atoi() instead of strtol(), and didn't check the
4032 termination when getting file sizes from file names by regex.
4034 19. Completely re-implemented the support for maildirsize files, in the light
4035 of a number of problems with the previous contributed implementation
4036 (4.30/29). In particular:
4038 . If the quota is zero, the maildirsize file is maintained, but no quota is
4041 . If the maildir directory does not exist, it is created before any attempt
4042 to write a maildirsize file.
4044 . The quota value in the file is just a cache; if the quota is changed in
4045 the transport, the new value overrides.
4047 . A regular expression is available for excluding directories from the
4050 20. The autoreply transport checks the characters in options that define the
4051 message's headers; it allows continued headers, but it was checking with
4052 isspace() after an embedded newline instead of explicitly looking for a
4055 21. If all the "regular" hosts to which an address was routed had passed their
4056 expiry times, and had not reached their retry times, the address was
4057 bounced, even if fallback hosts were defined. Now Exim should go on to try
4060 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4061 equivalent code in the SMTP transport. Some hosts send humungous responses
4062 to HELO/EHLO, more than 1024 it seems.
4064 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4065 gives compiler warnings in some environments. I've now done it "properly",
4068 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4069 (because of problems with the built-in one) was declared to return uschar *
4070 instead of char *, causing compiler failure.
4072 25. Fixed a file descriptor leak when processing alias/forward files.
4074 26. Fixed a minor format string issue in dbfn.c.
4076 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4078 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4079 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4082 29. When a sender address is verified, it is cached, to save repeating the test
4083 when there is more than one recipient in a message. However, when the
4084 verification involves a callout, it is possible for different callout
4085 options to be set for different recipients. It is too complicated to keep
4086 track of this in the cache, so now Exim always runs a verification when a
4087 callout is required, relying on the callout cache for the optimization.
4088 The overhead is duplication of the address routing, but this should not be
4091 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4092 to be verified with callout=postmaster, and the main callout worked but the
4093 postmaster check failed, the verification correctly failed. However, if a
4094 subsequent RCPT command asked for sender verification *without* the
4095 postmaster check, incorrect caching caused this verification also to fail,
4098 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4099 it was not caching the DNS options (qualify_single, search_parents) that
4100 were used when the lookup failed. A subsequent lookup with different
4101 options therefore always gave the same answer, though there were cases
4102 where it should not have. (Example: a "domains = !$mx_any" option on a
4103 dnslookup router: the "domains" option is always processed without any
4104 widening, but the router might have qualify_single set.) Now Exim uses the
4105 cached value only when the same options are set.
4107 32. Added John Jetmore's "exipick" utility to the distribution.
4109 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4110 than a timeout (e.g. a certificate is required, and is not provided), an
4111 Exim server now closes the connection immediately. Previously it waited for
4112 the client to close - but if the client is SSL, it seems that they each
4113 wait for each other, leading to a delay before one of them times out.
4115 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4116 maintained 0.8.x compatibility because I don't think many are using it, and
4117 it is clearly obsolete.
4119 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4122 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4123 client certificate was expired. A simple patch fixes this, though I don't
4124 understand the full logic of why the verify callback is called multiple
4127 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4128 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4129 which causes problems with some clients (such as the Certicom SSL Plus
4130 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4131 disables the coutermeasure allowing Eudora to connect."
4133 38. Exim was not checking that a write() to a log file succeeded. This could
4134 lead to Bad Things if a log got too big, in particular if it hit a file
4135 size limit. Exim now panics and dies if it cannot write to a log file, just
4136 as it does if it cannot open a log file.
4138 39. Modified OS/Makefile-Linux so that it now contains
4140 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4142 The two -D definitions ensure that Exim is compiled with large file
4143 support, which makes it possible to handle log files that are bigger than
4146 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4147 instance) a domain was checked against a named list that involved a lookup,
4148 causing $domain_data to be set, then another domain was checked against the
4149 same list, then the first domain was re-checked, the value of $domain_data
4150 after the final check could be wrong. In particular, if the second check
4151 failed, it could be set empty. This bug probably also applied to
4154 41. The strip_trailing_dot option was not being applied to the address given
4155 with the -f command-line option.
4157 42. The code for reading a message's header from the spool was incrementing
4158 $received_count, but never initializing it. This meant that the value was
4159 incorrect (doubled) while delivering a message in the same process in which
4160 it was received. In the most common configuration of Exim, this never
4161 happens - a fresh exec is done - but it can happen when
4162 deliver_drop_privilege is set.
4164 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4165 it now includes up to 150 characters of the unexpected data in the log
4168 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4169 and building data strings. The size of both of these buffers was 10 000
4170 bytes - far larger than anybody would *ever* want, thought I. Needless to
4171 say, somebody hit the limit. I have increased the maximum line length to
4172 20 000 and the maximum data length of concatenated lines to 100 000. I have
4173 also fixed two bugs, because there was no checking on these buffers. Tsk,
4174 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4175 buffer is too small.
4177 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4178 lsearch lookups. Now it does.
4180 46. When parsing a route_list item in a manualroute router, a fixed-length
4181 buffer was used for the list of hosts. I made this 1024 bytes long,
4182 thinking that nobody would ever have a list of hosts that long. Wrong.
4183 Somebody had a whole pile of complicated expansion conditions, and the
4184 string was silently truncated, leading to an expansion error. It turns out
4185 that it is easier to change to an unlimited length (owing to other changes
4186 that have happened since this code was originally written) than to build
4187 structure for giving a limitation error. The length of the item that
4188 expands into the list of hosts is now unlimited.
4190 47. The lsearch lookup could not handle data where the length of text line was
4191 more than 4095 characters. Such lines were truncated, leading to shortened
4192 data being returned. It should now handle lines of any length.
