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 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
11 PP/02 Harmonised TLS library version reporting; only show if debugging.
12 Layout now matches that introduced for other libraries in 4.74 PP/03.
14 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
16 PP/04 New "dns_use_edns0" global option.
18 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
21 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
22 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
24 TK/01 Updated PolarSSL code to 0.14.2.
25 Bugzilla 1097. Patch from Andreas Metzler.
27 PP/07 Catch divide-by-zero in ${eval:...}.
30 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
33 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
34 format-string attack -- SECURITY: remote arbitrary code execution.
36 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
37 time unintentionally subject to list matching rules, letting the header
38 cause arbitrary Exim lookups (of items which can occur in lists, *not*
39 arbitrary string expansion). This allowed for information disclosure.
41 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
42 INT_MIN/-1 -- value coerced to INT_MAX.
48 NM/01 Workround for PCRE version dependency in version reporting
51 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
52 This fixes portability to compilers other than gcc, notably
53 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
55 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
56 makefiles for portability to HP-UX and POSIX correctness.
58 PP/01 Permit LOOKUP_foo enabling on the make command-line.
59 Also via indented variable definition in the Makefile.
60 (Debugging by Oliver Heesakkers).
62 PP/02 Restore caching of spamd results with expanded spamd_address.
63 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
65 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
66 Improves build reliability. Fix from: Frank Elsner
68 NM/02 Fix wide character breakage in the rfc2047 coding
69 Fixes bug 1064. Patch from Andrey N. Oktyabrski
71 NM/03 Allow underscore in dnslist lookups
72 Fixes bug 1026. Patch from Graeme Fowler
74 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
75 Code patches from Adam Ciarcinski of NetBSD.
77 NM/04 Fixed exiqgrep to cope with mailq missing size issue
80 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
81 is logged, to avoid truncation. Patch from John Horne.
83 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
84 Patch from Jakob Hirsch.
86 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
87 SQL string expansion failure details.
88 Patch from Andrey Oktyabrski.
90 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
91 Patch from Simon Arlott.
93 PP/09 New lookups functionality failed to compile on old gcc which rejects
94 extern declarations in function scope.
95 Patch from Oliver Fleischmann
97 PP/10 Use sig_atomic_t for flags set from signal handlers.
98 Check getgroups() return and improve debugging.
99 Fixed developed for diagnosis in bug 927 (which turned out to be
102 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
103 Patch from Mark Zealey.
105 PP/12 Bugzilla 1056: Improved spamd server selection.
106 Patch from Mark Zealey.
108 PP/13 Bugzilla 1086: Deal with maildir quota file races.
109 Based on patch from Heiko Schlittermann.
111 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
112 Patch from Uwe Doering, sign-off by Michael Haardt.
114 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
115 variable declaration deep within a block. Bug and patch from
118 PP/15 lookups-Makefile IRIX compatibilty coercion.
120 PP/16 Make DISABLE_DKIM build knob functional.
122 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
123 Patch by Simon Arlott
125 TF/03 Fix valgrind.h portability to C89 compilers that do not support
126 variable argument macros. Our copy now differs from upstream.
132 TF/01 Failure to get a lock on a hints database can have serious
133 consequences so log it to the panic log.
135 TF/02 Log LMTP confirmation messages in the same way as SMTP,
136 controlled using the smtp_confirmation log selector.
138 TF/03 Include the error message when we fail to unlink a spool file.
140 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
141 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
142 for maintaining out-of-tree patches for some time.
144 PP/01 Bugzilla 139: Documentation and portability issues.
145 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
146 Handle per-OS dynamic-module compilation flags.
148 PP/02 Let /dev/null have normal permissions.
149 The 4.73 fixes were a little too stringent and complained about the
150 permissions on /dev/null. Exempt it from some checks.
151 Reported by Andreas M. Kirchwitz.
153 PP/03 Report version information for many libraries, including
154 Exim version information for dynamically loaded libraries. Created
155 version.h, now support a version extension string for distributors
156 who patch heavily. Dynamic module ABI change.
158 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
159 privilege escalation vulnerability whereby the Exim run-time user
160 can cause root to append content of the attacker's choosing to
163 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
166 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
167 If dropping privileges for untrusted macros, we disabled normal logging
168 on the basis that it would fail; for the Exim run-time user, this is not
169 the case, and it resulted in successful deliveries going unlogged.
170 Fixed. Reported by Andreas Metzler.
176 PP/01 Date: & Message-Id: revert to normally being appended to a message,
177 only prepend for the Resent-* case. Fixes regression introduced in
178 Exim 4.70 by NM/22 for Bugzilla 607.
180 PP/02 Include check_rfc2047_length in configure.default because we're seeing
181 increasing numbers of administrators be bitten by this.
183 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
185 PP/03 Bugzilla 994: added openssl_options main configuration option.
187 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
189 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
191 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
193 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
194 without a peer certificate, leading to a segfault because of an
195 assumption that peers always have certificates. Be a little more
196 paranoid. Problem reported by Martin Tscholak.
198 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
199 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
200 NB: ClamAV planning to remove STREAM in "middle of 2010".
201 CL also introduces -bmalware, various -d+acl logging additions and
202 more caution in buffer sizes.
204 PP/09 Implemented reverse_ip expansion operator.
206 PP/10 Bugzilla 937: provide a "debug" ACL control.
208 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
210 PP/12 Bugzilla 973: Implement --version.
212 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
214 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
216 PP/15 Bugzilla 816: support multiple condition rules on Routers.
218 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
219 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
220 ignore trailing whitespace.
222 JJ/02 prevent non-panic DKIM error from being sent to paniclog
224 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
227 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
228 Notification from Dr Andrew Aitchison.
230 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
231 ExtendedDetectionInfo response format.
232 Notification from John Horne.
234 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
237 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
238 XSL and documented dependency on system catalogs, with examples of how
241 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
244 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
245 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
246 configuration file which is writeable by the Exim user or group.
248 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
249 of configuration files to cover files specified with the -C option if
250 they are going to be used with root privileges, not just the default
253 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
254 option (effectively making it always true).
256 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
257 files to be used while preserving root privileges.
259 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
260 that rogue child processes cannot use them.
262 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
263 run-time user, instead of root.
265 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
266 Exim run-time user without dropping privileges.
268 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
269 result string, instead of calling string_vformat() twice with the same
272 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
273 for other users. Others should always drop root privileges if they use
274 -C on the command line, even for a whitelisted configure file.
276 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
278 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
284 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
285 $data_path, and $header_path variables; fixed documentation bugs and
288 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
289 exipick to access non-standard spools, including the "frozen" queue
292 NM/01 Bugzilla 965: Support mysql stored procedures.
293 Patch from Alain Williams
295 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
297 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
298 Patch from Andreas Metzler
300 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
301 Patch from Kirill Miazine
303 NM/05 Bugzilla 671: Added umask to procmail example.
305 JJ/03 installed exipick 20100323.0, fixing doc bug
307 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
308 directory. Notification and patch from Dan Rosenberg.
310 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
312 TK/02 Improve log output when DKIM signing operation fails.
314 MH/01 Treat the transport option dkim_domain as a colon separated
315 list, not as a single string, and sign the message with each element,
316 omitting multiple occurences of the same signer.
318 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
319 Bugzilla 985, 986. Patch by Simon Arlott
321 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
322 Patch by Simon Arlott
324 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
325 MBX locking. Notification from Dan Rosenberg.
331 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
333 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
335 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
337 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
339 NM/04 Bugzilla 915: Flag broken perl installation during build.
345 TK/01 Added patch by Johannes Berg that expands the main option
346 "spamd_address" if it starts with a dollar sign.
348 TK/02 Write list of recipients to X-Envelope-Sender header when building
349 the mbox-format spool file for content scanning (suggested by Jakob
352 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
353 (http://www.dcc-servers.net/) support via dccifd. Activated by
354 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
356 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
357 by Mark Daniel Reidel <mr@df.eu>.
359 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
360 When building exim an external PCRE library is now needed -
361 PCRE is a system library on the majority of modern systems.
362 See entry on PCRE_LIBS in EDITME file.
364 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
365 conversation. Added nologin parameter to request.
366 Patch contributed by Kirill Miazine.
368 TF/01 Do not log submission mode rewrites if they do not change the address.
370 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
372 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
373 log files in place. Contributed by Roberto Lima.
375 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
377 TF/03 Bugzilla 615: When checking the local_parts router precondition
378 after a local_part_suffix or local_part_prefix option, Exim now
379 does not use the address's named list lookup cache, since this
380 contains cached lookups for the whole local part.
382 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
383 Robert Millan. Documentation is in experimental-spec.txt.
385 TF/04 Bugzilla 668: Fix parallel build (make -j).
387 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
389 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
390 Patch provided by Jan Srzednicki.
392 TF/05 Leading white space used to be stripped from $spam_report which
393 wrecked the formatting. Now it is preserved.
395 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
396 that they are available at delivery time.
398 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
400 TF/08 TLS error reporting now respects the incoming_interface and
401 incoming_port log selectors.
403 TF/09 Produce a more useful error message if an SMTP transport's hosts
404 setting expands to an empty string.
406 NM/06 Bugzilla 744: EXPN did not work under TLS.
407 Patch provided by Phil Pennock.
409 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
410 Patch provided by Richard Godbee.
412 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
413 acl_smtp_notquit, added index entry.
415 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
416 Patch provided by Eugene Bujak.
418 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
419 accept(). Patch provided by Maxim Dounin.
421 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
422 Patch provided by Phil Pennock.
424 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
426 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
427 Patch provided by Brad "anomie" Jorsch.
429 NM/14 Bugzilla 622: Added timeout setting to transport filter.
430 Patch provided by Dean Brooks.
432 TK/05 Add native DKIM support (does not depend on external libraries).
434 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
435 Patch provided by Graeme Fowler.
437 NM/16 Bugzilla 851: Documentation example syntax fix.
439 NM/17 Changed NOTICE file to remove references to embedded PCRE.
441 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
444 NM/19 Bugzilla 745: TLS version reporting.
445 Patch provided by Phil Pennock.
447 NM/20 Bugzilla 167: bool: condition support.
448 Patch provided by Phil Pennock.
450 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
451 clients. Patch provided by Phil Pennock.
453 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
454 Patch provided by Brad "anomie" Jorsch.
456 NM/23 Bugzilla 687: Fix misparses in eximstats.
457 Patch provided by Heiko Schlittermann.
459 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
460 Patch provided by Heiko Schlittermann.
462 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
463 plus update to original patch.
465 NM/26 Bugzilla 799: Documentation correction for ratelimit.
467 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
468 Patch provided by David Brownlee.
470 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
472 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
474 NM/30 Bugzilla 888: TLS documentation bugfixes.
476 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
478 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
479 Unlike the original bugzilla I have changed all shell scripts in src tree.
481 NM/33 Bugzilla 898: Transport filter timeout fix.
482 Patch by Todd Rinaldo.
484 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
485 Patch by Serge Demonchaux.
487 NM/35 Bugzilla 39: Base64 decode bug fixes.
488 Patch by Jakob Hirsch.
490 NM/36 Bugzilla 909: Correct connect() call in dcc code.
492 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
494 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
496 NM/39 Bugzilla 911: Fixed MakeLinks build script.
502 TK/01 Add preliminary DKIM support. Currently requires a forked version of
503 ALT-N's libdkim that I have put here:
504 http://duncanthrax.net/exim-experimental/
506 Note to Michael Haardt: I had to rename some vars in sieve.c. They
507 were called 'true' and it seems that C99 defines that as a reserved
508 keyword to be used with 'bool' variable types. That means you could
509 not include C99-style headers which use bools without triggering
510 build errors in sieve.c.
512 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
513 as mailq or other aliases. Changed the --help handling significantly
514 to do whats expected. exim_usage() emits usage/help information.
516 SC/01 Added the -bylocaldomain option to eximstats.
518 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
520 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
522 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
528 PH/01 Another patch from the Sieve maintainer.
530 PH/02 When an IPv6 address is converted to a string for single-key lookup
531 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
532 dots are used instead of colons so that keys in lsearch files need not
533 contain colons. This was done some time before quoting was made available
534 in lsearch files. However, iplsearch files do require colons in IPv6 keys
535 (notated using the quote facility) so as to distinguish them from IPv4
536 keys. This meant that lookups for IP addresses in host lists did not work
537 for iplsearch lookups.
539 This has been fixed by arranging for IPv6 addresses to be expressed with
540 colons if the lookup type is iplsearch. This is not incompatible, because
541 previously such lookups could never work.
543 The situation is now rather anomolous, since one *can* have colons in
544 ordinary lsearch keys. However, making the change in all cases is
545 incompatible and would probably break a number of configurations.
547 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
550 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
551 conversion specification without a maximum field width, thereby enabling
552 a rogue spamd server to cause a buffer overflow. While nobody in their
553 right mind would setup Exim to query an untrusted spamd server, an
554 attacker that gains access to a server running spamd could potentially
555 exploit this vulnerability to run arbitrary code as the Exim user.
557 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
558 $primary_hostname instead of what libspf2 thinks the hosts name is.
560 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
561 a directory entry by the name of the lookup key. Previously, if a
562 symlink pointed to a non-existing file or a file in a directory that
563 Exim lacked permissions to read, a lookup for a key matching that
564 symlink would fail. Now it is enough that a matching directory entry
565 exists, symlink or not. (Bugzilla 503.)
567 PH/03 The body_linecount and body_zerocount variables are now exported in the
570 PH/04 Added the $dnslist_matched variable.
572 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
573 This means they are set thereafter only if the connection becomes
576 PH/06 Added the client_condition to authenticators so that some can be skipped
577 by clients under certain conditions.
579 PH/07 The error message for a badly-placed control=no_multiline_responses left
580 "_responses" off the end of the name.
582 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
584 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
585 (without spaces) instead of just copying the configuration text.
587 PH/10 Added the /noupdate option to the ratelimit ACL condition.
589 PH/11 Added $max_received_linelength.
591 PH/12 Added +ignore_defer and +include_defer to host lists.
593 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
594 way in which PCRE > 7.0 is built.
596 PH/14 Implemented queue_only_load_latch.
598 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
599 MAIL command. The effect was to mangle the value on 64-bit systems.
601 PH/16 Another patch from the Sieve maintainer.
603 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
605 PH/18 If a system quota error occurred while trying to create the file for
606 a maildir delivery, the message "Mailbox is full" was not appended to the
607 bounce if the delivery eventually timed out. Change 4.67/27 below applied
608 only to a quota excession during the actual writing of the file.
610 PH/19 It seems that peer DN values may contain newlines (and other non-printing
611 characters?) which causes problems in log lines. The DN values are now
612 passed through string_printing() before being added to log lines.
614 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
615 and InterBase are left for another time.)
617 PH/21 Added message_body_newlines option.
619 PH/22 Guard against possible overflow in moan_check_errorcopy().
621 PH/23 POSIX allows open() to be a macro; guard against that.
623 PH/24 If the recipient of an error message contained an @ in the local part
624 (suitably quoted, of course), incorrect values were put in $domain and
625 $local_part during the evaluation of errors_copy.
631 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
632 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
635 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
636 issue a MAIL command.
638 PH/02 In an ACL statement such as
640 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
642 if a client was not listed at all, or was listed with a value other than
643 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
644 the condition was not true (as it should be), so access was not denied.
645 The bug was that the ! inversion was incorrectly passed on to the second
646 item. This has been fixed.
648 PH/03 Added additional dnslists conditions == and =& which are different from
649 = and & when the dns lookup returns more than one IP address.
651 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
652 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
654 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
655 FSYNC, which compiles an option called disable_fsync that allows for
656 bypassing fsync(). The documentation is heavily laced with warnings.
658 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
660 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
661 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
662 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
663 including adding "make clean"; (3) Added -fPIC when compiling the test
664 dynamically loaded module, to get rid of a warning.
666 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
667 message fails, move_frozen_messages = true and ignore_bounce_errors_after
668 = 0s. The bug is otherwise harmless.
670 PH/07 There was a bug in the dovecot authenticator such that the value of
671 $auth1 could be overwritten, and so not correctly preserved, after a
672 successful authentication. This usually meant that the value preserved by
673 the server_setid option was incorrect.
675 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
677 PH/09 Installed PCRE release 7.0.
679 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
680 run for batched SMTP input. It is now run at the start of every message
681 in the batch. While fixing this I discovered that the process information
682 (output by running exiwhat) was not always getting set for -bs and -bS
683 input. This is fixed, and it now also says "batched" for BSMTP.
685 PH/11 Added control=no_pipelining.
687 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
688 patch, slightly modified), and move the expansion of helo_data till after
689 the connection is made in the smtp transport (so it can use these
692 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
694 PH/14 Added log_selector = +pid.
696 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
698 PH/16 Add ${if forany and ${if forall.
700 PH/17 Added dsn_from option to vary the From: line in DSNs.
702 PH/18 Flush SMTP output before performing a callout, unless control =
703 no_callout_flush is set.
705 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
706 was true (the default) a successful delivery failed to delete the retry
707 item, thus causing premature timeout of the address. The bug is now
710 PH/20 Added hosts_avoid_pipelining to the smtp transport.
712 PH/21 Long custom messages for fakedefer and fakereject are now split up
713 into multiline reponses in the same way that messages for "deny" and
714 other ACL rejections are.
716 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
717 with slight modification.
719 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
720 draft, changing the syntax and factoring some duplicate code".
722 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
723 for deliveries of the second and subsequent messages over the same SMTP
726 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
727 ${reduce, with only minor "tidies".
729 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
731 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
732 expansion side effects.
734 PH/27 When a message times out after an over-quota error from an Exim-imposed
735 quota, the bounce message says "mailbox is full". This message was not
736 being given when it was a system quota that was exceeded. It now should
739 MH/03 Made $recipients available in local_scan(). local_scan() already has
740 better access to the recipient list through recipients_list[], but
741 $recipients can be useful in postmaster-provided expansion strings.
743 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
744 in the case of a MAIL command with additional options following the
745 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
746 were accidentally chopped off.