4194 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4195 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4198 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4199 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4200 no longer happens in comment lines. A list of replacements is now placed
4201 at the head of all of the source files, except those whose only change is
4202 to replace PERL_COMMAND in the very first #! line.
4204 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4205 messages on the queue, with a bottom-up merge sort, using code contributed
4206 by Michael Haardt. This should make operations like -bp somewhat faster on
4207 large queues. It won't affect queue runners, except when queue_run_in_order
4210 51. Installed eximstats 1.31 in the distribution.
4212 52. Added support for SRV lookups to the dnslookup router.
4214 53. If an ACL referred to $message_body or $message_body_end, the value was not
4215 reset for any messages that followed in the same SMTP session.
4217 54. The store-handling optimization for building very long strings was not
4218 differentiating between the different store pools. I don't think this
4219 actually made any difference in practice, but I've tidied it.
4221 55. While running the routers to verify a sender address, $sender_address
4222 was still set to the sender address. This is wrong, because when routing to
4223 send a bounce to the sender, it would be empty. Therefore, I have changed
4224 it so that, while verifying a sender address, $sender_address is set to <>.
4225 (There is no change to what happens when verifying a recipient address.)
4227 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4228 target A or AAAA records (if not already returned) without resetting the
4229 qualify_single or search_parents options of the DNS resolver. These are
4230 inappropriate in this case because the targets of MX and SRV records must
4231 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4232 target that, when qualified, matched something in the local domain. These
4233 two options are now turned off when doing these lookups.
4235 57. It seems that at least some releases of Reiserfs (which does not have the
4236 concept of a fixed number of inodes) returns zero and not -1 for the
4237 number of available inodes. This interacted badly with check_spool_inodes,
4238 which assumed that -1 was the "no such thing" setting. What I have done is
4239 to check that the total number of inodes is greater than zero before doing
4240 the test of how many are available.
4242 58. When a "warn" ACL statement has a log_message modifier, the message is
4243 remembered, and not repeated. This is to avoid a lot of repetition when a
4244 message has many recipients that cause the same warning to be written.
4245 Howewer, Exim was preserving the list of already written lines for an
4246 entire SMTP session, which doesn't seem right. The memory is now reset if a
4247 new message is started.
4249 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4250 debugging output unless log_rewrite was also set.
4252 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4253 of (int)(handle) when we know that handle contains (void *)(-1).
4255 61. The Exim daemon panic-logs an error return when it closes the incoming
4256 connection. However "connection reset by peer" seems to be common, and
4257 isn't really an error worthy of noting specially, so that particular error
4260 62. When Exim is trying to find all the local interfaces, it used to panic and
4261 die if the ioctl to get the interface flags failed. However, it seems that
4262 on at least one OS (Solaris 9) it is possible to have an interface that is
4263 included in the list of interfaces, but for which you get a failure error
4264 for this call. This happens when the interface is not "plumbed" into a
4265 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4266 failure of the "get flags" call assumes that the interface is down.
4268 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4269 makes life easier for people who are doing arithmetic on fields extracted
4270 from dates, where you often get leading zeros that should not be
4271 interpreted as octal.
4273 64. Added qualify_domain to the redirect router, to override the global
4276 65. If a pathologically long header line contained very many addresses (the
4277 report of this problem mentioned 10 000) and each of them was rewritten,
4278 Exim could use up a very large amount of memory. (It kept on making new
4279 copies of the header line as it rewrote, and never released the old ones.)
4280 At the expense of a bit more processing, the header rewriting function has
4281 been changed so that it no longer eats memory in this way.
4283 66. The generation of the Received: header has been moved from the time that a
4284 message starts to be received, to the time that it finishes. The timestamp
4285 in the Received: header should now be very close to that of the <= log
4286 line. There are two side-effects of this change:
4288 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4289 logged header lines no longer include the local Received: line, because
4290 it has not yet been created. The same applies to a copy of the message
4291 that is returned to a non-SMTP sender when a message is rejected.
4293 (b) When a filter file is tested using -bf, no additional Received: header
4294 is added to the test message. After some thought, I decided that this
4297 This change does not affect the value of $received_for. It is still set
4298 after address rewriting, but before local_scan() is called.
4300 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4302 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4303 gave an unhelpful panic error message, and a defer error. I have managed to
4304 change this behaviour so that it now rejects any supplied certificate,
4305 which seems right, as the list of acceptable certificates is empty.
4307 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4308 gave an unhelpful defer error. I have not managed to make this reject any
4309 supplied certificates, but the error message it gives is "no certificate
4310 supplied", which is not helpful.
4312 70. exigrep's output now also includes lines that are not associated with any
4313 message, but which match the given pattern. Implemented by a patch from
4314 Martin Sluka, which also tidied up the Perl a bit.
4316 71. Recipient callout verification, like sender verification, was using <> in
4317 the MAIL FROM command. This isn't really the right thing, since the actual
4318 sender may affect whether the remote host accepts the recipient or not. I
4319 have changed it to use the actual sender in the callout; this means that
4320 the cache record is now keyed on a recipient/sender pair, not just the
4321 recipient address. There doesn't seem to be a real danger of callout loops,
4322 since a callout by the remote host to check the sender would use <>.
4323 [SEE ABOVE: changed after hitting problems.]
4325 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4326 temporary errors. However, in the case of such a code being given after
4327 the end of a data transmission (i.e. after ".") Exim was failing to write
4328 a retry record for the message. (Yes, there was some broken host that was
4329 actually sending 8xx at this point.)
4331 73. An unknown lookup type in a host list could cause Exim to panic-die when
4332 the list was checked. (An example that provoked this was putting <; in the
4333 middle of a list instead of at the start.) If this happened during a DATA
4334 ACL check, a -D file could be left lying around. This kind of configuration
4335 error no longer causes Exim to die; instead it causes a defer errror. The
4336 incident is still logged to the main and panic logs.
4338 74. Buglet left over from Exim 3 conversion. The message "too many messages
4339 in one connection" was written to the rejectlog but not the mainlog, except
4340 when address rewriting (yes!) was being logged.
4342 75. Added write_rejectlog option.
4344 76. When a system filter was run not as root (that is, when system_filter_user
4345 was set), the values of the $n variables were not being returned to the
4346 main process; thus, they were not subsequently available in the $sn
4349 77. Added +return_path_on_delivery log selector.
4351 78. A connection timeout was being treated differently from recipients deferred
4352 when testing hosts_max_try with a message that was older than the host's
4353 retry timeout. (The host should not be counted, thus allowing all hosts to
4354 be tried at least once before bouncing.) This may have been the cause of an
4355 occasionally reported bug whereby a message would remain on the queue
4356 longer than the retry timeout, but would be bounced if a delivery was
4357 forced. I say "may" because I never totally pinned down the problem;
4358 setting up timeout/retry tests is difficult. See also the next item.
4360 79. The ultimate address timeout was not being applied to errors that involved
4361 a combination of host plus message (for example, a timeout on a MAIL
4362 command). When an address resolved to a number of possible hosts, and they
4363 were not all tried for each delivery (e.g. because of hosts_max_try), a
4364 message could remain on the queue longer than the retry timeout.
4366 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4369 81. Fixed an obscure SMTP outgoing bug which required at least the following
4370 conditions: (a) there was another message waiting for the same server;
4371 (b) the server returned 5xx to all RCPT commands in the first message so
4372 that the message was not completed; (c) the server dropped the connection
4373 or gave a negative response to the RSET that Exim sends to abort the
4374 transaction. The observed case was a dropped connection after DATA that had
4375 been sent in pipelining mode. That is, the server had advertised PIPELINING
4376 but was not implementing it correctly. The effect of the bug was incorrect
4377 behaviour, such as trying another host, and this could lead to a crash.
4383 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4384 and daemon.c were passed as pointers to ints; they should have been
4385 pointers to socklen_t variables (which are typically unsigned ints).
4387 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4390 3. Fixed a really odd bug that affected only the testing scheme; patching a
4391 certain fixed string in the binary changed the value of another string that
4392 happened to be identical to the end of the original first string.
4394 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4395 name", it returns that address as the IP address. On some operating
4396 systems (e.g. Solaris), it also passes back the IP address string as the
4397 "host name". However, on others (e.g. Linux), it passes back an empty
4398 string. Exim wasn't checking for this, and was changing the host name to an
4399 empty string, assuming it had been canonicized.
4401 5. Although rare, it is permitted to have more than one PTR record for a given
4402 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4403 all the names associated with an address, because they do in Solaris.
4404 However, it seems that they do not in Linux for data that comes from the
4405 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4406 I found this out when I moved to a new Linux workstation and tried to run
4407 the Exim test suite.
4409 To get round this problem I have changed the code so that it now does its
4410 own call to the DNS to look up PTR records when searching for a host name.
4411 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4412 addresses that are only in /etc/hosts are still found.
4414 This behaviour is, however, controlled by an option called host_lookup_
4415 order, which defaults to "bydns:byaddr". If people want to use the other
4416 order, or indeed, just use one or the other means of lookup, they can
4417 specify it in this variable.
4419 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4420 some operating systems, this comes back from gethostbyaddr() as an empty
4421 string, and this is what Exim used to test for. However, it seems that in
4422 other systems, "." is yielded. Exim now tests for this case too.
4424 7. The values of check_spool_space and check_log_space are now held internally
4425 as a number of kilobytes instead of an absolute number of bytes. If a
4426 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4427 kilobyte. This means that much larger values can be stored.
4429 8. Exim monitor: an attempt to get the action menu when not actually pointing
4430 at a message produces an empty menu entitled "No message selected". This
4431 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4432 no entries in it ("Shell widget menu has zero width and/or height"). So I
4433 have added a single, blank menu entry in this case.
4435 9. Added ${quote_local_part.
4437 10. MIME decoding is now applied to the contents of Subject: header lines when
4440 11. Now that a reference to $sender_host_address automatically causes a reverse
4441 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4442 host lookup before query-style lookups in lists that might use this
4443 variable. This has therefore been abolished, and the "net-" prefix is no
4444 longer necessary for query-style lookups.
4446 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4447 have been a typo for LFLAGS, so it has been changed.
4449 13. The install script calls Exim with "-C /dev/null" in order to find the
4450 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4451 to be output. Howeve, since Exim outputs its version number before the
4452 error, it didn't break the script. It just looked ugly. I fixed this by
4453 always allowing "-C /dev/null" if the caller is root.
4455 14. Ignore overlarge ACL variable number when reading spool file - insurance
4456 against a later release with more variables having written the file.
4458 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4459 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4460 this, as well as the form without the "IPv6" on the front.
4462 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4463 OS/Makefile-Darwin file.
4465 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4466 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4468 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4469 variable. However, some OS use ints for inodes. I've added cast to long int
4470 to get rid of the compiler warning.
4472 19. I had forgotten to lock out "/../" in configuration file names when
4473 ALT_CONFIG_PREFIX was set.
4475 20. Routers used for verification do not need to specify transports. However,
4476 if such a router generated a host list, and callout was configured, Exim
4477 crashed, because it could not find a port number from the (non-existent)
4478 transport. It now assumes port 25 in this circumstance.
4480 21. Added the -t option to exigrep.
4482 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4483 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4484 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4485 them is set, LOOKUP_LSEARCH is forced.