748 PH/29 SMTP synchronization checks are implemented when a command is read -
749 there is a check that no more input is waiting when there shouldn't be
750 any. However, for some commands, a delay in an ACL can mean that it is
751 some time before the response is written. In this time, more input might
752 arrive, invalidly. So now there are extra checks after an ACL has run for
753 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
754 pipelining has not been advertised.
756 PH/30 MH's patch to allow iscntrl() characters to be list separators.
758 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
759 returned in the SMTP response when smtp_return_error_details was false.
762 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
763 instead of the C I/O that was originally supplied, because problems were
766 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
767 Exim which did not show up earlier: it was assuming that a call to
768 SSL_CTX_set_info_callback() might give an error value. In fact, there is
769 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
770 was a macro that became an assignment, so it seemed to work. This has
771 changed to a proper function call with a void return, hence the compile
772 error. Exim's code has been fixed.
774 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
777 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
779 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
781 PH/37 If a message is not accepted after it has had an id assigned (e.g.
782 because it turns out to be too big or there is a timeout) there is no
783 "Completed" line in the log. When some messages of this type were
784 selected by exigrep, they were listed as "not completed". Others were
785 picked up by some special patterns. I have improved the selection
786 criteria to be more general.
788 PH/38 The host_find_failed option in the manualroute router can now be set
789 to "ignore", to completely ignore a host whose IP address cannot be
790 found. If all hosts are ignored, the behaviour is controlled by the new
791 host_all_ignored option.
793 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
794 homing or because of multiple MX records with /mx) generated more than
795 one IP address, and the following item turned out to be the local host,
796 all the secondary addresses of the first item were incorrectly removed
797 from the list, along with the local host and any following hosts (which
798 is what is supposed to happen).
800 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
801 whoever called Exim into the -H file. In the case of the daemon it was
802 behaving confusingly. When first started, it used values for whoever
803 started the daemon, but after a SIGHUP it used the Exim user (because it
804 calls itself on a restart). I have changed the code so that it now always
807 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
808 message are rejected with the same error (e.g. no authentication or bad
809 sender address), and a DATA command is nevertheless sent (as can happen
810 with PIPELINING or a stupid MUA), the error message that was given to the
811 RCPT commands is included in the rejection of the DATA command. This is
812 intended to be helpful for MUAs that show only the final error to their
815 PH/42 Another patch from the Sieve maintainer.
817 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
818 Eximstats - Fixed some broken HTML links and added missing column headers
820 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
821 columns for Rejects, Temp Rejects, Ham, and Spam rows.
823 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
825 PH/43 Yet another patch from the Sieve maintainer.
827 PH/44 I found a way to check for a TCP/IP connection going away before sending
828 the response to the final '.' that terminates a message, but only in the
829 case where the client has not sent further data following the '.'
830 (unfortunately, this is allowed). However, in many cases there won't be
831 any further data because there won't be any more messages to send. A call
832 to select() can be used: if it shows that the input is "ready", there is
833 either input waiting, or the socket has been closed. An attempt to read
834 the next input character can distinguish the two cases. Previously, Exim
835 would have sent an OK response which the client would never have see.
836 This could lead to message repetition. This fix should cure that, at
837 least in a lot of common cases.
839 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
840 advertised in response to EHLO.
846 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
847 fixed by 4.65/MH/01 (is this a record?) are fixed:
849 (i) An empty string was always treated as zero by the numeric comparison
850 operators. This behaviour has been restored.
852 (ii) It is documented that the numeric comparison operators always treat
853 their arguments as decimal numbers. This was broken in that numbers
854 starting with 0 were being interpreted as octal.
856 While fixing these problems I realized that there was another issue that
857 hadn't been noticed. Values of message_size_limit (both the global option
858 and the transport option) were treated as octal if they started with 0.
859 The documentation was vague. These values are now always treated as
860 decimal, and I will make that clear in the documentation.
866 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
867 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
870 MH/01 Don't check that the operands of numeric comparison operators are
871 integers when their expansion is in "skipping" mode (fixes bug
872 introduced by 4.64-PH/07).
874 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
875 child addresses, Exim now panics and dies. Previously, because the count
876 is held in a short int, deliveries were likely to be lost. As such a
877 large number of recipients for a single message is ridiculous
878 (performance will be very, very poor), I have chosen to impose a limit
879 rather than extend the field.
885 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
886 leftover -K file (the existence of which was triggered by #402).
887 While we were at it, introduced process PID as part of the -K
888 filename. This should rule out race conditions when creating
891 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
892 processing considerably. Previous code took too long for large mails,
893 triggering a timeout which in turn triggers #401.
895 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
896 in the DK code in transports.c. sendfile() is not really portable,
897 hence the _LINUX specificness.
899 TF/01 In the add_headers option to the mail command in an Exim filter,
900 there was a bug that Exim would claim a syntax error in any
901 header after the first one which had an odd number of characters
904 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
905 callout verification, Exim cached a "reject" for the entire domain. This
906 is correct for most verifications, but it is not correct for a recipient
907 verification with use_sender or use_postmaster set, because in that case
908 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
909 case of MAIL FROM:<> rejection from other early rejections (e.g.
910 rejection of HELO). When verifying a recipient using a non-null MAIL
911 address, the cache is ignored if it shows MAIL FROM:<> rejection.
912 Whatever the result of the callout, the value of the domain cache is
913 left unchanged (for any other kind of callout, getting as far as trying
914 RCPT means that the domain itself is ok).
916 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
919 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
920 manifest itself as EPIPE rather than ECONNECT. When tidying away a
921 session, the daemon ignores ECONNECT errors and logs others; it now
922 ignores EPIPE as well.
924 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
925 (quoted-printable decoding).
927 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
928 later the small subsequent patch to fix an introduced bug.
930 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
932 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
934 PH/08 An error is now given if message_size_limit is specified negative.
936 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
937 to be given (somewhat) arbitrary names.
939 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
942 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
943 miscellaneous code fixes
945 PH/10 Added the log_reject_target ACL modifier to specify where to log
948 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
949 hostname. This is wrong, because it relates to the incoming message (and
950 probably the interface on which it is arriving) and not to the outgoing
951 callout (which could be using a different interface). This has been
952 changed to use the value of the helo_data option from the smtp transport
953 instead - this is what is used when a message is actually being sent. If
954 there is no remote transport (possible with a router that sets up host
955 addresses), $smtp_active_hostname is used.
957 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
958 tweaks were necessary in order to get it to work (see also 21 below):
959 (a) The code assumed that strncpy() returns a negative number on buffer
960 overflow, which isn't the case. Replaced with Exim's string_format()
962 (b) There were several signed/unsigned issues. I just did the minimum
963 hacking in of casts. There is scope for a larger refactoring.
964 (c) The code used strcasecmp() which is not a standard C function.
965 Replaced with Exim's strcmpic() function.
966 (d) The code set only $1; it now sets $auth1 as well.
967 (e) A simple test gave the error "authentication client didn't specify
968 service in request". It would seem that Dovecot has changed its
969 interface. Fortunately there's a specification; I followed it and
970 changed what the client sends and it appears to be working now.
972 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
975 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
976 address A is aliased to B and C, where B exists and C does not. Without
977 -v the output is "A verified" because verification stops after a
978 successful redirection if more than one address is generated. However,
979 with -v the child addresses are also verified. Exim was outputting "A
980 failed to verify" and then showing the successful verification for C,
981 with its parentage. It now outputs "B failed to verify", showing B's
982 parentage before showing the successful verification of C.
984 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
985 look up a TXT record in a specific list after matching in a combined
988 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
989 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
990 they consult the DNS. I had assumed they would set it the way they
991 wanted; and indeed my experiments on Linux seem to show that in some
992 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
993 To be on the safe side, however, I have now made the interface to
994 host_find_byname() similar to host_find_bydns(), with an argument
995 containing the DNS resolver options. The host_find_byname() function now
996 sets these options at its start, just as host_find_bydns() does. The smtp
997 transport options dns_qualify_single and dns_search_parents are passed to
998 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
999 of host_find_byname() use the default settings of RES_DEFNAMES
1000 (qualify_single) but not RES_DNSRCH (search_parents).
1002 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1003 spool_read_header() do less string testing, by means of a preliminary
1004 switch on the second character of optional "-foo" lines. (This is
1005 overdue, caused by the large number of possibilities that now exist.
1006 Originally there were few.) While I was there, I also converted the
1007 str(n)cmp tests so they don't re-test the leading "-" and the first
1008 character, in the hope this might squeeze out yet more improvement.
1010 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1011 flag allowing group syntax was set by the header_syntax check but not
1012 turned off, possible causing trouble later; (2) The flag was not being
1013 set at all for the header_verify test, causing "group"-style headers to
1014 be rejected. I have now set it in this case, and also caused header_
1015 verify to ignore an empty address taken from a group. While doing this, I
1016 came across some other cases where the code for allowing group syntax
1017 while scanning a header line wasn't quite right (mostly, not resetting
1018 the flag correctly in the right place). These bugs could have caused
1019 trouble for malformed header lines. I hope it is now all correct.
1021 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1022 with the "reply" argument non-NULL. The code, however (which originally
1023 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1024 but it didn't always do it. This confused somebody who was copying the
1025 code for some other use. I have removed all the tests.
1027 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1028 feature that was used to support insecure browsers during the U.S. crypto
1029 embargo. It requires special client support, and Exim is probably the
1030 only MTA that supported it -- and would never use it because real RSA is
1031 always available. This code has been removed, because it had the bad
1032 effect of slowing Exim down by computing (never used) parameters for the
1033 RSA_EXPORT functionality.
1035 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1036 authenticator to fail if there's a tab character in the incoming data
1037 (there should never be unless someone is messing about, as it's supposed
1038 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1039 if the connection is using TLS or if the remote IP is the same as the
1040 local IP, and the "valid-client-cert option" if a client certificate has
1043 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1044 authenticators. This can be used for authorization after authentication
1045 succeeds. (In the case of plaintext, it servers for both authentication
1048 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1049 if any retry times were supplied.
1051 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1052 connection, where there is no client IP address to check. In this
1053 situation, the verify now always succeeds.
1055 PH/25 Applied John Jetmore's -Mset patch.
1057 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1059 PH/27 In a string expansion for a processed (not raw) header when multiple
1060 headers of the same name were present, leading whitespace was being
1061 removed from all of them, but trailing whitespace was being removed only
1062 from the last one. Now trailing whitespace is removed from each header
1063 before concatenation. Completely empty headers in a concatenation (as
1064 before) are ignored.
1066 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1067 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1069 PH/29 [Removed. This was a change that I later backed out, and forgot to
1070 correct the ChangeLog entry (that I had efficiently created) before
1071 committing the later change.]
1073 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1074 address errors (4xx response to RCPT) over the same connection as other
1075 messages routed to the same hosts. Such deliveries are always "forced",
1076 so retry times are not inspected. This resulted in far too many retries
1077 for the affected addresses. The effect occurred only when there were more
1078 hosts than the hosts_max_try setting in the smtp transport when it had
1079 the 4xx errors. Those hosts that it had tried were not added to the list
1080 of hosts for which the message was waiting, so if all were tried, there
1081 was no problem. Two fixes have been applied:
1083 (i) If there are any address or message errors in an SMTP delivery, none
1084 of the hosts (tried or untried) are now added to the list of hosts
1085 for which the message is waiting, so the message should not be a
1086 candidate for sending over the same connection that was used for a
1087 successful delivery of some other message. This seems entirely
1088 reasonable: after all the message is NOT "waiting for some host".
1089 This is so "obvious" that I'm not sure why it wasn't done
1090 previously. Hope I haven't missed anything, but it can't do any
1091 harm, as the worst effect is to miss an optimization.
1093 (ii) If, despite (i), such a delivery is accidentally attempted, the
1094 routing retry time is respected, so at least it doesn't keep
1095 hammering the server.
1097 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1098 in ${readsocket because some servers need this prod.
1100 PH/32 Added some extra debug output when updating a wait-xxx database.
1102 PH/33 The hint "could be header name not terminated by colon", which has been
1103 given for certain expansion errors for a long time, was not being given
1104 for the ${if def:h_colon_omitted{... case.
1106 PH/34 The spec says: "With one important exception, whenever a domain list is
1107 being scanned, $domain contains the subject domain." There was at least
1108 one case where this was not true.
1110 PH/35 The error "getsockname() failed: connection reset by peer" was being
1111 written to the panic log as well as the main log, but it isn't really
1112 panic-worthy as it just means the connection died rather early on. I have
1113 removed the panic log writing for the ECONNRESET error when getsockname()
1116 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1117 runs only) independently of the message's sender address. This meant
1118 that, if the 4xx error was in fact related to the sender, a different
1119 message to the same recipient with a different sender could confuse
1120 things. In particualar, this can happen when sending to a greylisting
1121 server, but other circumstances could also provoke similar problems.
1122 I have changed the default so that the retry time for these errors is now
1123 based a combination of the sender and recipient addresses. This change
1124 can be overridden by setting address_retry_include_sender=false in the
1127 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1128 remote server are returned as part of bounce messages. This was not
1129 happening for LMTP over a pipe (the lmtp transport), but now it is the
1130 same for both kinds of LMTP.
1132 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1133 in header lines that were in fact CNAMEs. This is no longer the case.
1135 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1136 and queue runs started by the daemon processed all messages. This has
1137 been fixed so that -R and -S can now usefully be given with -q<time>.
1139 PH/40 Import PCRE release 6.7 (fixes some bugs).
1141 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1143 PH/42 Give an error if -q is specified more than once.
1145 PH/43 Renamed the variables $interface_address and $interface_port as
1146 $received_ip_address and $received_port, to make it clear that these
1147 values apply to message reception, and not to the outgoing interface when
1148 a message is delivered. (The old names remain recognized, of course.)
1150 PH/44 There was no timeout on the connect() call when using a Unix domain
1151 socket in the ${readsocket expansion. There now is.
1153 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1154 be meaningful with "accept".
1156 SC/01 Eximstats V1.43
1157 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1159 SC/02 Eximstats V1.44
1160 Use a glob alias rather than an array ref in the generated
1161 parser. This improves both readability and performance.
1163 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1164 Collect SpamAssassin and rejection statistics.
1165 Don't display local sender or destination tables unless
1166 there is data to show.
1167 Added average volumes into the top table text output.
1169 SC/04 Eximstats V1.46
1170 Collect data on the number of addresses (recipients)
1171 as well as the number of messages.
1173 SC/05 Eximstats V1.47
1174 Added 'Message too big' to the list of mail rejection
1175 reasons (thanks to Marco Gaiarin).
1177 SC/06 Eximstats V1.48
1178 Mainlog lines which have GMT offsets and are too short to
1179 have a flag are now skipped.
1181 SC/07 Eximstats V1.49 (Alain Williams)
1182 Added the -emptyok flag.
1184 SC/08 Eximstats V1.50
1185 Fixes for obtaining the IP address from reject messages.
1187 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1188 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1189 whitesspace changes from 4.64-PH/27
1191 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1194 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1195 are found, allow negative numbers in numeric criteria)
1197 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1199 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1200 to match changes made in 4.64-PH/43
1202 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1204 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1205 "domain" is actually an IP address, and force a failure. This locks out
1206 those revolvers/nameservers that support "A-for-A" lookups, in
1207 contravention of the specifications.
1209 PH/48 When a host name was looked up from an IP address, and the subsequent
1210 forward lookup of the name timed out, the host name was left in
1211 $sender_host_name, contrary to the specification.
1213 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1214 restricted to single-key lookups, Exim was not diagnosing an error if
1215 * or *@ was used with a query-style lookup.
1217 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1219 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1220 long ago, but noone interested enough thought of it. Let's just say that
1221 the "1.1" means that there are some new functions that weren't there at
1222 some point in the past.
1224 PH/51 Error processing for expansion failure of helo_data from an smtp
1225 transport during callout processing was broken.
1227 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1228 tested/used via the -bh/-bhc/-bs options.
1230 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1231 bug, fixed in subsequent PCRE releases).
1233 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1234 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1236 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1242 SC/01 Use a glob alias rather than an array ref in eximstats generated
1243 parser. This improves both readability and performance.
1245 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1246 Don't display local sender or destination tables in eximstats unless
1247 there is data to show.
1248 Added average volumes into the eximstats top table text output.
1250 SC/03 Collect data on the number of addresses (recipients) as well
1251 as the number of messages in eximstats.
1253 TF/01 Correct an error in the documentation for the redirect router. Exim
1254 does (usually) call initgroups() when daemonizing.
1256 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1257 with consistent privilege compared to when running as a daemon.
1259 TF/03 Note in the spec that $authenticated_id is not set for local
1260 submissions from trusted users.
1262 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1263 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1265 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1266 by adding some example configuration directives to the default
1267 configuration file. A little bit of work is required to uncomment the
1268 directives and define how usernames and passwords are checked, but
1269 there is now a framework to start from.
1271 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1272 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1273 without this. I don't know how relevant this is to other LDAP libraries.
1275 PH/02 Add the verb name to the "unknown ACL verb" error.
1277 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1279 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1281 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1282 directory not expanded when it should be if an expanded home directory
1283 was set for the address (which is overridden by the transport).
1285 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1288 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1289 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1290 because it is too late at that time, and has no effect.
1292 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1293 security issue with \' (bugzilla #107). I could not use the
1294 PQescapeStringConn() function, because it needs a PGconn value as one of
1297 PH/08 When testing addresses using -bt, indicate those final addresses that
1298 are duplicates that would not cause an additional delivery. At least one
1299 person was confused, thinking that -bt output corresponded to deliveries.
1300 (Suppressing duplicates isn't a good idea as you lose the information
1301 about possibly different redirections that led to the duplicates.)
1303 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1304 systems where poll() doesn't work, in particular OS X.
1306 PH/10 Added more information to debugging output for retry time not reached.
1308 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1309 operations in malware.c.
1311 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1314 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1315 syslog_duplication set false, log lines that would normally be written
1316 both the the main log and to the reject log were not written to syslog at
1319 PH/14 In the default configuration, change the use of "message" in ACL warn
1320 statements to "add_header".