4487 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4489 24. Added sender and host information to the "rejected by local_scan()" log
4490 line; previously there was no indication of these.
4492 25. Added .include_if_exists.
4494 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4495 Linux. It turns out that not all file systems support this. Apparently some
4496 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4497 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4498 not supported on the file descriptor, is now ignored when Exim is trying to
4499 sync a directory. This applies only to Linux.
4501 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4503 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4504 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4507 29. Added support for maildirsize files from supplied patch (modified a bit).
4509 30. The use of :fail: followed by an empty string could lead Exim to respond to
4510 sender verification failures with (e.g.):
4512 550 Verification failed for <xxx>
4513 550 Sender verify failed
4515 where the first response line was missing the '-' that indicates it is not
4516 the final line of the response.
4518 31. The loop for finding the name of the user that called Exim had a hardwired
4519 limit of 10; it now uses the value of finduser_retries, which is used for
4520 all other user lookups.
4522 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4525 33. Exim was neglecting to zero errno before one call of strtol() when
4526 expanding a string and expecting an integer value. On some systems this
4527 resulted in spurious "integer overflow" errors. Also, it was casting the
4528 result into an int without checking.
4530 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4531 did not work. The code looks as if it has *never* worked, though it appears
4532 to have been documented since at least releast 1.62. I have made it work.
4534 35. The "timeout_DNS" error in retry rules, also documented since at least
4535 1.62, also never worked. As it isn't clear exactly what this means, and
4536 clearly it isn't a major issue, I have abolished the feature by treating it
4537 as "timeout", and writing a warning to the main and panic logs.
4539 36. The display of retry rules for -brt wasn't always showing the error code
4542 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4543 timeout_connect_A, timeout_connect_MX.
4545 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4546 to the empty sender.
4548 39. The daemon was not analysing the content of -oX till after it had closed
4549 stderr and disconnected from the controlling terminal. This meant that any
4550 syntax errors were only noted on the panic log, and the return code from
4551 the command was 0. By re-arranging the code a little, I've made the
4552 decoding happen first, so such errors now appear on stderr, and the return
4553 code is 1. However, the actual setting up of the sockets still happens in
4554 the disconnected process, so errors there are still only recorded on the
4557 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4558 connections (as happens on some IP stacks) was logged at start up time as
4559 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4560 it from "IPv6 and IPv4", which means that two separate sockets are being
4563 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4564 says whether AF_INET or AF_INET6 was passed as an argument.
4566 42. Exiwhat output was messed up when time zones were included in log
4569 43. Exiwhat now gives more information about the daemon's listening ports,
4570 and whether -tls-on-connect was used.
4572 44. The "port" option of the smtp transport is now expanded.
4574 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4575 silently ignored. Now an error message is written to the main and panic
4578 46. There's a new ACL modifier called "logwrite" which writes to a log file
4579 as soon as it is encountered.
4581 47. Added $local_user_uid and $local_user_gid at routing time.
4583 48. Exim crashed when trying to verify a sender address that was being
4586 49. Exim was recognizing only a space character after ".include". It now also
4587 recognizes a tab character.
4589 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4590 extracting the relevant information from the specification. The man page no
4591 longer contains scrambled data for the -d option, and I've added a section
4592 at the front about calling Exim under different names.
4594 51. Added "extra_headers" argument to the "mail" command in filter files.
4596 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4599 53. Installed eximstats 1.29.
4601 54. Added transport_filter_timeout as a generic transport option.
4603 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4604 Cc: header lines. This was required by RFC 822, but it not required by RFC
4607 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4608 incoming messages that did not have them. Now it does so only if the
4609 message originates locally, that is, if there is no associated remote host
4610 address. When Resent- header lines are present, this applies to the Resent-
4611 lines rather than the non-Resent- lines.
4613 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4614 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4616 58. Messages for configuration errors now include the name of the main
4617 configuration file - useful now that there may be more than one file in a
4618 list (.included file names were always shown).
4620 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4621 for those rare installations that do not start the daemon as root or run it
4622 setuid root. I've cut out the call to initgroups() if the daemon is not
4625 60. The Exim user and group can now be bound into the binary as text strings
4626 that are looked up at the start of Exim's processing.
4628 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4630 62. Added $mailstore_basename variable.
4632 63. Installed patch to sieve.c from Michael Haardt.
4634 64. When Exim failed to open the panic log after failing to open the main log,
4635 the original message it was trying to log was written to stderr and debug
4636 output, but if they were not available (the usual case in production), it
4637 was lost. Now it is written to syslog before the two lines that record the
4638 failures to open the logs.
4640 65. Users' Exim filters run in subprocesses under the user's uid. It is
4641 possible for a "deliver" command or an alias in a "personal" command to
4642 provoke an address rewrite. If logging of address rewriting is configured,
4643 this fails because the process is not running as root or exim. There may be
4644 a better way of dealing with this, but for the moment (because 4.30 needs
4645 to be released), I have disabled address rewrite logging when running a
4646 filter in a non-root, non-exim process.
4652 1. The buildconfig auxiliary program wasn't quoting the value set for
4653 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4654 not defined. This bug was masked in 4.22 by the effect that was fixed in
4657 2. Some messages that were rejected after a message id was allocated were
4658 shown as "incomplete" by exigrep. It no longer does this for messages that
4659 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4661 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4662 have allow_domain_literals set, the ID did not get logged in the <= line.
4663 Domain literals are now always recognized in Message-ID: header lines.
4665 4. The first argument for a ${extract expansion item is the key name or field
4666 number. Leading and trailing spaces in this item were not being ignored,
4667 causing some misleading effects.
4669 5. When deliver_drop_privilege was set, single queue runner processes started
4670 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4671 same command in the process it spins off) were not dropping privilege.
4673 6. When the daemon running as "exim" started a queue runner, it always
4674 re-executed Exim in the spun-off process. This is a waste of effort when
4675 deliver_drop_privilege is set. The new process now just calls the
4676 queue-runner function directly.
4682 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4685 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4686 ignored. Though the use of -oP was forcing the writing of a pid file, it
4687 was always written to the default place.
4689 3. If the message "no IP address found for host xxxx" is generated during
4690 incoming verification, it is now followed by identification of the incoming
4691 connection (so you can more easily find what provoked it).
4693 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4695 5. Added some features to "harden" Exim a bit more against certain attacks:
4697 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4698 be put in Local/Makefile. This is like the never_users runtime option,
4699 but it cannot be overridden. The default setting is "root".
4701 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4702 prefix string with which any file named in a -C command line option
4705 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4706 is retained for -C and -D only if the caller of Exim is root. Without
4707 it, the exim user may also use -C and -D and retain privilege.
4709 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4710 command line option is disabled.
4712 6. Macro names set by the -D option must start with an upper case letter, just
4713 like macro names defined in the configuration file.
4715 7. Added "dereference=" facility to LDAP.
4717 8. Two instances of the typo "uknown" in the source files are fixed.
4719 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4720 the Configure-Makefile script screwed up while processing it.
4722 10. Incorporated PCRE 4.4.
4724 11. The SMTP synchronization check was not operating right at the start of an
4725 SMTP session. For example, it could not catch a HELO sent before the client
4726 waited for the greeting. There is now a check for outstanding input at the
4727 point when the greeting is written. Because of the duplex, asynchronous
4728 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4729 way, but not yet received, when the check is performed.
4731 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4732 on TCP/IP sockets, because this apparently causes some broken clients to
4735 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4736 unchanged) from the Cygwin maintainer.
4738 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4739 was supported instead of testing for mbx. Effectively a typo.
4741 15. The spa authenticator server code was not checking that the input it
4742 received was valid base64.
4744 16. The debug output line for the "set" modifier in ACLs was not showing the
4745 name of the variable that was being set.
4747 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4749 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4750 lookup on its own. Something else was needed to trigger the lookup. For
4751 example, a match in host_lookup or the need for a host name in a host list.
4752 Now, if $sender_host_name is referenced and the host name has not yet been
4753 looked up, a lookup is performed. If the lookup fails, the variable remains
4754 empty, and $host_lookup_failed is set to "1".
4756 19. Added "eqi" as a case-independent comparison operator.
4758 20. The saslauthd authentication condition could segfault if neither service
4759 nor realm was specified.
4761 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4762 error message that was logged was misleading, and incoming SMTP
4763 connections were dropped. The message is now more accurate, and temporary
4764 errors are given to SMTP connections.
4766 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4767 (except RSET) with a 421 error, until QUIT is received. However, it was
4768 failing to send a response to QUIT.
4770 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4771 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4772 ACL is now run after the helo_try_verify_hosts code.
4774 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4775 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4776 case-independent, so other case variants are also recognized). Apparently
4777 some systems use these upper case variants.
4779 25. If more than two messages were waiting for the same host, and a transport
4780 filter was specified for the transport, Exim sent two messages over the
4781 same TCP/IP connection, and then failed with "socket operation on non-
4782 socket" when it tried to send the third.
4784 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4786 27. The extern definition of crypt16() in expand.c was not being excluded when
4787 the OS had its own crypt16() function.
4789 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4790 as a preferred synonym for return_size_limit, both as an option and as an
4793 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4795 30. Changed the default configuration ACL to relax the local part checking rule
4796 for addresses that are not in any local domains. For these addresses,
4797 slashes and pipe symbols are allowed within local parts, but the sequence
4798 /../ is explicitly forbidden.
4800 31. SPA server authentication was not clearing the challenge buffer before
4803 32. log_message in a "warn" ACL statement was writing to the reject log as
4804 well as to the main log, which contradicts the documentation and doesn't
4805 seem right (because no rejection is happening). So I have stopped it.
4807 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4808 However, I am unable to do any testing of this.
4810 34. Fixed an infelicity in the appendfile transport. When checking directories
4811 for a mailbox, to see if any needed to be created, it was accidentally
4812 using path names with one or more superfluous leading slashes; tracing
4813 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4815 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4816 contents are added to the log line that is written for every discarded
4817 recipient. (Previously a log_message setting was ignored.)
4819 36. The ${quote: operator now quotes the string if it is empty.
4821 37. The install script runs exim in order to find its version number. If for
4822 some reason other than non-existence or emptiness, which it checks, it
4823 could not run './exim', it was installing it with an empty version number,
4824 i.e. as "exim-". This error state is now caught, and the installation is
4827 38. An argument was missing from the function that creates an error message
4828 when Exim fails to connect to the socket for saslauthd authentication.
4829 This could cause Exim to crash, or give a corrupted message.
4831 39. Added isip, isip4, and isip6 to ${if conditions.
4833 40. The ACL variables $acl_xx are now saved with the message, and can be
4834 accessed later in routers, transports, and filters.
4836 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4837 strings in the file are not string-expanded.
4839 42. If a MAIL command specified a SIZE value that was too large to fit into an
4840 int variable, the check against message_size_limit failed. Such values are
4841 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4842 day this will have to be increased, but I don't think I want to be around
4843 when emails are that large.
4850 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4851 iconv() is not standard in FreeBSD.
4853 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4854 IPv6 enabled. The observed symptom was a segmentation fault on return from
4855 the function os_common_find_running_interfaces() in src/os.c.
4857 3. In the check_special_case() function in daemon.c I had used "errno" as an
4858 argument name, which causes warnings on some systems. This was basically a
4859 typo, since it was named "eno" in the comments!