1322 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1323 not followed by a command (e.g. "seen endif").
1325 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1326 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1329 PH/17 Added extra conditions to the default value of delay_warning_condition
1333 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1334 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1335 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1338 The Auto-Submitted: and various List- headers are standardised, whereas I
1339 don't think Precedence: ever was.
1341 PH/18 Refactored debugging code in route_finduser() to show more information,
1342 in particular, the error code if getpwnam() issues one.
1344 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1345 This is apparently needed in addition to the PH/07 change above to avoid
1346 any possible encoding problems.
1348 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1349 but not after initializing Perl.
1351 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1352 output them only if debugging. By default they are written stderr,
1353 apparently, which is not desirable.
1355 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1358 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1361 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1363 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1364 authenticated or an ident call has been made. Suppress the default
1365 values for $authenticated_id and $authenticated_sender (but permit -oMai
1366 and -oMas) when testing with -bh.
1368 PH/24 Re-jigged the order of the tests in the default configuration so that the
1369 tests for valid domains and recipients precede the DNS black list and CSA
1370 tests, on the grounds that those ones are more expensive.
1372 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1373 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1374 This bug exists in every version of Exim that I still have, right back to
1377 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1378 However, an attempt to turn on case-sensitivity in a regex key by
1379 including (?-i) didn't work because the subject string was already
1380 lowercased, and the effects were non-intuitive. It turns out that a
1381 one-line patch can be used to allow (?-i) to work as expected.
1387 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1388 other effects) broke the use of negated acl sub-conditions.
1390 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1393 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1394 "Deny" causes Exim to reject the incoming connection with a 554 error.
1395 Unfortunately, if there is a major crisis, such as a disk failure,
1396 tcp-wrappers gives "deny", whereas what one would like would be some
1397 kind of temporary error. A kludge has been added to help with this.
1398 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1399 554 error is used if errno is still zero or contains ENOENT (which occurs
1400 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1403 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1405 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1406 errors. Otherwise a message that provokes a temporary error (when other
1407 messages do not) can cause a whole host to time out.
1409 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1410 addresses were routed directly to files or pipes from a redirect router.
1411 File deliveries just didn't batch; pipe deliveries might have suffered
1414 PH/06 A failure to get a lock for a hints database would erroneously always say
1415 "Failed to get write lock", even when it was really a read lock.
1417 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1418 of 0600. This has been changed to use the value of the lockfile_mode
1419 option (which defaults to 0600).
1421 PH/08 Applied small patch from the Sieve maintainer.
1423 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1424 folder from quota calculations, a direct delivery into this folder messed
1425 up the contents of the maildirsize file. This was because the regex was
1426 used only to exclude .Trash (or whatever) when the size of the mailbox
1427 was calculated. There was no check that a delivery was happening into an
1428 excluded directory. This bug has been fixed by ignoring all quota
1429 processing for deliveries into excluded directories.
1431 PH/10 Added the maildirfolder_create_regex option to appendfile.
1437 PH/01 The code for finding all the local interface addresses on a FreeBSD
1438 system running IPv6 was broken. This may well have applied to all BSD
1439 systems, as well as to others that have similar system calls. The broken
1440 code found IPv4 interfaces correctly, but gave incorrect values for the
1441 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1442 that it would not match correctly against @[] and not recognize the IPv6
1445 PH/02 The ipliteral router was not recognizing addresses of the form user@
1446 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1448 PH/03 Added disable_ipv6.
1450 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1451 decoded form, because it is most often used to construct To: headers
1452 lines in autoreplies, and the decoded form may well be syntactically
1453 invalid. However, $reply_address has leading white space removed, and all
1454 newlines turned into spaces so that the autoreply transport does not
1457 PH/05 If group was specified without a user on a router, and no group or user
1458 was specified on a transport, the group from the router was ignored.
1460 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1461 for visible compile-time settings that can be used to change these
1462 numbers, for those that want even more. Backwards compatibility with old
1463 spool files has been maintained. However, going back to a previous Exim
1464 release will lost any variables that are in spool files.
1466 PH/07 Two small changes when running in the test harness: increase delay when
1467 passing a TCP/IP connection to a new process, in case the original
1468 process has to generate a bounce, and remove special handling of
1469 127.0.0.2 (sic), which is no longer necessary.
1471 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1472 be the same on different OS.
1474 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1477 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1478 whether --show-vars was specified or not
1480 JJ/02 exipick: Added support for new ACL variable spool format introduced
1483 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1484 syntactically invalid From: or Reply-to: line, and a filter used this to
1485 generate an autoreply, and therefore failed to obtain an address for the
1486 autoreply, Exim could try to deliver to a non-existent relative file
1487 name, causing unrelated and misleading errors. What now happens is that
1488 it logs this as a hard delivery error, but does not attempt to create a
1491 PH/11 The exinext utility has a -C option for testing purposes, but although
1492 the given file was scanned by exinext itself; it wasn't being passed on
1493 when Exim was called.
1495 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1496 an end-of-file indication when reading a command response.
1498 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1499 compiled. In many other places in Exim, IPv6 addresses are always
1500 recognized, so I have changed this. It also means that IPv4 domain
1501 literals of the form [IPV4:n.n.n.n] are now always recognized.
1503 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1504 used if the router is not running as root, for example, when verifying at
1505 ACL time, or when using -bh. The debugging output from this situation was
1506 non-existent - all you got was a failure to exec. I have made two
1509 (a) Failures to set uid/gid, the current directory, or a process leader
1510 in a subprocess such as that created by queryprogram now generate
1511 suitable debugging ouput when -d is set.
1513 (b) The queryprogram router detects when it is not running as root,
1514 outputs suitable debugging information if -d is set, and then runs
1515 the subprocess without attempting to change uid/gid.
1517 PH/15 Minor change to Makefile for building test_host (undocumented testing
1520 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1521 additional section of a DNS packet that returns MX or SRV records.
1522 Instead, it always explicitly searches for A/AAAA records. This avoids
1523 major problems that occur when a DNS server includes only records of one
1524 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1525 fixed another bug: if SRV records were looked up and the corresponding
1526 address records were *not* found in the additional section, the port
1527 values from the SRV records were lost.
1529 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1530 using the correct key (the original address) when searching the retry
1531 rules in order to find which one to use for generating the retry hint.
1533 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1534 adding the default one. Similarly, if it contains a Reply-To: header, the
1535 errors_reply_to option, if set, is not used.
1537 PH/19 When calculating a retry time, Exim used to measure the "time since
1538 failure" by looking at the "first failed" field in the retry record. Now
1539 it does not use this if it is later than than the arrival time of the
1540 message. Instead it uses the arrival time. This makes for better
1541 behaviour in cases where some deliveries succeed, thus re-setting the
1542 "first failed" field. An example is a quota failure for a huge message
1543 when small messages continue to be delivered. Without this change, the
1544 "time since failure" will always be short, possible causing more frequent
1545 delivery attempts for the huge message than are intended.
1546 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1548 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1549 $1, $2, $3) because the numerical variables can be reset during some
1550 expansion items (e.g. "match"), thereby losing the authentication data.
1552 PH/21 Make -bV show the size of off_t variables so that the test suite can
1553 decide whether to run tests for quotas > 2G.
1555 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1556 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1557 filecount value is greater than 2G or if a quota value is greater than 2G
1558 on a system where the size of off_t is not greater than 4, a panic error
1561 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1562 never match. The debug and -bh output now contains an explicit error
1563 message indicating a malformed IPv4 address or mask.
1565 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1566 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1567 PH/23 above applies.
1569 PH/25 Do not write to syslog when running in the test harness. The only
1570 occasion when this arises is a failure to open the main or panic logs
1571 (for which there is an explicit test).
1573 PH/26 Added the /no_tell option to "control=freeze".
1575 PH/27 If a host name lookup failed very early in a connection, for example, if
1576 the IP address matched host_lookup and the reverse lookup yielded a name
1577 that did not have a forward lookup, an error message of the form "no IP
1578 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1579 could be logged. Now it outputs the IP address instead of "NULL".
1581 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1582 allows the sender and the authenticated sender to be set when
1583 submitting a message from within Exim. Since child_open_exim() is
1584 documented for local_scan(), the new function should be too.
1586 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1587 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1588 results in an empty string is now treated as unset.
1590 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1592 PH/31 Added sender_verify_fail logging option.
1594 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1595 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1596 tidied the source and removed it altogether.
1598 PH/33 When a queue run was abandoned because the load average was too high, a
1599 log line was always written; now it is written only if the queue_run log
1600 selector is set. In addition, the log line for abandonment now contains
1601 information about the queue run such as the pid. This is always present
1602 in "start" and "stop" lines but was omitted from the "abandon" line.
1604 PH/34 Omit spaces between a header name and the colon in the error message that
1605 is given when verify = headers_syntax fails (if there are lots of them,
1606 the message gets confusing).
1608 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1609 names, as there are now some PTR records that contain slashes. This check
1610 is only to protect against broken name servers that fall over on strange
1611 characters, so the fact that it applies to all lookups doesn't matter.
1613 PH/36 Now that the new test suite is complete, we can remove some of the
1614 special code in Exim that was needed for the old test suite. For example,
1615 sorting DNS records because real resolvers return them in an arbitrary
1616 order. The new test suite's fake resolver always returns records in the
1619 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1620 bounces) except when queue_only is set, to avoid logging races between
1621 the different processes.
1623 PH/38 Panic-die if .include specifies a non-absolute path.
1625 PH/39 A tweak to the "H" retry rule from its user.
1627 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1628 a label. They prevented compilation on older perls.
1630 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1631 a warning to be raised on newish perls.
1633 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1634 on queue. Changes to match documented behaviour of showing count of
1635 messages matching specified criteria.
1637 PH/40 Changed the default ident timeout from 30s to 5s.
1639 PH/41 Added support for the use of login_cap features, on those BSD systems
1640 that have them, for controlling the resources used by pipe deliveries.
1642 PH/42 The content-scanning code uses fopen() to create files in which to put
1643 message data. Previously it was not paying any attention to the mode of
1644 the files. Exim runs with umask(0) because the rest of the code creates
1645 files with open(), and sets the required mode explicitly. Thus, these
1646 files were ending up world-writeable. This was not a big issue, because,
1647 being within the spool directory, they were not world-accessible. I have
1648 created a function called modefopen, which takes an additional mode
1649 argument. It sets umask(777), creates the file, chmods it to the required
1650 mode, then resets the umask. All the relevant calls to fopen() in the
1651 content scanning code have been changed to use this function.
1653 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1654 to 24 hours. This avoids potential overflow problems when processing G
1655 and H retry rules. I suspect nobody ever tinkers with this value.
1657 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1659 PH/45 When the plaintext authenticator is running as a client, the server's
1660 challenges are checked to ensure they are valid base64 strings. By
1661 default, the authentication attempt is cancelled if an invalid string is
1662 received. Setting client_ignore_invalid_base64 true ignores these errors.
1663 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1664 they are received. Thus, the responses can be made to depend on the
1665 challenges. If an invalid string is ignored, an empty string is placed in
1668 PH/46 Messages that are created by the autoreply transport now contains a
1669 References: header, in accordance with RFCs 2822 and 3834.
1671 PH/47 Added authenticated_sender_force to the smtp transport.
1673 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1675 PH/49 Installed latest patch from the Sieve maintainer.
1677 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1678 was also set, the appendfile transport was unnecessarily scanning a
1679 directory of message files (e.g. for maildir delivery) to find the count
1680 of files (along with the size), even though it did not need this
1681 information. It now does the scan only if it needs to find either the
1682 size of the count of files.
1684 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1686 PH/52 Two bugs concerned with error handling when the smtp transport is
1689 (i) Exim was not creating retry information for temporary errors given
1690 for individual recipients after the DATA command when the smtp transport
1691 was used in LMTP mode. This meant that they could be retried too
1692 frequently, and not timed out correctly.
1694 (ii) Exim was setting the flag that allows error details to be returned
1695 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1696 recipients that were returned after the DATA command.
1698 PH/53 This is related to PH/52, but is more general: for any failing address,
1699 when detailed error information was permitted to be returned to the
1700 sender, but the error was temporary, then after the final timeout, only
1701 "retry timeout exceeded" was returned. Now it returns the full error as
1702 well as "retry timeout exceeded".
1704 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1705 do this, and (what is worse) MTAs that accept it.
1707 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1708 will now be deprecated.
1710 PH/56 New os.c-cygwin from the Cygwin maintainer.
1712 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1713 formats (previously only available in exim formats via -bpr, -bpru,
1714 and -bpra. Now also available in native and exiqgrep formats)
1716 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1717 with very large, slow to parse queues
1719 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1721 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1723 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1724 responses to authentication challenges, though it was showing the
1725 challenges; (ii) I've removed the CR characters from the debug output for
1728 PH/58 Allow for the insertion of a newline as well as a space when a string
1729 is turned into more than one encoded-word during RFC 2047 encoding. The
1730 Sieve code now uses this.
1732 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1733 data_4xx, lost_connection, tls_required.
1735 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1736 message was being sent as an SMTP response.
1738 PH/61 Add -l and -k options to exicyclog.
1740 PH/62 When verifying, if an address was redirected to one new address, so that
1741 verification continued, and the new address failed or deferred after
1742 having set something in $address_data, the value of $address_data was not
1743 passed back to the ACL. This was different to the case when no
1744 redirection occurred. The value is now passed back in both cases.
1746 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1747 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1748 use login_cap.h, so on its own it isn't the distinguishing feature. The
1749 new name refers directly to the setclassresources() function.
1751 PH/65 Added configuration files for NetBSD3.
1753 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1755 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1756 is preferred over IPv4.
1758 PH/68 The bounce_return_message and bounce_return_body options were not being
1759 honoured for bounces generated during the reception of non-SMTP messages.
1760 In particular, this applied to messages rejected by the ACL. This bug has
1761 been fixed. However, if bounce_return_message is true and bounce_return_
1762 body is false, the headers that are returned for a non-SMTP message
1763 include only those that have been read before the error was detected.
1764 (In the case of an ACL rejection, they have all been read.)
1766 PH/69 The HTML version of the specification is now built in a directory called
1767 spec_html instead of spec.html, because the latter looks like a path with
1768 a MIME-type, and this confuses some software.
1770 PH/70 Catch two compiler warnings in sieve.c.
1772 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1773 function verify_get_ident() calls ip_connect() to connect a socket, but
1774 if the "connect()" function timed out, ip_connect() used to close the
1775 socket. However, verify_get_ident() also closes the socket later, and in
1776 between Exim writes to the log, which may get opened at this point. When
1777 the socket was closed in ip_connect(), the log could get the same file
1778 descriptor number as the socket. This naturally causes chaos. The fix is
1779 not to close the socket in ip_connect(); the socket should be closed by
1780 the function that creates it. There was only one place in the code where
1781 this was missing, in the iplookup router, which I don't think anybody now
1782 uses, but I've fixed it anyway.
1784 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1785 well as to direct DNS lookups. Otherwise the handling of names in host
1786 lists is inconsistent and therefore confusing.
1792 PH/01 Two changes to the default runtime configuration:
1794 (1) Move the checks for relay_from_hosts and authenticated clients from
1795 after to before the (commented out) DNS black list checks.
1797 (2) Add control=submission to the relay_from_hosts and authenticated
1798 clients checks, on the grounds that messages accepted by these
1799 statements are most likely to be submissions.
1801 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1803 (1) Generate an error if the third argument for the ${prvs expansion is
1806 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1809 (3) Reset the variables that are obtained from the first argument of
1810 ${prvscheck and used in the second argument before leaving the code,
1811 because their memory is reclaimed, so using them afterwards may do
1814 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1815 one (it's much easier than Tom thought :-).
1817 (5) Because of (4), we can now allow for the use of $prvscheck_result
1818 inside the third argument.
1820 PH/03 For some reason, the default setting of PATH when running a command from
1821 a pipe transport was just "/usr/bin". I have changed it to
1824 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1825 anything to be listed in the output from -bV.
1827 PH/05 When a filter generated an autoreply, the entire To: header line was
1828 quoted in the delivery log line, like this:
1830 => >A.N.Other <ano@some.domain> <original@ddress> ...
1832 This has been changed so that it extracts the operative address. There
1833 may be more than one such address. If so, they are comma-separated, like
1836 => >ano@some.domain,ona@other.domain <original@ddress> ...
1838 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1839 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1840 not being looked up in the rDNS to get a host name, Exim was showing the
1841 IP address twice in Received: lines, even though the IP addresses were
1842 identical. For example:
1844 Received: from [1.2.3.4] (helo=[1.2.3.4])
1846 However, if the real host name was known, it was omitting the HELO data
1847 if it matched the actual IP address. This has been tidied up so that it
1848 doesn't show the same IP address twice.
1850 PH/07 When both +timestamp and +memory debugging was on, the value given by
1851 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1852 called by the debug printing, thereby overwriting the timestamp buffer.
1853 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1855 PH/08 When the original message was included in an autoreply transport, it
1856 always said "this is a copy of the message, including all the headers",
1857 even if body_only or headers_only was set. It now gives an appropriate
1860 PH/09 Applied a patch from the Sieve maintainer which:
1862 o fixes some comments
1863 o adds the (disabled) notify extension core
1864 o adds some debug output for the result of if/elsif tests
1865 o points to the current vacation draft in the documentation
1866 and documents the missing references header update
1870 o fixes a bug in processing the envelope test (when testing
1871 multiple envelope elements, the last element determinted the
1874 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1875 Electronic Mail") by including:
1877 Auto-submitted: auto-generated
1879 in the messages that it generates (bounce messages and others, such as
1880 warnings). In the case of bounce messages for non-SMTP mesages, there was
1881 also a typo: it was using "Auto_submitted" (underscore instead of
1882 hyphen). Since every message generated by Exim is necessarily in response
1883 to another message, thes have all been changed to:
1885 Auto-Submitted: auto-replied
1887 in accordance with these statements in the RFC:
1889 The auto-replied keyword:
1891 - SHOULD be used on messages sent in direct response to another
1892 message by an automatic process,
1894 - MUST NOT be used on manually-generated messages,
1896 - MAY be used on Delivery Status Notifications (DSNs) and Message
1897 Disposition Notifications (MDNs),
1899 - MUST NOT be used on messages generated by automatic or periodic
1900 processes, except for messages which are automatic responses to
1903 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1904 to the default Received: header definition.