4861 4. The code that waits for the clock to tick (at a resolution of some fraction
4862 of a second) so as to ensure message-id uniqueness was always waiting for
4863 at least one whole tick, when it could have waited for less. [This is
4864 almost certainly not relevant at current processor speeds, where it is
4865 unlikely to ever wait at all. But we try to future-proof.]
4867 5. The function that sleeps for a time interval that includes fractions of a
4868 second contained a race. It did not block SIGALRM between setting the
4869 timer, and suspending (a couple of lines later). If the interval was short
4870 and the sigsuspend() was delayed until after it had expired, the suspension
4871 never ended. On busy systems this could lead to processes getting stuck for
4874 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4875 process, before it forks any delivery processes. The open lookup caching
4876 mechanism meant that the open file or database connection was passed into
4877 the delivery process. The problem was that delivery processes always tidy
4878 up cached lookup data. This could cause a problem for the next delivery
4879 process started by the queue runner, because the external queue runner
4880 process does not know about the closure. So the next delivery process
4881 still has data in the lookup cache. In the case of a file lookup, there was
4882 no problem because closing a file descriptor in a subprocess doesn't affect
4883 the parent. However, if the lookup was caching a connection to a database,
4884 the connection was closed, and the second delivery process was likely to
4885 see errors such as "PGSQL: query failed: server closed the connection
4886 unexpectedly". The problem has been fixed by closing all cached lookups
4887 in a queue runner before running a delivery process.
4889 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4890 seem to have the "const" qualifier which it has on other OS. I've
4893 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4894 *of the same type* (client or server) with the same public name that an
4895 error should be diagnosed.
4897 9. When Exim looked up a host name for an IP address, but failed to find the
4898 original IP address when looking up the host name (a safety check), it
4899 output the message "<ip address> does not match any IP for NULL", which was
4900 confusing, to say the least. The bug was that the host name should have
4901 appeared instead of "NULL".
4903 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4904 user that is built into the binary, and the -C or -D options is used, root
4905 privilege is dropped before the configuration file is read. In addition,
4906 logging is switched to stderr instead of the normal log files. If the
4907 configuration then re-defines the Exim user, the unprivileged environment
4908 is probably not what is expected, so Exim logs a panic warning message (but
4911 However, if deliver_drop_privilege is set, the unprivileged state may well
4912 be exactly what is intended, so the warning has been cut out in that case,
4913 and Exim is allowed to try to write to its normal log files.
4919 1. smtp_return_error_details was not giving details for temporary sender
4920 or receiver verification errors.
4922 2. Diagnose a configuration error if two authenticators have the same public
4925 3. Exim used not to create the message log file for a message until the first
4926 delivery attempt. This could be confusing when incoming messages were held
4927 for policy or load reasons. The message log file is now created at the time
4928 the message is received, and an initial "Received" line is written to it.
4930 4. The automatically generated man page for command line options had a minor
4931 bug that caused no ill effects; however, a more serious problem was that
4932 the procedure for building the man page automatically didn't always
4933 operate. Consequently, release 4.20 contains an out-of-date version. This
4934 shouldn't happen again.
4936 5. When building Exim with embedded Perl support, the script that builds the
4937 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4938 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4940 6. The freeze_tell option was not being used for messages that were frozen on
4941 arrival, either by an ACL or by local_scan().
4943 7. Added the smtp_incomplete_transaction log selector.
4945 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4946 was accepting AUTH without a new EHLO.
4948 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4949 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4952 10. Logging of TCP/IP connections (when configured) now happens in the main
4953 daemon process instead of the child process, so that the TCP/IP connection
4954 count is more accurate (but it can never be perfect).
4956 11. The use of "drop" in a nested ACL was not being handled correctly in the
4957 outer ACL. Now, if condition failure induced by the nested "drop" causes
4958 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4959 or "require"), the connection is dropped.
4961 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4962 that yield "discard" can now be used with an "accept" or a "discard" verb,
4963 but an error is generated for any others (because I can't see a useful way
4964 to define what should happen).
4966 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4967 are now processed in the same way as lines in the Exim configuration file.
4968 In particular, continuation lines are supported.
4970 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4972 15. Added -ti meaning -t -i.
4974 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4975 domains, and warn by logging if others are found.
4977 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4978 returned from GIFCONF when Exim is trying to find the list of interfaces on
4979 a host. The code in os.c has been modified to copy each ifreq to an aligned
4980 structure in all cases.
4982 Also, in some cases, the returned ifreq's were being copied to a 'struct
4983 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4984 means the last couple of bytes of an IPv6 address could be chopped if the
4985 ifreq contained only a normal sockaddr (14 bytes storage).
4987 18. Named domain lists were not supported in the hosts_treat_as_local option.
4988 An entry such as +xxxx was not recognized, and was treated as a literal
4991 19. Ensure that header lines added by a DATA ACL are included in the reject log
4992 if the ACL subsequently rejects the message.
4994 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4995 MD5 (which is deprecated).
4997 21. When testing a filter file using -bf, Exim was writing a message when it
4998 took the sender from a "From " line in the message, but it was not doing so
4999 when it took $return_path from a Return-Path: header line. It now does.
5001 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5002 with a valid header line field name (a series of printing characters
5003 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5005 23. Changed "disc" in the source to "disk" to conform to the documentation and
5006 the book and for uniformity.
5008 24. Ignore Sendmail's -Ooption=value command line item.
5010 25. When execve() failed while trying to run a command in a pipe transport,
5011 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5012 could be confused with a return value of 69 from the command itself. This
5013 has been changed to 127, the value the shell returns if it is asked to run
5014 a non-existent command. The wording for the related log line suggests a
5015 non-existent command as the problem.
5017 26. If received_header_text expands to an empty string, do not add a Received:
5018 header line to the message. (Well, it adds a token one on the spool, but
5019 marks it "old" so that it doesn't get used or transmitted.)
5021 27. Installed eximstats 1.28 (addition of -nt option).
5023 28. There was no check for failure on the call to getsockname() in the daemon
5024 code. This can fail if there is a shortage of resources on the system, with
5025 ENOMEM, for example. A temporary error is now given on failure.
5027 29. Contrary to the C standard, it seems that in some environments, the
5028 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5029 program. Exim now does this explicitly; it affects the formatting of
5030 timestamps using strftime().
5032 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5033 complaints. I've now initialized all the variables, to avoid this.
5035 32. Header lines added by a system filter were not being "seen" during
5036 transport-time rewrites.
5038 33. The info_callback() function passed to OpenSSL is set up with type void
5039 (*)(SSL *, int, int), as described somewhere. However, when calling the
5040 function (actually a macro) that sets it up, the type void(*)() is
5041 expected. I've put in a cast to prevent warnings from picky compilers.
5043 34. If a DNS black list lookup found a CNAME record, but there were no A
5044 records associated with the domain it pointed at, Exim crashed.
5046 35. If a DNS black list lookup returned more than one A record, Exim ignored
5047 all but the first. It now scans all returned addresses if a particular IP
5048 value is being sought. In this situation, the contents of the
5049 $dnslist_value variable are a list of all the addresses, separated by a
5052 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5053 to accept a host name and all its aliases if the forward lookup for any of
5054 them yielded the IP address of the incoming connection. Now it accepts only
5055 those names whose forward lookup yields the correct IP address. Any other
5056 names are discarded. This closes a loophole whereby a rogue DNS
5057 administrator could create reverse DNS records to break through a
5058 wildcarded host restriction in an ACL.
5060 37. If a user filter or a system filter that ran in a subprocess used any of
5061 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5062 the wrong values were passed to the pipe command ($thisaddress had the
5063 value of $0, $0 had the value of $1, etc). This bug was introduced by
5064 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5066 38. Improved the line breaking for long SMTP error messages from ACLs.
5067 Previously, if there was no break point between 40 and 75 characters, Exim
5068 left the rest of the message alone. Two changes have been made: (a) I've
5069 reduced the minimum length to 35 characters; (b) if it can't find a break
5070 point between 35 and 75 characters, it looks ahead and uses the first one
5071 that it finds. This may give the occasional overlong line, but at least the
5072 remaining text gets split now.
5074 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5075 file descriptors might be low, and that setting 1000 would always raise it.
5076 It turns out that in some environments, the limit is already over 1000 and
5077 that lowering it causes trouble. So now Exim takes care not to decrease it.
5079 40. When delivering a message, the value of $return_path is set to $sender_
5080 address at the start of routing (routers may change the value). By an
5081 oversight, this default was not being set up when an address was tested by
5082 -bt or -bv, which affected the outcome if any router or filter referred to
5085 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5086 log only when "message" or "log_message" are set. However, if one of the
5087 conditions was an address verification, or a call to a nested ACL, the
5088 messages generated by the underlying test were being passed through. This
5089 no longer happens. The underlying message is available in $acl_verify_
5090 message for both "message" and "log_message" expansions, so it can be
5091 passed through if needed.
5093 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5094 new expansion $bh_ to give the encoded byte string without charset
5095 translation. Translation happens only if iconv() is available; HAVE_ICONV
5096 indicates this at build time. HEADERS_CHARSET gives the charset to
5097 translate to; headers_charset can change it in the configuration, and
5098 "headers charset" can change it in an individual filter file.
5100 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5101 that creates RFC 2047 encoded "words" labels them as that charset instead
5102 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5103 expansion operator; (ii) when Exim creates a From: line for a local
5104 message; (iii) when a header line is rewritten to include a "phrase" part.
5106 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5107 buggy, causing it to skip the first lines of messages whose message ID
5108 ended in 'D'. This would not have bitten before Exim release 4.14, because
5109 message IDs were unlikely to end in 'D' before then. The effect was to have
5110 incorrect size information for certain domains.
5112 45. #include "config.h" was missing at the start of the crypt16.c module. This
5113 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5116 46. If there was a timeout during a "random" callout check, Exim treated it as
5117 a failure of the random address, and carried on sending RSET and the real
5118 address. If the delay was just some slowness somewhere, the response to the
5119 original RCPT would be taken as a response to RSET and so on, causing
5120 mayhem of various kinds.
5122 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5123 when I implemented it. It didn't allow for the fact that some option values
5124 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5125 the right test for positive values.
5127 48. Domain names in DNS records are case-independent. Exim always looks them up
5128 in lower case. Some resolvers return domain names in exactly the case they
5129 appear in the zone file, that is, they may contain uppercase letters. Not
5130 all resolvers do this - some return always lower case. Exim was treating a
5131 change of case by a resolver as a change of domain, similar to a widening
5132 of a domain abbreviation. This triggered its re-routing code and so it was
5133 trying to route what was effectively the same domain again. This normally
5134 caused routing to fail (because the router wouldn't handle the domain
5135 twice). Now Exim checks for this case specially, and just changes the
5136 casing of the domain that it ultimately uses when it transmits the message
5139 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5142 50. If a filter generated a file delivery with a non-absolute name (possible if
5143 no home directory exists for the router), the forbid_file option was not
5146 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5147 the existing equality matching.
5149 52. Exim was using ints instead of ino_t variables in some places where it was
5150 dealing with inode numbers.
5152 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5153 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5154 called TMPDIR, and if it finds it is different, it changes its value.
5156 54. The smtp_printf() function is now made available to local_scan() so
5157 additional output lines can be written before returning. There is also an
5158 smtp_fflush() function to enable the detection of a dropped connection.