1906 PH/12 Added log selector acl_warn_skipped (default on).
1908 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1909 variables because (a) they are in the wrong storage pool and (b) even if
1910 they were copied, it wouldn't work properly because of the caching.
1912 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1913 checking when decoding. Apparently there are clients that generate
1914 overlong encoded strings. Why am I not surprised?
1916 PH/15 If the first argument of "${if match_address" was not empty, but did not
1917 contain an "@" character, Exim crashed. Now it writes a panic log message
1918 and treats the condition as false.
1920 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1922 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1923 extension "envelope-auth". The code is finished and in agreement with
1924 other implementations, but there is no documentation so far and in fact,
1925 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1926 not changing the active code.
1928 Print executed "if" and "elsif" statements when debugging is used. This
1929 helps a great deal to understand what a filter does.
1931 Document more things not specified clearly in RFC3028. I had all this
1932 sorted out, when out of a sudden new issues came to my mind. Oops."
1934 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1937 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1938 canonical form (as documented). However, after a host name lookup from
1939 the IP address, check_host() was doing a simple string comparison with
1940 addresses acquired from the DNS when checking that the found name did
1941 have the original IP as one of its addresses. Since any found IPv6
1942 addresses are likely to be in abbreviated form, the comparison could
1943 fail. Luckily, there already exists a function for doing the comparison
1944 by converting both addresses to binary, so now that is used instead of
1945 the text comparison.
1947 PH/20 There was another similar case to PH/19, when a complete host name was
1948 given in a host list; looking up its IP address could give an abbreviated
1949 form, whereas the current host's name might or might not be abbreviated.
1950 The same fix has been applied.
1956 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1957 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1960 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1961 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1963 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1965 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1966 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1967 Also, for "Auto-Submitted", treat anything other than "no" as
1968 non-personal, in accordance with RFC 3834. (Previously it treated
1969 anything starting "auto-" as non-personal.)
1971 TF/01 The control=submission/name=... option had a problem with syntax
1972 errors if the name included a slash character. The /name= option
1973 now slurps the rest of the string, so it can include any characters
1974 but it must come last in the list of options (after /sender_retain
1977 PH/05 Some modifications to the interface to the fake nameserver for the new
1985 TK/01 Added the "success_on_redirect" address verification option. See
1986 NewStuff for rationale and an example.
1988 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1990 PH/02 Patch to exigrep to allow it to work on syslog lines.
1992 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1993 fread() to skip over the body file's header line, because in Cygwin the
1994 header line is locked and is inaccessible.
1996 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1997 co-exist for some time) to make it clear that it is the Exim ID that is
1998 referenced, not the Message-ID: header line.
2000 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2001 string_format() function, because snprintf() does not exist on all
2004 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2005 ${stat: expansion item.
2007 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2008 protocol synchronization error", to keep the pedants happy.
2010 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2011 well as for IRIX systems, when gcc is being used. See the host.c source
2014 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2016 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2019 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2020 transport and to the smtp transport in LMTP mode.
2022 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2024 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2025 run a filter in a subprocess. This could lead to confusion in subsequent
2026 lookups in the parent process. There should also be a search_tidyup() at
2027 the end of the subprocess.
2029 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2030 only if the host matched helo_try_verify_hosts, which caused the
2031 verification to occur when the EHLO/HELO command was issued. The ACL just
2032 tested the remembered result. Now, if a previous verification attempt has
2033 not happened, "verify = helo" does it there and then.
2035 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2037 TK/03 Fix log output including CR from clamd.
2039 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2040 exist provoked a memory error which could cause a segfault.
2042 PH/15 Installed PCRE 6.2
2044 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2046 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2047 of the problem. Specifically, suggested +O2 rather than +O1 for the
2050 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2052 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2053 record, the port was not correctly logged when the outgoing_port log
2054 selector was set (it logged the transort's default port).
2056 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2057 fallback_hosts, and "hosts" in the smtp transport.
2059 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2060 host errors such as "Connection refused".
2062 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2063 authentication with radiusclient 0.4.9:
2065 - Error returned from rc_read_config was caught wrongly
2066 - Username/password not passed on to radius server due to wrong length.
2068 The presumption is that some radiusclient API changes for 4.51/PH/17
2069 were not taken care of correctly. The code is still untested by me (my
2070 Linux distribution still has 0.3.2 of radiusclient), but it was
2071 contributed by a Radius user.
2073 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2074 expanding the "port" option of the smtp transport.
2076 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2077 while reading a MIME header. Thanks to Tom Hughes for a patch.
2079 PH/24 Include config.h inside local_scan.h so that configuration settings are
2082 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2083 that in an ACL for RCPT (for example), you can examine exactly what was
2086 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2087 commands, but it was not correctly comparing the address with the actual
2088 client host address. Thus, it would show the EHLO address in Received:
2089 header lines when this was not necessary.
2091 PH/27 Added the % operator to ${eval:}.
2093 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2094 it should be ignoring failures (because with -C, for example, it has lost
2095 privilege). It wasn't ignoring creation failures other than "already
2098 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2101 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2102 because an input line was too long, either on its own, or by virtue of
2103 too many continuations, the temporary file was not being removed, and the
2104 return code was incorrect.
2106 PH/31 Missing "BOOL" in function definition in filtertest.c.
2108 PH/32 Applied Sieve patches from the maintainer.
2110 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2112 PH/33 Added "verify = not_blind".
2114 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2115 Local/Makefile (with some defaults set). These are used in built scripts
2116 such as exicyclog, but they have never been used in the exim_install
2117 script (though there are many overriding facilities there). I have
2118 arranged that the exim_install script now takes note of these two
2121 PH/35 Installed configuration files for Dragonfly.
2123 PH/36 When a locally submitted message by a trusted user did not contain a
2124 From: header, and the sender address was obtained from -f or from an SMTP
2125 MAIL command, and the trusted user did not use -F to supply a sender
2126 name, $originator_name was incorrectly used when constructing a From:
2127 header. Furthermore, $originator_name was used for submission mode
2128 messages from external hosts without From: headers in a similar way,
2129 which is clearly wrong.
2131 PH/37 Added control=suppress_local_fixups.
2133 PH/38 When log_selector = +received_sender was set, and the addition of the
2134 sender made the log line's construction buffer exactly full, or one byte
2135 less than full, an overflow happened when the terminating "\n" was
2138 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2139 when the result of a list match is failure because a DNS lookup failed.
2141 PH/40 RM_COMMAND is now used in the building process.
2143 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2144 the "build-* directories that it finds.
2146 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2147 address in a domain literal was a prefix of an interface address.
2149 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2150 when verifying a sender address, unless rewrite_headers is false.
2152 PH/44 Wrote a long comment about why errors_to addresses are verified as
2153 recipients, not senders.
2155 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2156 the ratelimit ACL was added.
2158 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2160 PH/46 Added extra information about PostgreSQL errors to the error string.
2162 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2163 avoiding the need to install special zones in a real server. This is
2164 backwards compatible; if it can't find the fake resolver, it drops back.
2165 Thus, both old and new test suites can be run.
2167 TF/02 Added util/ratelimit.pl
2169 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2170 clock is set back in time.
2172 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2173 Candler <B.Candler@pobox.com>.
2175 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2176 OK for addresses that are the result of redirections.
2178 PH/48 A number of further additions for the benefit of the new test suite,
2179 including a fake gethostbyname() that interfaces to the fake DNS resolver
2182 TF/06 The fix for widen_domains has also been applied to qualify_single and
2183 search_parents which are the other dnslookup options that can cause
2186 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2189 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2191 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2192 service" if it encountered an MX record with an empty target hostname.
2193 The message is now "an MX or SRV record indicated no SMTP service".
2195 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2196 if the DNS of the sending site is misconfigured. This is quite a
2197 common situation. This change restores the behaviour of treating a
2198 helo verification defer as a failure.
2200 PH/51 If self=fail was set on a router, the bounce message did not include the
2201 actual error message.
2207 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2209 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2210 command itself ended in error, the underlying message about the transport
2211 filter timeout was being overwritten with the pipe command error. Now the
2212 underlying error message should be appended to the second error message.
2214 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2216 PH/02 Reduce the amount of output that "make" produces by default. Full output
2217 can still be requested.
2219 PH/03 The warning log line about a condition test deferring for a "warn" verb
2220 was being output only once per connection, rather than after each
2221 occurrence (because it was using the same function as for successful
2222 "warn" verbs). This seems wrong, so I have changed it.
2224 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2225 it should not have, which might have caused a crash in the right
2226 circumstances, but probably never did.
2228 PH/04 Installed a modified version of Tony Finch's patch to make submission
2229 mode fix the return path as well as the Sender: header line, and to
2230 add a /name= option so that you can make the user's friendly name appear
2233 TF/03 Added the control = fakedefer ACL modifier.
2235 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2236 Mark Lowes for thorough testing.
2238 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2240 TK/03 Merged latest SRS patch from Miles Wilton.
2242 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2243 with the definition in sysexits.h (which is #included earlier).
2244 Fortunately, Exim does not actually use EX_OK. The code used to try to
2245 preserve the sysexits.h value, by assumimg that macro definitions were
2246 scanned for macro replacements. I have been disabused of this notion,
2247 so now the code just undefines EX_OK before #including unistd.h.
2249 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2250 in the smtp transport. When a block could not be written in a single
2251 write() function, the timeout was being re-applied to each part-write.
2252 This seems wrong - if the receiver was accepting one byte at a time it
2253 would take for ever. The timeout is now adjusted when this happens. It
2254 doesn't have to be particularly precise.
2256 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2257 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2259 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2260 without a domain if the check to <postmaster@domain> fails.
2262 SC/01 Eximstats: added -xls and the ability to specify output files
2263 (patch written by Frank Heydlauf).
2265 SC/02 Eximstats: use FileHandles for outputing results.
2267 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2269 SC/04 Eximstats: fixed display of large numbers with -nvr option
2271 SC/05 Eximstats: fixed merging of reports with empty tables.
2273 SC/06 Eximstats: added the -include_original_destination flag
2275 SC/07 Eximstats: removed tabs and trailing whitespace.
2277 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2279 TK/06 MBOX spool code: Add real "From " MBOX separator line
2280 so the .eml file is really in mbox format (even though
2281 most programs do not really care). Patch from Alex Miller.
2283 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2284 The latter is generated from $received_to and is only set if the
2285 message has one envelope recipient. SA can use these headers,
2286 obviously out-of-the-box. Patch from Alex Miller.
2288 PH/08 The ${def test on a variable was returning false if the variable's
2289 value was "0", contrary to what the specification has always said!
2290 The result should be true unless the variable is empty.
2292 PH/09 The syntax error of a character other than { following "${if
2293 def:variable_name" (after optional whitespace) was not being diagnosed.
2294 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2295 accidental colon was present, for example, could give incorrect results.
2297 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2298 result is used (not including appendfile, where other changes are about
2301 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2302 This involved changing a lot of size variables from int to off_t. It
2303 should work with maildirs and everything.
2305 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2306 spamd dying while we are connected to it.
2308 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2311 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2312 function for BDB 4.3.
2314 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2316 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2317 the sources for PCRE have been re-arranged and more files are now
2320 PH/15 The code I had for printing potentially long long variables in PH/11
2321 above was not the best (it lost precision). The length of off_t variables
2322 is now inspected at build time, and an appropriate printing format (%ld
2323 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2324 to be "long long int" or "long int". This is needed for the internal
2325 formatting function string_vformat().
2327 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2328 the configuration file to be ":syslog", then the script "guesses" where
2329 the logs files are, rather than using the compiled in default. In our
2330 case the guess is not the same as the compiled default, so the script
2331 suddenly stopped working when I started to use syslog. The patch checks
2332 to see if log_file_path is "". If so, it attempts to read it from exim
2333 with no configuration file to get the compiled in version, before it
2334 falls back to the previous guessing code."
2336 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2337 implementing BATV in an Exim configuration. See NewStuff for the gory
2340 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2341 Makefile that are specific to HP-UX.
2343 PH/18 If the "use_postmaster" option was set for a recipient callout together
2344 with the "random" option, the postmaster address was used as the MAIL
2345 FROM address for the random test, but not for the subsequent recipient
2346 test. It is now used for both.
2348 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2349 patch removes a few documentation additions to RFC 3028, because the
2350 latest draft now contains them. It adds the new en;ascii-case comparator
2351 and a new error check for 8bit text in MIME parts. Comparator and
2352 require names are now matched exactly. I enabled the subaddress
2353 extension, but it is not well tested yet (read: it works for me)."
2355 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2356 rework some of the code of TK/09 above to avoid the hardwired use of
2357 "%lld" and "long long". Replaced the call to snprintf() with a call to
2360 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2361 records point to non-existent hosts", "retry timeout exceeded", and
2362 "retry time not reached for any host after a long failure period".
2364 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2365 experimental DomainKeys support:
2367 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2368 (2) On an error such as an illegally used "control", the wrong name for
2369 the control was given.
2371 These problems did NOT occur unless DomainKeys support was compiled.
2373 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2375 PH/24 Added ${if match_ip condition.
2377 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2378 databases so that it will be absolutely obvious if a crash occurs in the
2379 DB library. This is a regular occurrence (often caused by mis-matched
2382 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2383 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2384 picked up on a user's system that detects such things. There doesn't seem
2385 to be a gcc warning option for this - only an attribute that has to be
2386 put on the function's prototype. It seems that in Fedora Core 4 they have
2387 set this on a number of new functions. No doubt there will be more in due
2390 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2391 specify a transport. However, if an address that was verified by such a
2392 router was the subject of a callout, Exim crashed because it tried to
2393 read the rcpt_include_affixes from the non-existent transport. Now it
2394 just assumes that the setting of that option is false. This bug was
2395 introduced by 4.51/PH/31.
2397 PH/28 Changed -d+all to exclude +memory, because that information is very
2398 rarely of interest, but it makes the output a lot bigger. People tend to
2399 do -d+all out of habit.
2401 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2402 code in os-type was giving problems when libc.so lives in lib64, like on
2405 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2406 aren't the modern standard, and it seems that some systems' include files
2407 don't always have them. Exim was already checking for some of the newer
2408 ones like T_AAAA, and defining it itself. I've added checks for all the
2409 record types that Exim uses.
2411 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2412 not automatically generating a new one, as it is supposed to. This
2413 prevented TLS from working. If the file did exist, but contained invalid
2414 data, a new version was generated, as expected. It was only the case of a
2415 non-existent file that was broken.
2417 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2418 with a change in libdomainkeys > 0.64.
2420 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2421 from DNS. If the selector record carries the flag, it now has
2422 precedence over the domain-wide flag.
2424 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2426 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2427 the use of a port name that isn't defined in /etc/services) occurred, the
2428 message was deferred as in a normal delivery, and thus remained on the
2429 spool, instead of being failed because of the mua_wrapper setting. This
2430 is now fixed, and I tidied up some of the mua_wrapper messages at the
2433 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2434 the messages in a hash of arrays rather than using individual hashes.
2435 This is a bit cleaner and results in dramatic memory savings, albeit
2436 at a slight CPU cost.
2438 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2439 as requested by Marc Sherman.
2441 SC/10 Eximstats: added histograms for user specified patterns as requested
2444 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2446 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2447 fopen() in the content-scanning modules that did not already have it.
2453 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2454 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2456 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2458 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2460 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2461 to test Sieve filters that use "vacation".
2463 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2464 that changes the way the GnuTLS parameters are stored in the cache file.
2465 The new format can be generated externally. For backward compatibility,
2466 if the data in the cache doesn't make sense, Exim assumes it has read an
2467 old-format file, and it generates new data and writes a new file. This
2468 means that you can't go back to an older release without removing the
2471 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2472 work if there are any delivery delays because "one_time" forces the
2473 parent to be marked "delivered", so its unseen clone is never tried
2474 again. For this reason, Exim now forbids the simultaneous setting of
2477 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2478 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2479 entry if you want to know the details. The fix, however, neglected to
2480 consider the case when local delivery batching is involved. The test for
2481 "previously delivered" was not happening when checking to see if an
2482 address could be batched with a previous (undelivered) one; under
2483 certain circumstances this could lead to multiple deliveries to the same
2486 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2487 in its include files, and this causes problems building Exim.
2489 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2490 header_syntax) but Exim was just ignoring anything given after a slash.
2491 In particular, this caused confusion with an attempt to use "verify =
2492 reverse_host_lookup/defer_ok". An error is now given when options are
2493 supplied for verify items that do not have them. (Maybe reverse_host_
2494 lookup should have a defer_ok option, but that's a different point.)
2496 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2497 defined by RFC 821) to 2048, because there were problems with some AUTH
2498 commands, and RFC 1869 says the size should be increased for extended
2499 SMTP commands that take arguments.
2501 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2504 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2505 "unknown" error; now it says that the functionality isn't in the binary.
2507 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2508 an address' error message when a string expansion fails (syntax or
2509 whatever). Otherwise the password may appear in the log. Following change
2510 PH/42 below, there is no longer a chance of it appearing in a bounce
2513 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2515 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2516 of its addresses was ever tried. (Bugzilla bug #2.)
2518 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2519 the result incorrectly in the debug output. (It correctly added a newline
2520 to what was transported.)
2522 TF/01 Added $received_time.
2524 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2525 commented out examples of how to interface to a virus scanner and to
2526 SpamAssassin. Also added commented examples of av_scanner and
2527 spamd_address settings.
2529 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2530 and controls are allowed in which ACLs. There were a couple of minor
2531 errors. Some of the entries in the conditions table (which is a table of
2532 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2533 negation of where the condition IS allowed.
2535 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2537 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2538 header file does not have a version number, so I've had to invent a new
2539 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2540 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2541 radiusclient), but it was contributed by a Radius user.
2543 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2544 files or directories, for OpenSSL.
2546 PH/19 When an Exim process that is running as root has to create an Exim log
2547 file, it does so in a subprocess that runs as exim:exim so as to get the
2548 ownership right at creation (otherwise, other Exim processes might see
2549 the file with the wrong ownership). There was no test for failure of this
2550 fork() call, which would lead to the process getting stuck as it waited
2551 for a non-existent subprocess. Forks do occasionally fail when resources
2552 run out. I reviewed all the other calls to fork(); they all seem to check
2555 PH/20 When checking for unexpected SMTP input at connect time (before writing
2556 the banner), Exim was not dealing correctly with a non-positive return
2557 from the read() function. If the client had disconnected by this time,
2558 the result was a log entry for a synchronization error with an empty
2559 string after "input=" when read() returned zero. If read() returned -1
2560 (an event I could not check), uninitialized data bytes were printed.
2561 There were reports of junk text (parts of files, etc) appearing after
2564 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2566 PH/22 Added support for macro redefinition, and (re)definition in between
2567 driver and ACL definitions.
2569 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2570 forgetting to use the resulting value; it was using the unexpanded value.
2572 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2573 hadn't been configured. The fix is from Juergen Kreileder, who
2574 understands it better than I do:
2576 "Here's what I see happening with three configured cyrus_sasl
2577 authenticators configured (plain, login, cram-md5):
2579 On startup auth_cyrus_sasl_init() gets called for each of these.
2580 This means three calls to sasl_listmech() without a specified mech_list.
2581 => SASL tests which mechs of all available mechs actually work
2582 => three warnings about OTP not working
2583 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2585 With the patch, sasl_listmech() also gets called three times. But now
2586 SASL's mech_list option is set to the server_mech specified in the the
2587 authenticator. Or in other words, the answer from sasl_listmech()
2588 gets limited to just the mech you're testing for (which is different
2590 => the return list contains just 'plain' or 'login', 'cram-md5' or
2591 nothing depending on the value of ob->server_mech.
2593 I've just tested the patch: Authentication still works fine,
2594 unavailable mechs specified in the exim configuration are still
2595 caught, and the auth.log warnings about OTP are gone."
2597 PH/25 When debugging is enabled, the contents of the command line are added
2598 to the debugging output, even when log_selector=+arguments is not
2601 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2602 answer is "GNU", and only if the return is "GNU/something" is the answer
2605 PH/27 $acl_verify_message is now set immediately after the failure of a
2606 verification in an ACL, and so is available in subsequent modifiers. In
2607 particular, the message can be preserved by coding like this:
2609 warn !verify = sender
2610 set acl_m0 = $acl_verify_message
2612 Previously, $acl_verify_message was set only while expanding "message"
2613 and "log_message" when a very denied access.
2615 PH/28 Modified OS/os.c-Linux with
2617 -#ifndef OS_LOAD_AVERAGE
2618 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2620 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2621 nomenclature these days.)
2623 PH/29 Installed patch from the Sieve maintainer that adds the options
2624 sieve_useraddress and sieve_subaddress to the redirect router.
2626 PH/30 In these circumstances:
2627 . Two addresses routed to the same list of hosts;
2628 . First host does not offer TLS;
2629 . First host accepts first address;
2630 . First host gives temporary error to second address;
2631 . Second host offers TLS and a TLS session is established;
2632 . Second host accepts second address.
2633 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2634 and peerdn, if requested) that were in fact used only for the second
2637 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2638 attention to any local part prefix or suffix that was matched by the
2639 router that accepted the address. It now behaves in the same way as it
2640 does for delivery: the affixes are removed from the local part unless
2641 rcpt_include_affixes is set on the transport.
2643 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2644 timeout during the DATA phase of an incoming message.
2646 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2647 applied a patch sanctioned by the Sieve maintainer.
2649 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2650 the uid or gid is negative. A case of a negative gid caused this to be
2651 noticed. The fix allows for either to be negative.
2653 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2654 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2657 PH/36 The change PH/12 above was broken. Fixed it.
2659 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2660 the grounds that routing the same address twice would always produce the
2661 same answer. This might have been true once, but it is certainly no
2662 longer true now. Routing a child address may depend on the previous
2663 routing that produced that child. Some complicated redirection strategies
2664 went wrong when messages had multiple recipients, and made Exim's
2665 behaviour dependent on the order in which the addresses were given.
2667 I have moved the duplicate checking until after the routing is complete.
2668 Exim scans the addresses that are assigned to local and remote
2669 transports, and removes any duplicates. This means that more work will be
2670 done, as duplicates will always all be routed, but duplicates are
2671 presumably rare, so I don't expect this is of any significance.
2673 For deliveries to pipes, files, and autoreplies, the duplicate checking
2674 still happens during the routing process, since they are not going to be
2677 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2678 It corrects a timeout issue with spamd. This is Ian's comment: "The
2679 background is that sometimes spamd either never reads data from a
2680 connection it has accepted, or it never writes response data. The exiscan
2681 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2682 blindly assumes that writes won't block so it may never time out."
2684 PH/39 Allow G after quota size as well as K and M.
2686 PH/40 The value set for $authenticated_id in an authenticator may not contain
2687 binary zeroes or newlines because the value is written to log lines and
2688 to spool files. There was no check on this. Now the value is run through
2689 the string_printing() function so that such characters are converted to
2690 printable escape sequences.
2692 PH/41 $message_linecount is a new variable that contains the total number of
2693 lines in the message. Compare $body_linecount, which is the count for the
2696 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2697 bounce and delay warning messages, except in certain special cases, which
2700 (a) An SMTP error message from a remote host;
2701 (b) A message specified in a :fail: redirection;
2702 (c) A message specified in a "fail" command in a system filter;
2703 (d) A message specified in a FAIL return from the queryprogram router;
2704 (e) A message specified by the cannot_route_message router option.
2706 In these cases only, Exim does include the error details in bounce and
2707 warning messages. There are also a few cases where bland messages such
2708 as "unrouteable address" or "local delivery error" are given.
2710 PH/43 $value is now also set for the "else" part of a ${run expansion.
2712 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2713 being worked on, but at least Exim now implements the latest version to
2716 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2717 process to complete was treated as a delivery failure, a timeout while
2718 writing the message to the pipe was logged, but erroneously treated as a
2719 successful delivery. Such timeouts include transport filter timeouts. For
2720 consistency with the overall process timeout, these timeouts are now
2721 treated as errors, giving rise to delivery failures by default. However,
2722 there is now a new Boolean option for the pipe transport called
2723 timeout_defer, which, if set TRUE, converts the failures into defers for
2724 both kinds of timeout. A transport filter timeout is now identified in
2727 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2728 systems where "make" and "gmake" are different, calling "gmake" at top
2729 level broke things. I've arranged for the value of $(MAKE) to be passed
2730 from the Makefile to this script so that it can call the same version of
2734 A note about Exim versions 4.44 and 4.50
2735 ----------------------------------------
2737 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2738 changes of various kinds. As a consequence, a big documentation update was
2739 needed. This delayed the release for rather longer than seemed good, especially
2740 in the light of a couple of (minor) security issues. Therefore, the changes
2741 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2742 release. So 4.44 and 4.50 are in effect two different branches that both start
2745 I have left the 4.50 change log unchanged; it contains all the changes since
2746 4.43. The change log for 4.44 is below; many of its items are identical to
2747 those for 4.50. This seems to be the most sensible way to preserve the
2748 historical information.
2754 1. Minor wording change to the doc/README.SIEVE file.
2756 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2757 computation of the current number of files was incorrect.
2759 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2760 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2763 4. Give more explanation in the error message when the command for a transport
2764 filter fails to execute.
2766 5. There are several places where Exim runs a non-Exim command in a
2767 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2768 was being done only for the command run by the queryprogram router. It is
2769 now done for all such subprocesses. The other cases are: ${run, transport
2770 filters, and the commands run by the lmtp and pipe transports.
2772 6. Added CONFIGURE_GROUP build-time option.
2774 7. Some older OS have a limit of 256 on the maximum number of file
2775 descriptors. Exim was using setrlimit() to set 1000 as a large value
2776 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2777 systems. I've change it so that if it can't get 1000, it tries for 256.
2779 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2780 was an oversight, and furthermore, ever since the addition of extra
2781 controls (e.g. 4.43/32), the checks on when to allow different forms of
2782 "control" were broken. There should now be diagnostics for all cases when a
2783 control that does not make sense is encountered.
2785 9. Added the /retain_sender option to "control=submission".
2787 10. $recipients is now available in the predata ACL (oversight).
2789 11. Tidy the search cache before the fork to do a delivery from a message
2790 received from the command line. Otherwise the child will trigger a lookup
2791 failure and thereby defer the delivery if it tries to use (for example) a
2792 cached ldap connection that the parent has called unbind on.
2794 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2795 of $address_data from the recipient verification was clobbered by the
2796 sender verification.
2798 13. The value of address_data from a sender verification is now available in
2799 $sender_address_data in subsequent conditions in the ACL statement.
2801 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2803 15. Added a new option "connect=<time>" to callout options, to set a different
2806 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2807 was its contents. (It was OK if the option was not defined at all.)
2809 17. A "Completed" log line is now written for messages that are removed from
2810 the spool by the -Mrm option.
2812 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2813 information about exactly what failed.
2815 19. Added -dd to debug only the daemon process.
2817 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2818 handles timeouts, both on the server side and network timeouts. Renamed the
2819 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2821 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2822 It is now set to "smtps".
2824 22. $host_address is now set to the target address during the checking of
2825 ignore_target_hosts.
2827 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2828 being passed; this would have caused $sender_host_name to have been used if
2829 matching the list had actually called for a host name (not very likely,
2830 since this list is usually IP addresses). A host name is now passed as
2833 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2834 code that specifies a non-restarting handler (typically sigaction() in
2835 modern systems) in an attempt to fix a rare and obscure crash bug.
2837 25. Narrowed the window for a race in the daemon that could cause it to ignore
2838 SIGCHLD signals. This is not a major problem, because they are used only to
2839 wake it up if nothing else does.
2841 26. A malformed maildirsize file could cause Exim to calculate negative values
2842 for the mailbox size or file count. Odd effects could occur as a result.
2843 The maildirsize information is now recalculated if the size or filecount
2846 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2847 support for a long time. Removed HAVE_SYS_VFS_H.
2849 28. Installed the latest version of exipick from John Jetmore.
2851 29. In an address list, if the pattern was not a regular expression, an empty
2852 subject address (from a bounce message) matched only if the pattern was an
2853 empty string. Non-empty patterns were not even tested. This was the wrong
2854 because it is perfectly reasonable to use an empty address as part of a
2855 database query. An empty address is now tested by patterns that are
2856 lookups. However, all the other forms of pattern expect the subject to
2857 contain a local part and a domain, and therefore, for them, an empty
2858 address still always fails if the pattern is not itself empty.
2860 30. Exim went into a mad DNS loop when attempting to do a callout where the
2861 host was specified on an smtp transport, and looking it up yielded more
2862 than one IP address.
2864 31. Re-factored the code for checking spool and log partition space into a
2865 function that finds that data and another that does the check. The former
2866 is then used to implement four new variables: $spool_space, $log_space,
2867 $spool_inodes, and $log_inodes.
2869 32. The RFC2047 encoding function was originally intended for short strings
2870 such as real names; it was not keeping to the 75-character limit for
2871 encoded words that the RFC imposes. It now respects the limit, and
2872 generates multiple encoded words if necessary. To be on the safe side, I
2873 have increased the buffer size for the ${rfc2047: expansion operator from
2876 33. It is now permitted to omit both strings after an "if" condition; if the
2877 condition is true, the result is "true". As before, when the second string
2878 is omitted, a false condition yields an empty string. This makes it less
2879 cumbersome to write custom ACL and router conditions.
2881 34. Failure to deliver a bounce message always caused it to be frozen, even if
2882 there was an errors_to setting on the router. The errors_to setting is now
2885 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2886 canonical form (fully expanded) before being placed in
2887 $sender_host_address.
2889 36. The table in the code that translates DNS record types into text (T_A to
2890 "A" for instance) was missing entries for NS and CNAME. It is just possible
2891 that this could have caused confusion if both these types were looked up
2892 for the same domain, because the text type is used as part of Exim's
2893 per-process caching. But the chance of anyone hitting this buglet seems
2896 37. The dnsdb lookup has been extended in a number of ways.
2898 (1) There is a new type, "zns", which walks up the domain tree until it
2899 finds some nameserver records. It should be used with care.
2901 (2) There is a new type, "mxh", which is like "mx" except that it returns
2902 just the host names, not the priorities.
2904 (3) It is now possible to give a list of domains (or IP addresses) to be
2905 looked up. The behaviour when one of the lookups defers can be
2906 controlled by a keyword.
2908 (4) It is now possible to specify the separator character for use when
2909 multiple records are returned.
2911 38. The dnslists ACL condition has been extended: it is now possible to supply
2912 a list of IP addresses and/or domains to be looked up in a particular DNS
2915 39. Added log_selector=+queue_time_overall.
2917 40. When running the queue in the test harness, wait just a tad after forking a
2918 delivery process, to get repeatability of debugging output.
2920 41. Include certificate and key file names in error message when GnuTLS fails
2921 to set them up, because the GnuTLS error message doesn't include the name
2922 of the failing file when there is a problem reading it.
2924 42. Allow both -bf and -bF in the same test run.
2926 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2928 44. The "Exiscan patch" is now merged into the mainline Exim source.
2930 45. Sometimes the final signoff response after QUIT could fail to get
2931 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2932 < 0 before doing a fflush(). This bug looks as though it goes back to the
2933 introduction of TLS in release 3.20, but "sometimes" must have been rare
2934 because the tests only now provoked it.
2936 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2937 (this can affect the format of dates).
2939 47. exim_tidydb, when checking for the continued existence of a message for
2940 which it has found a message-specific retry record, was not finding
2941 messages that were in split spool directories. Consequently, it was
2942 deleting retry records that should have stayed in existence.
2944 48. Steve fixed some bugs in eximstats.
2946 49. The SPA authentication driver was not abandoning authentication and moving
2947 on to the next authenticator when an expansion was forced to fail,
2948 contradicting the general specification for all authenticators. Instead it
2949 was generating a temporary error. It now behaves as specified.
2951 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2952 (the order specifies the preference for clients). The order is now AES256,
2953 AES128, 3DES, ARCFOUR128.
2955 51. Small patch to Sieve code - explicitly set From: when generating an
2958 52. Exim crashed if a remote delivery caused a very long error message to be
2959 recorded - for instance if somebody sent an entire SpamAssassin report back
2960 as a large number of 550 error lines. This bug was coincidentally fixed by
2961 increasing the size of one of Exim's internal buffers (big_buffer) that
2962 happened as part of the Exiscan merge. However, to be on the safe side, I
2963 have made the code more robust (and fixed the comments that describe what
2966 53. Now that there can be additional text after "Completed" in log lines (if
2967 the queue_time_overall log selector is set), a one-byte patch to exigrep
2968 was needed to allow it to recognize "Completed" as not the last thing in
2971 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2972 patch that reportedly fixes this has been added. I am not expert enough to
2973 create a test for it. This is what the patch creator wrote:
2975 "I found a little strange behaviour of ldap code when working with
2976 Windows 2003 AD Domain, where users was placed in more than one
2977 Organization Units. When I tried to give exim partial DN, the exit code
2978 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2979 But simultaneously result of request was absolutely normal ldap result,
2980 so I produce this patch..."
2982 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2983 REFERENCE, so I have modified the code to exclude the patch when that macro
2986 55. Some experimental protocols are using DNS PTR records for new purposes. The
2987 keys for these records are domain names, not reversed IP addresses. The
2988 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2989 leaves it alone. Component reversal etc. now happens only for IP addresses.
2992 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2994 57. Double the size of the debug message buffer (to 2048) so that more of very
2995 long debug lines gets shown.
2997 58. The exicyclog utility now does better if the number of log files to keep
2998 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3000 59. Two changes related to the smtp_active_hostname option:
3002 (1) $smtp_active_hostname is now available as a variable.
3003 (2) The default for smtp_banner uses $smtp_active_hostname instead
3004 of $primary_hostname.
3006 60. The host_aton() function is supposed to be passed a string that is known
3007 to be a valid IP address. However, in the case of IPv6 addresses, it was
3008 not checking this. This is a hostage to fortune. Exim now panics and dies
3009 if the condition is not met. A case was found where this could be provoked
3010 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3011 components; fortuitously, this particular loophole had already been fixed
3012 by change 4.50/55 above.
3014 If there are any other similar loopholes, the new check in host_aton()
3015 itself should stop them being exploited. The report I received stated that
3016 data on the command line could provoke the exploit when Exim was running as
3017 exim, but did not say which command line option was involved. All I could
3018 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3019 running as the user.
3022 61. There was a buffer overflow vulnerability in the SPA authentication code
3023 (which came originally from the Samba project). I have added a test to the
3024 spa_base64_to_bits() function which I hope fixes it.
3027 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3028 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3030 63. The daemon start-up calls getloadavg() while still root for those OS that
3031 need the first call to be done as root, but it missed one case: when
3032 deliver_queue_load_max is set with deliver_drop_privilege. This is
3033 necessary for the benefit of the queue runner, because there is no re-exec
3034 when deliver_drop_privilege is set.
3036 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3037 This has been fixed.
3039 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3040 was in use, was not putting the data itself into the right store pool;
3041 consequently, it could be overwritten for a subsequent message in the same
3042 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3045 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3047 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3048 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3049 the function were treating the return as a boolean value, which happened to
3050 work because 0=false and not-0=true, but is not correct code.
3052 68. The host_aton() function was not handling scoped IPv6 addresses (those
3053 with, for example, "%eth0" on the end) correctly.
3055 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3056 negated items (that is, ~something) in unsigned ints. Some compilers
3057 apparently mutter when there is no cast.
3059 70. If an address verification called from an ACL failed, and did not produce a
3060 user-specific message (i.e. there was only a "system" message), nothing was
3061 put in $acl_verify_message. In this situation, it now puts the system
3064 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3065 session; change 4.31/43 added the unwanted input to the log line - except
3066 that it did not do this in the start of session case. It now does.
3068 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3069 This is wrong and can cause the other end to generate a synchronization
3070 error if it is another Exim or anything else that does the synchronization
3071 check. A QUIT command is no longer sent after a timeout.
3073 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3074 during host lookups.
3076 74. The defer_ok option of callout verification was not working if it was used
3077 when verifying addresses in header lines, that is, for this case:
3079 verify = header_sender/callout=defer_ok
3081 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3082 those file descriptors could be used for SMTP connections. If anything
3083 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3084 could be sent to the SMTP client, causing chaos. The daemon now opens
3085 stdin, stdout, and stderr to /dev/null when it puts itself into the
3088 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3089 log by default. The user can override this with suitable Perl magic.
3091 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3092 the log message when discard triggers, was not working for the DATA ACL or
3093 for the non-SMTP ACL.
3095 78. Error message wording change in sieve.c.
3097 79. If smtp_accept_max_per_host was set, the number of connections could be
3098 restricted to fewer than expected, because the daemon was trying to set up
3099 a new connection before checking whether the processes handling previous
3100 connections had finished. The check for completed processes is now done
3101 earlier. On busy systems, this bug wouldn't be noticed because something
3102 else would have woken the daemon, and it would have reaped the completed
3105 80. If a message was submitted locally by a user whose login name contained one
3106 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3107 It caused a spool format error. I have fixed the spool reading code. A
3108 related problem was that the "from" clause in the Received: line became
3109 illegal because of the space(s). It is now covered by ${quote_local_part.
3111 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3114 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3116 83. Patch from Sieve maintainer to fix unterminated string problem in
3117 "vacation" handling.
3119 84. Some minor changes to the Linux configuration files to help with other
3120 OS variants using glibc.
3122 85. One more patch for Sieve to update vacation handling to latest spec.
3125 ----------------------------------------------------
3126 See the note above about the 4.44 and 4.50 releases.
3127 ----------------------------------------------------
3133 1. Change 4.43/35 introduced a bug that caused file counts to be
3134 incorrectly computed when quota_filecount was set in an appendfile
3137 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3138 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3141 3. Give more explanation in the error message when the command for a transport
3142 filter fails to execute.
3144 4. There are several places where Exim runs a non-Exim command in a
3145 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3146 was being done only for the command run by the queryprogram router. It is
3147 now done for all such subprocesses. The other cases are: ${run, transport
3148 filters, and the commands run by the lmtp and pipe transports.
3150 5. Some older OS have a limit of 256 on the maximum number of file
3151 descriptors. Exim was using setrlimit() to set 1000 as a large value
3152 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3153 systems. I've change it so that if it can't get 1000, it tries for 256.
3155 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3156 was an oversight, and furthermore, ever since the addition of extra
3157 controls (e.g. 4.43/32), the checks on when to allow different forms of
3158 "control" were broken. There should now be diagnostics for all cases when a
3159 control that does not make sense is encountered.
3161 7. $recipients is now available in the predata ACL (oversight).
3163 8. Tidy the search cache before the fork to do a delivery from a message
3164 received from the command line. Otherwise the child will trigger a lookup
3165 failure and thereby defer the delivery if it tries to use (for example) a
3166 cached ldap connection that the parent has called unbind on.
3168 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3169 of $address_data from the recipient verification was clobbered by the
3170 sender verification.
3172 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3173 was its contents. (It was OK if the option was not defined at all.)
3175 11. A "Completed" log line is now written for messages that are removed from
3176 the spool by the -Mrm option.
3178 12. $host_address is now set to the target address during the checking of
3179 ignore_target_hosts.
3181 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3182 being passed; this would have caused $sender_host_name to have been used if
3183 matching the list had actually called for a host name (not very likely,
3184 since this list is usually IP addresses). A host name is now passed as
3187 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3188 code that specifies a non-restarting handler (typically sigaction() in
3189 modern systems) in an attempt to fix a rare and obscure crash bug.
3191 15. Narrowed the window for a race in the daemon that could cause it to ignore
3192 SIGCHLD signals. This is not a major problem, because they are used only to
3193 wake it up if nothing else does.
3195 16. A malformed maildirsize file could cause Exim to calculate negative values
3196 for the mailbox size or file count. Odd effects could occur as a result.
3197 The maildirsize information is now recalculated if the size or filecount
3200 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3201 support for a long time. Removed HAVE_SYS_VFS_H.
3203 18. Updated exipick to current release from John Jetmore.
3205 19. Allow an empty sender to be matched against a lookup in an address list.
3206 Previously the only cases considered were a regular expression, or an
3209 20. Exim went into a mad DNS lookup loop when doing a callout where the
3210 host was specified on the transport, if the DNS lookup yielded more than
3213 21. The RFC2047 encoding function was originally intended for short strings
3214 such as real names; it was not keeping to the 75-character limit for
3215 encoded words that the RFC imposes. It now respects the limit, and
3216 generates multiple encoded words if necessary. To be on the safe side, I
3217 have increased the buffer size for the ${rfc2047: expansion operator from
3220 22. Failure to deliver a bounce message always caused it to be frozen, even if
3221 there was an errors_to setting on the router. The errors_to setting is now
3224 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3225 canonical form (fully expanded) before being placed in
3226 $sender_host_address.
3228 24. Updated eximstats to version 1.33
3230 25. Include certificate and key file names in error message when GnuTLS fails
3231 to set them up, because the GnuTLS error message doesn't include the name
3232 of the failing file when there is a problem reading it.
3234 26. Expand error message when OpenSSL has problems setting up cert/key files.
3237 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3238 (this can affect the format of dates).
3240 28. exim_tidydb, when checking for the continued existence of a message for
3241 which it has found a message-specific retry record, was not finding
3242 messages that were in split spool directories. Consequently, it was
3243 deleting retry records that should have stayed in existence.
3245 29. eximstats updated to version 1.35
3246 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3247 1.35 - bugfix such that pie charts by volume are generated correctly
3249 30. The SPA authentication driver was not abandoning authentication and moving
3250 on to the next authenticator when an expansion was forced to fail,
3251 contradicting the general specification for all authenticators. Instead it
3252 was generating a temporary error. It now behaves as specified.
3254 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3255 (the order specifies the preference for clients). The order is now AES256,
3256 AES128, 3DES, ARCFOUR128.
3258 31. Small patch to Sieve code - explicitly set From: when generating an
3261 32. Exim crashed if a remote delivery caused a very long error message to be
3262 recorded - for instance if somebody sent an entire SpamAssassin report back
3263 as a large number of 550 error lines. This bug was coincidentally fixed by
3264 increasing the size of one of Exim's internal buffers (big_buffer) that
3265 happened as part of the Exiscan merge. However, to be on the safe side, I
3266 have made the code more robust (and fixed the comments that describe what
3269 33. Some experimental protocols are using DNS PTR records for new purposes. The
3270 keys for these records are domain names, not reversed IP addresses. The
3271 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3272 leaves it alone. Component reversal etc. now happens only for IP addresses.
3275 34. The host_aton() function is supposed to be passed a string that is known
3276 to be a valid IP address. However, in the case of IPv6 addresses, it was
3277 not checking this. This is a hostage to fortune. Exim now panics and dies
3278 if the condition is not met. A case was found where this could be provoked
3279 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3280 components; fortuitously, this particular loophole had already been fixed
3281 by change 4.50/55 or 4.44/33 above.
3283 If there are any other similar loopholes, the new check in host_aton()
3284 itself should stop them being exploited. The report I received stated that
3285 data on the command line could provoke the exploit when Exim was running as
3286 exim, but did not say which command line option was involved. All I could
3287 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3288 running as the user.
3291 35. There was a buffer overflow vulnerability in the SPA authentication code
3292 (which came originally from the Samba project). I have added a test to the
3293 spa_base64_to_bits() function which I hope fixes it.
3296 36. The daemon start-up calls getloadavg() while still root for those OS that
3297 need the first call to be done as root, but it missed one case: when
3298 deliver_queue_load_max is set with deliver_drop_privilege. This is
3299 necessary for the benefit of the queue runner, because there is no re-exec
3300 when deliver_drop_privilege is set.
3302 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3303 was in use, was not putting the data itself into the right store pool;
3304 consequently, it could be overwritten for a subsequent message in the same
3305 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3308 38. Sometimes the final signoff response after QUIT could fail to get
3309 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3310 < 0 before doing a fflush(). This bug looks as though it goes back to the
3311 introduction of TLS in release 3.20, but "sometimes" must have been rare
3312 because the tests only now provoked it.
3318 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3319 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3320 Now it returns an integer. A number of calls were still expecting a T/F
3321 return. Fortuitously, in all cases, the tests worked in OK situations,
3322 which is the norm. However, things would have gone wrong on any write
3323 failures on the smtp file descriptor. This function is used when sending
3324 messages over SMTP and also when doing verify callouts.
3326 2. When Exim is called to do synchronous delivery of a locally submitted
3327 message (the -odf or -odi options), it no longer closes stderr before doing
3330 3. Implemented the mua_wrapper option.
3332 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3334 5. Implemented the functions header_remove(), header_testname(),
3335 header_add_at_position(), and receive_remove_recipient(), and exported them
3338 6. If an ACL "warn" statement specified the addition of headers, Exim already
3339 inserted X-ACL-Warn: at the start if there was no header name. However, it
3340 was not making this test for the second and subsequent header lines if
3341 there were newlines in the string. This meant that an invalid header could
3342 be inserted if Exim was badly configured.
3344 7. Allow an ACL "warn" statement to add header lines at the start or after all
3345 the Received: headers, as well as at the end.
3347 8. Added the rcpt_4xx retry error code.
3349 9. Added postmaster_mailfrom=xxx to callout verification option.
3351 10. Added mailfrom=xxxx to the callout verification option, for verify=
3354 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3355 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3357 12. Inserted some casts to stop certain compilers warning when using pointer
3358 differences as field lengths or precisions in printf-type calls (mostly
3359 affecting debugging statements).
3361 13. Added optional readline() support for -be (dynamically loaded).
3363 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3364 same clock tick as a message's arrival, so that its received time was the
3365 same as the "first fail" time on the retry record, and that message
3366 remained on the queue past the ultimate address timeout, every queue runner
3367 would try a delivery (because it was past the ultimate address timeout) but
3368 after another failure, the ultimate address timeout, which should have then
3369 bounced the address, did not kick in. This was a "< instead of <=" error;
3370 in most cases the first failure would have been in the next clock tick
3371 after the received time, and all would be well.
3373 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3374 being recognized when the domain list was tested by the match_domain
3375 condition in an expansion string.
3377 16. Added the ${str2b64: operator.
3379 17. Exim was always calling setrlimit() to set a large limit for the number of
3380 processes, without checking whether the existing limit was already
3381 adequate. (It did check for the limit on file descriptors.) Furthermore,
3382 errors from getrlimit() and setrlimit() were being ignored. Now they are
3383 logged to the main and panic logs, but Exim does carry on, to try to do its
3384 job under whatever limits there are.
3386 18. Imported PCRE 5.0.
3388 19. Trivial typo in log message " temporarily refused connection" (the leading
3391 20. If the log selector return_path_on_delivery was set and an address was
3392 redirected to /dev/null, the delivery process crashed because it assumed
3393 that a return path would always be set for a "successful" delivery. In this
3394 case, the whole delivery is bypassed as an optimization, and therefore no
3397 21. Internal re-arrangement: the function for sending a challenge and reading
3398 a response while authentication was assuming a zero-terminated challenge
3399 string. It's now changed to take a pointer and a length, to allow for
3400 binary data in such strings.
3402 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3404 23. Exim was not respecting finduser_retries when seeking the login of the
3405 uid under which it was called; it was always trying 10 times. (The default
3406 setting of finduser_retries is zero.) Also, it was sleeping after the final
3407 failure, which is pointless.
3409 24. Implemented tls_on_connect_ports.
3411 25. Implemented acl_smtp_predata.
3413 26. If the domain in control=submission is set empty, Exim assumes that the
3414 authenticated id is a complete email address when it generates From: or
3415 Sender: header lines.
3417 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3418 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3419 chown and chgrp in /bin and hostname in /usr/bin.
3421 28. Exim was keeping the "process log" file open after each use, just as it
3422 does for the main log. This opens the possibility of it remaining open for
3423 long periods when the USR1 signal hits a daemon. Occasional processlog
3424 errors were reported, that could have been caused by this. Anyway, it seems
3425 much more sensible not to leave this file open at all, so that is what now
3428 29. The long-running daemon process does not normally write to the log once it
3429 has entered its main loop, and it closes the log before doing so. This is
3430 so that log files can straightforwardly be renamed and moved. However,
3431 there are a couple of unusual error situations where the daemon does write
3432 log entries, and I had neglected to close the log afterwards.
3434 30. The text of an SMTP error response that was received during a remote
3435 delivery was being truncated at 512 bytes. This is too short for some of
3436 the long messages that one sometimes sees. I've increased the limit to
3439 31. It is now possible to make retry rules that apply only when a message has a
3440 specific sender, in particular, an empty sender.
3442 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3443 it possible to be selective about when SMTP synchronization is enforced.
3445 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3447 32. Implemented hosts_connection_nolog.
3449 33. Added an ACL for QUIT.
3451 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3454 35. Added mailbox_size and mailbox_filecount to appendfile.
3456 36. Added control = no_multiline_responses to ACLs.
3458 37. There was a bug in the logic of the code that waits for the clock to tick
3459 in the case where the clock went backwards by a substantial amount such
3460 that the microsecond fraction of "now" was more than the microsecond
3461 fraction of "then" (but the whole seconds number was less).
3463 38. Added support for the libradius Radius client library this is found on
3464 FreeBSD (previously only the radiusclient library was supported).
3470 1. When certain lookups returned multiple values in the form name=value, the
3471 quoting of the values was not always being done properly. Specifically:
3472 (a) If the value started with a double quote, but contained no whitespace,
3474 (b) If the value contained whitespace other than a space character (i.e.
3475 tabs or newlines or carriage returns) it was not quoted.
3476 This fix has been applied to the mysql and pgsql lookups by writing a
3477 separate quoting function and calling it from the lookup code. The fix
3478 should probably also be applied to nisplus, ibase and oracle lookups, but
3479 since I cannot test any of those, I have not disturbed their existing code.
3481 2. A hit in the callout cache for a specific address caused a log line with no
3482 reason for rejecting RCPT. Now it says "Previous (cached) callout
3483 verification failure".
3485 3. There was an off-by-one bug in the queryprogram router. An over-long
3486 return line was truncated at 256 instead of 255 characters, thereby
3487 overflowing its buffer with the terminating zero. As well as fixing this, I
3488 have increased the buffer size to 1024 (and made a note to document this).
3490 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3491 when Exim is waiting for an SMTP response from a remote server, Exim
3492 restarts its select() call on the socket, thereby resetting its timeout.
3493 This is not a problem when such interrupts are rare. Somebody set up a cron
3494 job to run exiwhat every 2 minutes, which is less than the normal select()
3495 timeout (5 or 10 minutes). This meant that the select() timeout never
3496 kicked in because it was always reset. I have fixed this by comparing the
3497 time when an interrupt arrives with the time at the start of the first call
3498 to select(). If more time than the timeout has elapsed, the interrupt is
3499 treated as a timeout.
3501 5. Some internal re-factoring in preparation for the addition of Sieve
3502 extensions (by MH). In particular, the "personal" test is moved to a
3503 separate function, and given an option for scanning Cc: and Bcc: (which is
3504 not set for Exim filters).
3506 6. When Exim created an email address using the login of the caller as the
3507 local part (e.g. when creating a From: or Sender: header line), it was not
3508 quoting the local part when it contained special characters such as @.
3510 7. Installed new OpenBSD configuration files.
3512 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3513 try to make them clearer.
3515 9. Callout options, other than the timeout value, were being ignored when
3516 verifying sender addresses in header lines. For example, when using
3518 verify = header_sender/callout=no_cache
3520 the cache was (incorrectly) being used.
3522 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3523 only the Cygwin environment.
3525 11. When return_path_on_delivery was set as a log selector, if different remote
3526 addresses in the same message used different return paths and parallel
3527 remote delivery occurred, the wrong values would sometimes be logged.
3528 (Whenever a remote delivery process finished, the return path value from
3529 the most recently started remote delivery process was logged.)
3531 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3532 lines when AUTH and/or TLS are in use. This is the "received protocol"
3533 field. Exim used to use "asmtp" for authenticated SMTP, without any
3534 indication (in the protocol name) for TLS use. Now it follows the RFC and
3535 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3536 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3537 names appear in log lines as well as in Received: header lines.
3539 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3540 extensions, and comparison tests, and to fix some bugs.
3542 14. Changes to the "personal" filter test:
3544 (1) The test was buggy in that it was just doing the equivalent of
3545 "contains" tests on header lines. For example, if a user's address was
3546 anne@some.where, the "personal" test would incorrectly be true for
3548 To: susanne@some.where
3550 This test is now done by extracting each address from the header in turn,
3551 and checking the entire address. Other tests that are part of "personal"
3552 are now done using regular expressions (for example, to check local parts
3553 of addresses in From: header lines).
3555 (2) The list of non-personal local parts in From: addresses has been
3556 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3557 taken from the Sieve specification recommendations.
3559 (3) If the message contains any header line starting with "List-" it is
3560 treated as non-personal.
3562 (4) The test for "circular" in the Subject: header line has been removed
3563 because it now seems ill-conceived.
3565 15. Minor typos in src/EDITME comments corrected.
3567 16. Installed latest exipick from John Jetmore.
3569 17. If headers_add on a router specified a text string that was too long for
3570 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3571 of string_sprintf() is now avoided.
3573 18. $message_body_size was not set (it was always zero) when running the DATA
3574 ACL and the local_scan() function.
3576 19. For the "mail" command in an Exim filter, no default was being set for
3577 the once_repeat time, causing a random time value to be used if "once" was
3578 specified. (If the value happened to be <= 0, no repeat happened.) The
3579 default is now 0s, meaning "never repeat". The "vacation" command was OK
3580 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3581 (I found it when inspecting the code).
3583 20. There is now an overall timeout for performing a callout verification. It
3584 defaults to 4 times the callout timeout, which applies to individual SMTP
3585 commands during the callout. The overall timeout applies when there is more
3586 than one host that can be tried. The timeout is checked before trying the
3587 next host. This prevents very long delays if there are a large number of
3588 hosts and all are timing out (e.g. when the network connections are timing
3589 out). The value of the overall timeout can be changed by specifying an
3590 additional sub-option for "callout", called "maxwait". For example:
3592 verify = sender/callout=5s,maxwait=20s
3594 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3595 to the end before writing, but this should make it even safer).
3597 22. Exim was forgetting that it had advertised PIPELINING for the second and
3598 subsequent messages on an SMTP connection. It was also not resetting its
3599 memory on STARTTLS and an internal HELO.
3601 23. When Exim logs an SMTP synchronization error within a session, it now
3602 records whether PIPELINING has been advertised or not.
3604 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3605 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3606 rather than long int.
3608 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3610 26. Added the never_mail option to autoreply.
3616 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3617 crash if the getsockname() call failed; this can happen if a connection is
3618 closed very soon after it is established. The problem was simply in the
3619 order in which certain operations were done, causing Exim to try to write
3620 to the SMTP stream before it had set up the file descriptor. The bug has
3621 been fixed by making things happen in the correct order.
3627 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3628 before the connection was closed, thus losing the rejection response.
3630 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3631 some early Solaris releases, but causes trouble in current releases where
3632 socklen_t is defined.
3634 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3637 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3640 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3641 that releasing the top of it at the end releases what was used for sub-
3642 expansions (unless the block got too big). However, discard this block if
3643 the first thing is a variable or header, so that we can use its block when
3644 it is dynamic (useful for very large $message_headers, for example).
3646 6. Lookups now cache *every* query, not just the most recent. A new, separate
3647 store pool is used for this. It can be recovered when all lookup caches are
3648 flushed. Lookups now release memory at the end of their result strings.
3649 This has involved some general refactoring of the lookup sources.
3651 7. Some code has been added to the store_xxx() functions to reduce the amount
3652 of flapping under certain conditions.
3654 8. log_incoming_interface used to affect only the <= reception log lines. Now
3655 it causes the local interface and port to be added to several more SMTP log
3656 lines, for example "SMTP connection from", and rejection lines.
3658 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3660 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3662 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3663 could be overwritten at the end of the current message (or the start of a
3664 new message if it was set in a HELO ACL). The value is now preserved for
3665 the duration of the SMTP connection.
3667 12. If a transport had a headers_rewrite setting, and a matching header line
3668 contained an unqualified address, that address was qualified, even if it
3669 did not match any rewriting rules. The underlying bug was that the values
3670 of the flags that permit the existence of unqualified sender and recipient
3671 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3672 non-local messages, and by -bnq for local messages) were not being
3673 preserved with the message after it was received.
3675 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3676 "next input=" as part of the text comprising the host identity instead of
3677 the correct text. The code was using the same buffer for two different
3678 strings. However, depending on which order the printing function evaluated
3679 its arguments, the bug did not always show up. Under Linux, for example, my
3680 test suite worked just fine.
3682 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3683 doesn't work with some older versions of Perl. It has been changed to "my",
3684 which in any case is probably the better facility to use.
3686 15. A really picky compiler found some instances of statements for creating
3687 error messages that either had too many or two few arguments for the format
3690 16. The size of the buffer for calls to the DNS resolver has been increased
3691 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3692 for addresses that have a lot of PTR records. This alleviates a problem; it
3693 does not fully solve it.
3695 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3696 buffer now truncates the list and logs the incident, which is the same
3697 action as happens when Exim is looking up a host name and its aliases.
3698 Previously in this situation something unpredictable would happen;
3699 sometimes it was "internal error: store_reset failed".
3701 18. If a server dropped the connection unexpectedly when an Exim client was
3702 using GnuTLS and trying to read a response, the client delivery process
3703 crashed while trying to generate an error log message.
3705 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3706 string, for example:
3708 warn message = H1: something\nH2: something
3710 the text was added as a single header line from Exim's point of view
3711 though it ended up OK in the delivered message. However, searching for the
3712 second and subsequent header lines using $h_h2: did not work. This has been
3713 fixed. Similarly, if a system filter added multiple headers in this way,
3714 the routers could not see them.
3716 20. Expanded the error message when iplsearch is called with an invalid key to
3717 suggest using net-iplsearch in a host list.
3719 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3720 ACLs are no longer actually imposed (and a message to that effect is
3723 22. If a "gecos" field in a passwd entry contained escaped characters, in
3724 particular, if it contained a \" sequence, Exim got it wrong when building
3725 a From: or a Sender: header from that name. A second bug also caused
3726 incorrect handling when an unquoted " was present following a character
3727 that needed quoting.
3729 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3730 was not being matched caselessly.
3732 24. Arranged for all hyphens in the exim.8 source to be escaped with
3735 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3736 properly. Recipient callout cache records were still being keyed to include
3737 the sender, even when use_sender was set false. This led to far more
3738 callouts that were necessary. The sender is no longer included in the key
3739 when use_sender is false.
3741 26. Added "control = submission" modifier to ACLs.
3743 27. Added the ${base62d: operator to decode base 62 numbers.
3745 28. dnsdb lookups can now access SRV records.
3747 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3748 the configuration file.
3750 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3751 (-v) mode. This makes the output for a verbose queue run more intelligible.
3753 31. Added a use_postmaster feature to recipient callouts.
3755 32. Added the $body_zerocount variable, containing the number of binary zero
3756 bytes in the message body.
3758 33. The time of last modification of the "new" subdirectory is now used as the
3759 "mailbox time last read" when there is a quota error for a maildir
3762 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3764 35. Added +ignore_unknown as a special item in host lists.
3766 36. Code for decoding IPv6 addresses in host lists is now included, even if
3767 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3768 address was recognized as an IP address, but was then not correctly decoded
3769 into binary, causing unexpected and incorrect effects when compared with
3776 1. Very minor rewording of debugging text in manualroute to say "list of
3777 hosts" instead of "hostlist".
3779 2. If verify=header_syntax was set, and a header line with an unqualified
3780 address (no domain) and a large number of spaces between the end of the
3781 name and the colon was received, the reception process suffered a buffer
3782 overflow, and (when I tested it) crashed. This was caused by some obsolete
3783 code that should have been removed. The fix is to remove it!
3785 3. When running in the test harness, delay a bit after writing a bounce
3786 message to get a bit more predictability in the log output.
3788 4. Added a call to search_tidyup() just before forking a reception process. In
3789 theory, someone could use a lookup in the expansion of smtp_accept_max_
3790 per_host which, without the tidyup, could leave open a database connection.
3792 5. Added the variables $recipient_data and $sender_data which get set from a
3793 lookup success in an ACL "recipients" or "senders" condition, or a router
3794 "senders" option, similar to $domain_data and $local_part_data.
3796 6. Moved the writing of debug_print from before to after the "senders" test
3799 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3800 problems for message scanning, either using a data ACL, or using
3801 local_scan() because the Received: header was not generated till after they
3802 were called (in order to set the time as the time of reception completion).
3803 I have revised the way this works. The header is now generated after the
3804 body is received, but before the ACL or local_scan() are called. After they
3805 are run, the timestamp in the header is updated.
3811 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3812 before starting a queue runner without re-exec. This happened only when
3813 deliver_drop_privilege was set or when the Exim user was set to root. The
3814 effect of the bug was that timeouts during subsequent deliveries caused
3815 crashes instead of being properly handled. The handler is now left at its
3816 default (and expected) setting.
3818 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3819 message, again when deliver_drop_privilege is set or Exim is run as root.
3820 The bug described in (1) was not present in this case, but the tidying up
3821 of the other signals was missing. I have made the two cases consistent.
3823 3. The ignore_target_hosts setting on a manualroute router was being ignored
3824 for hosts that were looked up using the /MX notation.
3826 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3829 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3830 operated on the sender address. After changing the $sender_address to <>
3831 for the sender address verify, Exim was re-instated it as the original
3832 (before rewriting) address, but remembering that it had rewritten it, so it
3833 wasn't rewriting it again. This bug also had the effect of breaking the
3834 sender address verification caching when the sender address was rewritten.
3836 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3837 This has been changed so that if the ip literal address matches
3838 ignore_target_hosts, the router declines.
3840 7. Added expansion conditions match_domain, match_address, and match_local_
3841 part (NOT match_host).
3843 8. The placeholder for the Received: header didn't have a length field set.
3845 9. Added code to Exim itself and to exim_lock to test for a specific race
3846 condition that could lead to file corruption when using MBX delivery. The
3847 issue is with the lockfile that is created in /tmp. If this file is removed
3848 after a process has opened it but before that process has acquired a lock,
3849 there is the potential for a second process to recreate the file and also
3850 acquire a lock. This could lead to two Exim processes writing to the file
3851 at the same time. The added code performs the same test as UW imapd; it
3852 checks after acquiring the lock that its file descriptor still refers to
3853 the same named file.
3855 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3856 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3859 11. Added the smtp_active_hostname option. If used, this will typically be made
3860 to depend on the incoming interface address. Because $interface_address is
3861 not set up until the daemon has forked a reception process, error responses
3862 that can happen earlier (such as "too many connections") no longer contain
3865 12. If an expansion in a condition on a "warn" statement fails because a lookup
3866 defers, the "warn" statement is abandoned, and the next ACL statement is
3867 processed. Previously this caused the whole ACL to be aborted.
3869 13. Added the iplsearch lookup type.
3871 14. Added ident_timeout as a log selector.
3873 15. Added tls_certificate_verified as a log selector.
3875 16. Added a global option tls_require_ciphers (compare the smtp transport
3876 option of the same name). This controls incoming TLS connections.
3878 17. I finally figured out how to make tls_require_ciphers do a similar thing
3879 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3880 before starting the TLS session.
3882 18. Tabs are now shown as \t in -bP output.
3884 19. If the log selector return_path_on_delivery was set, Exim crashed when
3885 bouncing a message because it had too many Received: header lines.
3887 20. If two routers both had headers_remove settings, and the first one included
3888 a superfluous trailing colon, the final name in the first list and the
3889 first name in the second list were incorrectly joined into one item (with a
3890 colon in the middle).
3896 1. Added -C and -D options to the exinext utility, mainly to make it easier
3897 to include in the automated testing, but these could be helpful when
3898 multiple configurations are in use.
3900 2. The exinext utility was not formatting the output nicely when there was
3901 an alternate port involved in the retry record key, nor when there was a
3902 message id as well (for retries that were specific to a specific message
3903 and a specific host). It was also confused by IPv6 addresses, because of
3904 the additional colons they contain. I have fixed the IPv4 problem, and
3905 patched it up to do a reasonable job for IPv6.
3907 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3908 delivery, the log line now contains "pipelined" if PIPELINING was used.
3910 4. An SMTP transport process used to panic and die if the bind() call to set
3911 an explicit outgoing interface failed. This has been changed; it is now
3912 treated in the same way as a connect() failure.
3914 5. A reference to $sender_host_name in the part of a conditional expansion
3915 that was being skipped was still causing a DNS lookup. This no longer
3918 6. The def: expansion condition was not recognizing references to header lines
3919 that used bh_ and bheader_.
3921 7. Added the _cache feature to named lists.
3923 8. The code for checking quota_filecount in the appendfile transport was
3924 allowing one more file than it should have been.
3926 9. For compatibility with Sendmail, the command line option
3934 and sets the incoming protocol and host name (for trusted callers). The
3935 host name and its colon can be omitted when only the protocol is to be set.
3936 Note the Exim already has two private options, -pd and -ps, that refer to
3937 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3938 "s", but I don't think that's a major issue.
3940 10. A number of refactoring changes to the code, none of which should affect
3943 (a) The number of logging options was getting close to filling up the
3944 32-bit word that was used as a bit map. I have split them into two classes:
3945 those that are passed in the argument to log_write(), and those that are
3946 only ever tested independently outside of that function. These are now in
3947 separate 32-bit words, so there is plenty of room for expansion again.
3948 There is no change in the user interface or the logging behaviour.
3950 (b) When building, for example, log lines, the code previously used a
3951 macro that called string_cat() twice, in order to add two strings. This is
3952 not really sufficiently general. Furthermore, there was one instance where
3953 it was actually wrong because one of the argument was used twice, and in
3954 one call a function was used. (As it happened, calling the function twice
3955 did not affect the overall behaviour.) The macro has been replaced by a
3956 function that can join an arbitrary number of extra strings onto a growing
3959 (c) The code for expansion conditions now uses a table and a binary chop
3960 instead of a serial search (which was left over from when there were very
3961 few conditions). Also, it now recognizes conditions like "pam" even when
3962 the relevant support is not compiled in: a suitably worded error message is
3963 given if an attempt is made to use such a condition.
3965 11. Added ${time_interval:xxxxx}.
3967 12. A bug was causing one of the ddress fields not to be passed back correctly
3968 from remote delivery subprocesses. The field in question was not being
3969 subsequently used, so this caused to problems in practice.
3971 13. Added new log selectors queue_time and deliver_time.
3973 14. Might have fixed a bug in maildirsizefile handling that threw up
3974 "unexpected character" debug warnings, and recalculated the data
3975 unnecessarily. In any case, I expanded the warning message to give more
3978 15. Added the message "Restricted characters in address" to the statements in
3979 the default ACL that block characters like @ and % in local parts.
3981 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3982 Three changes have been made:
3984 (a) There was a serious bug; a negative response to MAIL caused the whole
3985 recipient domain to be cached as invalid, thereby blocking all messages
3986 to all local parts at the same domain, from all senders. This bug has
3987 been fixed. The domain is no longer cached after a negative response to
3988 MAIL if the sender used is not empty.
3990 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3993 (c) A new callout option, "use_sender" has been added for people who want
3994 the modified behaviour.
4000 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4003 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4004 indeed breaks things for older releases.
4006 3. Added additional logging to the case where there is a problem reading data
4007 from a filter that is running in a subprocess using a pipe, in order to
4008 try to track down a specific problem.
4010 4. Testing facility fudge: when running in the test harness and attempting
4011 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4012 getting "No route to host". Convert this to a timeout.
4014 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4017 6. Some OS don't have socklen_t but use size_t instead. This affects the
4018 fifth argument of getsockopt() amongst other things. This is now
4019 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4020 can be set for individual OS. I have set it for SunOS5, OSF1, and
4021 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4022 some earlier ones do not.
4024 7. Change 4.30/15 was not doing the test caselessly.
4026 8. The standard form for an IPv6 address literal was being rejected by address
4027 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4028 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4029 this, as well as the form without the "IPv6" on the front (but only when
4030 address literals are enabled, of course).
4032 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4034 10. Exim crashed if a message with an empty sender address specified by -f
4035 encountered a router with an errors_to setting. This could be provoked only
4036 by a command such as
4040 where an empty string was supplied; "<>" did not hit this bug.
4042 11. Installed PCRE release 4.5.
4044 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4045 remained set. It is now erased.
4047 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4048 times from message ids (which are base 36 rather than the normal 62).
4050 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4051 were being counted as actual protocol errors, and logged if the log
4052 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4053 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4054 and DATA following a set of rejected RCPTs do not count as protocol errors.
4055 In other words, Exim assumes they were pipelined, though this may not
4056 actually be the case. Of course, in all cases the client gets an
4057 appropriate error code.
4059 15. If a lookup fails in an ACL condition, a message about the failure may
4060 be available; it is used if testing the ACL cannot continue, because most
4061 such messages specify what the cause of the deferral is. However, some
4062 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4063 that caused an old message to be retained and used if a later statement
4064 caused a defer, replacing the real cause of the deferral.
4066 16. If an IP address had so many PTR records that the DNS lookup buffer
4067 was not large enough to hold them, Exim could crash while trying to process
4068 the truncated data. It now detects and logs this case.
4070 17. Further to 4.21/58, another change has been made: if (and only if) the
4071 first line of a message (the first header line) ends with CRLF, a bare LF
4072 in a subsequent header line has a space inserted after it, so as not to
4073 terminate the header.
4075 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4076 unnecessarily, used atoi() instead of strtol(), and didn't check the
4077 termination when getting file sizes from file names by regex.
4079 19. Completely re-implemented the support for maildirsize files, in the light
4080 of a number of problems with the previous contributed implementation
4081 (4.30/29). In particular:
4083 . If the quota is zero, the maildirsize file is maintained, but no quota is
4086 . If the maildir directory does not exist, it is created before any attempt
4087 to write a maildirsize file.
4089 . The quota value in the file is just a cache; if the quota is changed in
4090 the transport, the new value overrides.
4092 . A regular expression is available for excluding directories from the
4095 20. The autoreply transport checks the characters in options that define the
4096 message's headers; it allows continued headers, but it was checking with
4097 isspace() after an embedded newline instead of explicitly looking for a
4100 21. If all the "regular" hosts to which an address was routed had passed their
4101 expiry times, and had not reached their retry times, the address was
4102 bounced, even if fallback hosts were defined. Now Exim should go on to try
4105 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4106 equivalent code in the SMTP transport. Some hosts send humungous responses
4107 to HELO/EHLO, more than 1024 it seems.
4109 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4110 gives compiler warnings in some environments. I've now done it "properly",
4113 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4114 (because of problems with the built-in one) was declared to return uschar *
4115 instead of char *, causing compiler failure.
4117 25. Fixed a file descriptor leak when processing alias/forward files.
4119 26. Fixed a minor format string issue in dbfn.c.
4121 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4123 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4124 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4127 29. When a sender address is verified, it is cached, to save repeating the test
4128 when there is more than one recipient in a message. However, when the
4129 verification involves a callout, it is possible for different callout
4130 options to be set for different recipients. It is too complicated to keep
4131 track of this in the cache, so now Exim always runs a verification when a
4132 callout is required, relying on the callout cache for the optimization.
4133 The overhead is duplication of the address routing, but this should not be
4136 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4137 to be verified with callout=postmaster, and the main callout worked but the
4138 postmaster check failed, the verification correctly failed. However, if a
4139 subsequent RCPT command asked for sender verification *without* the
4140 postmaster check, incorrect caching caused this verification also to fail,
4143 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4144 it was not caching the DNS options (qualify_single, search_parents) that
4145 were used when the lookup failed. A subsequent lookup with different
4146 options therefore always gave the same answer, though there were cases
4147 where it should not have. (Example: a "domains = !$mx_any" option on a
4148 dnslookup router: the "domains" option is always processed without any
4149 widening, but the router might have qualify_single set.) Now Exim uses the
4150 cached value only when the same options are set.
4152 32. Added John Jetmore's "exipick" utility to the distribution.
4154 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4155 than a timeout (e.g. a certificate is required, and is not provided), an
4156 Exim server now closes the connection immediately. Previously it waited for
4157 the client to close - but if the client is SSL, it seems that they each
4158 wait for each other, leading to a delay before one of them times out.
4160 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4161 maintained 0.8.x compatibility because I don't think many are using it, and
4162 it is clearly obsolete.
4164 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4167 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4168 client certificate was expired. A simple patch fixes this, though I don't
4169 understand the full logic of why the verify callback is called multiple
4172 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4173 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4174 which causes problems with some clients (such as the Certicom SSL Plus
4175 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4176 disables the coutermeasure allowing Eudora to connect."
4178 38. Exim was not checking that a write() to a log file succeeded. This could
4179 lead to Bad Things if a log got too big, in particular if it hit a file
4180 size limit. Exim now panics and dies if it cannot write to a log file, just
4181 as it does if it cannot open a log file.
4183 39. Modified OS/Makefile-Linux so that it now contains
4185 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4187 The two -D definitions ensure that Exim is compiled with large file
4188 support, which makes it possible to handle log files that are bigger than
4191 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4192 instance) a domain was checked against a named list that involved a lookup,
4193 causing $domain_data to be set, then another domain was checked against the
4194 same list, then the first domain was re-checked, the value of $domain_data
4195 after the final check could be wrong. In particular, if the second check
4196 failed, it could be set empty. This bug probably also applied to
4199 41. The strip_trailing_dot option was not being applied to the address given
4200 with the -f command-line option.
4202 42. The code for reading a message's header from the spool was incrementing
4203 $received_count, but never initializing it. This meant that the value was
4204 incorrect (doubled) while delivering a message in the same process in which
4205 it was received. In the most common configuration of Exim, this never
4206 happens - a fresh exec is done - but it can happen when
4207 deliver_drop_privilege is set.
4209 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4210 it now includes up to 150 characters of the unexpected data in the log
4213 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4214 and building data strings. The size of both of these buffers was 10 000
4215 bytes - far larger than anybody would *ever* want, thought I. Needless to
4216 say, somebody hit the limit. I have increased the maximum line length to
4217 20 000 and the maximum data length of concatenated lines to 100 000. I have
4218 also fixed two bugs, because there was no checking on these buffers. Tsk,
4219 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4220 buffer is too small.
4222 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4223 lsearch lookups. Now it does.
4225 46. When parsing a route_list item in a manualroute router, a fixed-length
4226 buffer was used for the list of hosts. I made this 1024 bytes long,
4227 thinking that nobody would ever have a list of hosts that long. Wrong.
4228 Somebody had a whole pile of complicated expansion conditions, and the
4229 string was silently truncated, leading to an expansion error. It turns out
4230 that it is easier to change to an unlimited length (owing to other changes
4231 that have happened since this code was originally written) than to build
4232 structure for giving a limitation error. The length of the item that
4233 expands into the list of hosts is now unlimited.
4235 47. The lsearch lookup could not handle data where the length of text line was
4236 more than 4095 characters. Such lines were truncated, leading to shortened
4237 data being returned. It should now handle lines of any length.
4239 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4240 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4243 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4244 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4245 no longer happens in comment lines. A list of replacements is now placed
4246 at the head of all of the source files, except those whose only change is
4247 to replace PERL_COMMAND in the very first #! line.
4249 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4250 messages on the queue, with a bottom-up merge sort, using code contributed
4251 by Michael Haardt. This should make operations like -bp somewhat faster on
4252 large queues. It won't affect queue runners, except when queue_run_in_order
4255 51. Installed eximstats 1.31 in the distribution.
4257 52. Added support for SRV lookups to the dnslookup router.
4259 53. If an ACL referred to $message_body or $message_body_end, the value was not
4260 reset for any messages that followed in the same SMTP session.
4262 54. The store-handling optimization for building very long strings was not
4263 differentiating between the different store pools. I don't think this
4264 actually made any difference in practice, but I've tidied it.
4266 55. While running the routers to verify a sender address, $sender_address
4267 was still set to the sender address. This is wrong, because when routing to
4268 send a bounce to the sender, it would be empty. Therefore, I have changed
4269 it so that, while verifying a sender address, $sender_address is set to <>.
4270 (There is no change to what happens when verifying a recipient address.)
4272 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4273 target A or AAAA records (if not already returned) without resetting the
4274 qualify_single or search_parents options of the DNS resolver. These are
4275 inappropriate in this case because the targets of MX and SRV records must
4276 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4277 target that, when qualified, matched something in the local domain. These
4278 two options are now turned off when doing these lookups.
4280 57. It seems that at least some releases of Reiserfs (which does not have the
4281 concept of a fixed number of inodes) returns zero and not -1 for the
4282 number of available inodes. This interacted badly with check_spool_inodes,
4283 which assumed that -1 was the "no such thing" setting. What I have done is
4284 to check that the total number of inodes is greater than zero before doing
4285 the test of how many are available.
4287 58. When a "warn" ACL statement has a log_message modifier, the message is
4288 remembered, and not repeated. This is to avoid a lot of repetition when a
4289 message has many recipients that cause the same warning to be written.
4290 Howewer, Exim was preserving the list of already written lines for an
4291 entire SMTP session, which doesn't seem right. The memory is now reset if a
4292 new message is started.
4294 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4295 debugging output unless log_rewrite was also set.
4297 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4298 of (int)(handle) when we know that handle contains (void *)(-1).
4300 61. The Exim daemon panic-logs an error return when it closes the incoming
4301 connection. However "connection reset by peer" seems to be common, and
4302 isn't really an error worthy of noting specially, so that particular error
4305 62. When Exim is trying to find all the local interfaces, it used to panic and
4306 die if the ioctl to get the interface flags failed. However, it seems that
4307 on at least one OS (Solaris 9) it is possible to have an interface that is
4308 included in the list of interfaces, but for which you get a failure error
4309 for this call. This happens when the interface is not "plumbed" into a
4310 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4311 failure of the "get flags" call assumes that the interface is down.
4313 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4314 makes life easier for people who are doing arithmetic on fields extracted
4315 from dates, where you often get leading zeros that should not be
4316 interpreted as octal.
4318 64. Added qualify_domain to the redirect router, to override the global
4321 65. If a pathologically long header line contained very many addresses (the
4322 report of this problem mentioned 10 000) and each of them was rewritten,
4323 Exim could use up a very large amount of memory. (It kept on making new
4324 copies of the header line as it rewrote, and never released the old ones.)
4325 At the expense of a bit more processing, the header rewriting function has
4326 been changed so that it no longer eats memory in this way.
4328 66. The generation of the Received: header has been moved from the time that a
4329 message starts to be received, to the time that it finishes. The timestamp
4330 in the Received: header should now be very close to that of the <= log
4331 line. There are two side-effects of this change:
4333 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4334 logged header lines no longer include the local Received: line, because
4335 it has not yet been created. The same applies to a copy of the message
4336 that is returned to a non-SMTP sender when a message is rejected.
4338 (b) When a filter file is tested using -bf, no additional Received: header
4339 is added to the test message. After some thought, I decided that this
4342 This change does not affect the value of $received_for. It is still set
4343 after address rewriting, but before local_scan() is called.
4345 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4347 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4348 gave an unhelpful panic error message, and a defer error. I have managed to
4349 change this behaviour so that it now rejects any supplied certificate,
4350 which seems right, as the list of acceptable certificates is empty.
4352 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4353 gave an unhelpful defer error. I have not managed to make this reject any
4354 supplied certificates, but the error message it gives is "no certificate
4355 supplied", which is not helpful.
4357 70. exigrep's output now also includes lines that are not associated with any
4358 message, but which match the given pattern. Implemented by a patch from
4359 Martin Sluka, which also tidied up the Perl a bit.
4361 71. Recipient callout verification, like sender verification, was using <> in
4362 the MAIL FROM command. This isn't really the right thing, since the actual
4363 sender may affect whether the remote host accepts the recipient or not. I
4364 have changed it to use the actual sender in the callout; this means that
4365 the cache record is now keyed on a recipient/sender pair, not just the
4366 recipient address. There doesn't seem to be a real danger of callout loops,
4367 since a callout by the remote host to check the sender would use <>.
4368 [SEE ABOVE: changed after hitting problems.]
4370 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4371 temporary errors. However, in the case of such a code being given after
4372 the end of a data transmission (i.e. after ".") Exim was failing to write
4373 a retry record for the message. (Yes, there was some broken host that was
4374 actually sending 8xx at this point.)
4376 73. An unknown lookup type in a host list could cause Exim to panic-die when
4377 the list was checked. (An example that provoked this was putting <; in the
4378 middle of a list instead of at the start.) If this happened during a DATA
4379 ACL check, a -D file could be left lying around. This kind of configuration
4380 error no longer causes Exim to die; instead it causes a defer errror. The
4381 incident is still logged to the main and panic logs.
4383 74. Buglet left over from Exim 3 conversion. The message "too many messages
4384 in one connection" was written to the rejectlog but not the mainlog, except
4385 when address rewriting (yes!) was being logged.
4387 75. Added write_rejectlog option.
4389 76. When a system filter was run not as root (that is, when system_filter_user
4390 was set), the values of the $n variables were not being returned to the
4391 main process; thus, they were not subsequently available in the $sn
4394 77. Added +return_path_on_delivery log selector.
4396 78. A connection timeout was being treated differently from recipients deferred
4397 when testing hosts_max_try with a message that was older than the host's
4398 retry timeout. (The host should not be counted, thus allowing all hosts to
4399 be tried at least once before bouncing.) This may have been the cause of an
4400 occasionally reported bug whereby a message would remain on the queue
4401 longer than the retry timeout, but would be bounced if a delivery was
4402 forced. I say "may" because I never totally pinned down the problem;
4403 setting up timeout/retry tests is difficult. See also the next item.
4405 79. The ultimate address timeout was not being applied to errors that involved
4406 a combination of host plus message (for example, a timeout on a MAIL
4407 command). When an address resolved to a number of possible hosts, and they
4408 were not all tried for each delivery (e.g. because of hosts_max_try), a
4409 message could remain on the queue longer than the retry timeout.
4411 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4414 81. Fixed an obscure SMTP outgoing bug which required at least the following
4415 conditions: (a) there was another message waiting for the same server;
4416 (b) the server returned 5xx to all RCPT commands in the first message so
4417 that the message was not completed; (c) the server dropped the connection
4418 or gave a negative response to the RSET that Exim sends to abort the
4419 transaction. The observed case was a dropped connection after DATA that had
4420 been sent in pipelining mode. That is, the server had advertised PIPELINING
4421 but was not implementing it correctly. The effect of the bug was incorrect
4422 behaviour, such as trying another host, and this could lead to a crash.
4428 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4429 and daemon.c were passed as pointers to ints; they should have been
4430 pointers to socklen_t variables (which are typically unsigned ints).
4432 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4435 3. Fixed a really odd bug that affected only the testing scheme; patching a
4436 certain fixed string in the binary changed the value of another string that
4437 happened to be identical to the end of the original first string.
4439 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4440 name", it returns that address as the IP address. On some operating
4441 systems (e.g. Solaris), it also passes back the IP address string as the
4442 "host name". However, on others (e.g. Linux), it passes back an empty
4443 string. Exim wasn't checking for this, and was changing the host name to an
4444 empty string, assuming it had been canonicized.
4446 5. Although rare, it is permitted to have more than one PTR record for a given
4447 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4448 all the names associated with an address, because they do in Solaris.
4449 However, it seems that they do not in Linux for data that comes from the
4450 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4451 I found this out when I moved to a new Linux workstation and tried to run
4452 the Exim test suite.
4454 To get round this problem I have changed the code so that it now does its
4455 own call to the DNS to look up PTR records when searching for a host name.
4456 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4457 addresses that are only in /etc/hosts are still found.
4459 This behaviour is, however, controlled by an option called host_lookup_
4460 order, which defaults to "bydns:byaddr". If people want to use the other
4461 order, or indeed, just use one or the other means of lookup, they can
4462 specify it in this variable.
4464 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4465 some operating systems, this comes back from gethostbyaddr() as an empty
4466 string, and this is what Exim used to test for. However, it seems that in
4467 other systems, "." is yielded. Exim now tests for this case too.
4469 7. The values of check_spool_space and check_log_space are now held internally
4470 as a number of kilobytes instead of an absolute number of bytes. If a
4471 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4472 kilobyte. This means that much larger values can be stored.
4474 8. Exim monitor: an attempt to get the action menu when not actually pointing
4475 at a message produces an empty menu entitled "No message selected". This
4476 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4477 no entries in it ("Shell widget menu has zero width and/or height"). So I
4478 have added a single, blank menu entry in this case.
4480 9. Added ${quote_local_part.
4482 10. MIME decoding is now applied to the contents of Subject: header lines when
4485 11. Now that a reference to $sender_host_address automatically causes a reverse
4486 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4487 host lookup before query-style lookups in lists that might use this
4488 variable. This has therefore been abolished, and the "net-" prefix is no
4489 longer necessary for query-style lookups.
4491 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4492 have been a typo for LFLAGS, so it has been changed.
4494 13. The install script calls Exim with "-C /dev/null" in order to find the
4495 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4496 to be output. Howeve, since Exim outputs its version number before the
4497 error, it didn't break the script. It just looked ugly. I fixed this by
4498 always allowing "-C /dev/null" if the caller is root.
4500 14. Ignore overlarge ACL variable number when reading spool file - insurance
4501 against a later release with more variables having written the file.
4503 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4504 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4505 this, as well as the form without the "IPv6" on the front.
4507 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4508 OS/Makefile-Darwin file.
4510 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4511 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4513 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4514 variable. However, some OS use ints for inodes. I've added cast to long int
4515 to get rid of the compiler warning.
4517 19. I had forgotten to lock out "/../" in configuration file names when
4518 ALT_CONFIG_PREFIX was set.
4520 20. Routers used for verification do not need to specify transports. However,
4521 if such a router generated a host list, and callout was configured, Exim
4522 crashed, because it could not find a port number from the (non-existent)
4523 transport. It now assumes port 25 in this circumstance.
4525 21. Added the -t option to exigrep.
4527 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4528 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4529 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4530 them is set, LOOKUP_LSEARCH is forced.
4532 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4534 24. Added sender and host information to the "rejected by local_scan()" log
4535 line; previously there was no indication of these.
4537 25. Added .include_if_exists.
4539 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4540 Linux. It turns out that not all file systems support this. Apparently some
4541 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4542 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4543 not supported on the file descriptor, is now ignored when Exim is trying to
4544 sync a directory. This applies only to Linux.
4546 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4548 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4549 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4552 29. Added support for maildirsize files from supplied patch (modified a bit).
4554 30. The use of :fail: followed by an empty string could lead Exim to respond to
4555 sender verification failures with (e.g.):
4557 550 Verification failed for <xxx>
4558 550 Sender verify failed
4560 where the first response line was missing the '-' that indicates it is not
4561 the final line of the response.
4563 31. The loop for finding the name of the user that called Exim had a hardwired
4564 limit of 10; it now uses the value of finduser_retries, which is used for
4565 all other user lookups.
4567 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4570 33. Exim was neglecting to zero errno before one call of strtol() when
4571 expanding a string and expecting an integer value. On some systems this
4572 resulted in spurious "integer overflow" errors. Also, it was casting the
4573 result into an int without checking.
4575 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4576 did not work. The code looks as if it has *never* worked, though it appears
4577 to have been documented since at least releast 1.62. I have made it work.
4579 35. The "timeout_DNS" error in retry rules, also documented since at least
4580 1.62, also never worked. As it isn't clear exactly what this means, and
4581 clearly it isn't a major issue, I have abolished the feature by treating it
4582 as "timeout", and writing a warning to the main and panic logs.
4584 36. The display of retry rules for -brt wasn't always showing the error code
4587 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4588 timeout_connect_A, timeout_connect_MX.
4590 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4591 to the empty sender.
4593 39. The daemon was not analysing the content of -oX till after it had closed
4594 stderr and disconnected from the controlling terminal. This meant that any
4595 syntax errors were only noted on the panic log, and the return code from
4596 the command was 0. By re-arranging the code a little, I've made the
4597 decoding happen first, so such errors now appear on stderr, and the return
4598 code is 1. However, the actual setting up of the sockets still happens in
4599 the disconnected process, so errors there are still only recorded on the
4602 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4603 connections (as happens on some IP stacks) was logged at start up time as
4604 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4605 it from "IPv6 and IPv4", which means that two separate sockets are being
4608 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4609 says whether AF_INET or AF_INET6 was passed as an argument.
4611 42. Exiwhat output was messed up when time zones were included in log
4614 43. Exiwhat now gives more information about the daemon's listening ports,
4615 and whether -tls-on-connect was used.
4617 44. The "port" option of the smtp transport is now expanded.
4619 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4620 silently ignored. Now an error message is written to the main and panic
4623 46. There's a new ACL modifier called "logwrite" which writes to a log file
4624 as soon as it is encountered.
4626 47. Added $local_user_uid and $local_user_gid at routing time.
4628 48. Exim crashed when trying to verify a sender address that was being
4631 49. Exim was recognizing only a space character after ".include". It now also
4632 recognizes a tab character.
4634 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4635 extracting the relevant information from the specification. The man page no
4636 longer contains scrambled data for the -d option, and I