5159 The variables smtp_input and smtp_batched_input are exported to
5162 55. Changed the default runtime configuration: the message "Unknown user"
5163 has been removed from the ACL, and instead placed on the localuser router,
5164 using the cannot_route_message feature. This means that any verification
5165 failures that generate their own messages won't get overridden. Similarly,
5166 the "Unrouteable address" message that was in the ACL for unverifiable
5167 relay addresses has also been removed.
5169 56. Added hosts_avoid_esmtp to the smtp transport.
5171 57. The exicyclog script was not checking for the esoteric option
5172 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5173 will work only if exicyclog is run under the appropriate euid.
5175 58. Following a discussion on the list, the rules by which Exim recognises line
5176 endings on incoming messages have been changed. The -dropcr and drop_cr
5177 options are now no-ops, retained only for backwards compatibility. The
5178 following line terminators are recognized: LF CRLF CR. However, special
5179 processing applies to CR:
5181 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5182 nor a local message in the state where . is a terminator.
5184 (ii) If a bare CR is encountered in a header line, an extra space is added
5185 after the line terminator so as not to end the header. The reasoning
5186 behind this is that bare CRs in header lines are most likely either
5187 to be mistakes, or people trying to play silly games.
5189 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5190 was being incorrectly given as 18 bytes larger than it should have been.
5191 This is a VOB (very old bug).
5193 60. This may never have affected anything current, but just in case it has:
5194 When the local host is found other than at the start of a list of hosts,
5195 the local host, those with the same MX, and any that follow, are discarded.
5196 When the list in question was part of a longer list of hosts, the following
5197 hosts (not currently being processed) were also being discarded. This no
5198 longer happens. I'm not sure if this situation could ever has previously
5201 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5204 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5205 header. This is something that is recommended in a new Internet Draft, and
5206 is something that is documented as being done by Sendmail. There are two
5207 possible values. For messages generated by the autoreply transport, Exim
5210 Auto-Submitted: auto-replied
5212 whereas for all other generated messages (e.g. bounces) it adds
5214 Auto-Submitted: auto-generated
5216 63. The "personal" condition in filters now includes a test for the
5217 Auto-Submitted: header. If it contains the string "auto-" the message it
5218 not considered personal.
5220 64. Added rcpt_include_affixes as a generic transport option.
5222 65. Added queue_only_override (default true).
5224 66. Added the syslog_duplication option.
5226 67. If what should have been the first header line of a message consisted of
5227 a space followed by a colon, Exim was mis-interpreting it as a header line.
5228 It isn't of course - it is syntactically invalid and should therefore be
5229 treated as the start of the message body. The misbehaviour could have
5230 caused a number of strange effects, including loss of data in subsequent
5231 header lines, and spool format errors.
5233 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5234 client host had authenticated. This control can now be exercised by an ACL
5235 for more flexibility.
5237 69. By default, callouts do not happen when testing with -bh. There is now a
5238 variant, -bhc, which does actually run the callout code, including
5239 consulting and updating the callout cache.
5241 70. Added support for saslauthd authentication, courtesy of Alexander
5244 71. If statvfs() failed on the spool or log directories while checking their
5245 size for availability, Exim confusingly gave the error "space shortage".
5246 Furthermore, in debugging mode it crashed with a floating point exception.
5247 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5248 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5249 serious problem, Exim now writes to the main and panic logs when this
5250 happens, with details of the failure. It then refuses to accept the
5251 incoming message, giving the message "spool directory problem" or "log
5252 directory problem" with a 421 code for SMTP messages.
5254 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5255 0, 1, and 2 exist, because some OS complain for execs without them (see
5256 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5257 descriptors. However, the code omitted to check that the open succeeded,
5258 causing mysterious errors if for some reason the permissions on /dev/null
5259 got screwed. Now Exim writes a message to the main and panic logs, and
5260 bombs out if it can't open /dev/null.
5262 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5263 interact so that it is all more flexible. It is supposed to remain
5264 backwards compatible. Also added extra_local_interfaces.
5266 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5267 to bomb out with an assertion failure - to the client this appears as a
5268 connection drop. This problem occurs in the part of the code that was taken
5269 from the Samba project. Fortunately, the assertion is in a very simple
5270 function, so I have fixed this by reproducing the function inline in the
5271 one place where it is called, and arranging for authentication to fail
5272 instead of killing the process with assert().
5274 75. The SPA client code was not working when the server requested OEM rather
5275 than Unicode encoding.
5277 76. Added code to make require_files with a specific uid setting more usable in
5278 the case where statting the file as root fails - usually a non-root-mounted
5279 NFS file system. When this happens and the failure is EACCES, Exim now
5280 forks a subprocess and does the per-uid checking as the relevant uid.
5282 77. Added process_log_path.
5284 78. If log_file_path was not explicitly set, a setting of check_log_space or
5285 check_log_inodes was ignored.
5287 79. If a space check for the spool or log partitions fails, the incident is now
5288 logged. Of course, in the latter case the data may get lost...
5290 80. Added the %p formatting code to string_format() so that it can be used to
5291 print addresses in debug_print(). Adjusted all the address printing in the
5292 debugging in store.c to use %p rather than %d.
5294 81. There was a concern that a line of code in smtp_in.c could overflow a
5295 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5296 initially expressed, the concern was not well-founded, because trailing
5297 spaces are removed early. However, if the trailing spaces were followed by
5298 a NULL, they did not get removed, so the overflow was possible. Two fixes
5301 (a) I re-wrote the offending code in a cleaner fashion.
5302 (b) If an incoming SMTP command contains a NULL character, it is rejected
5305 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5306 runs initgroups(), so that if the Exim user is in any additional groups,
5307 they will be used during message reception.
5313 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob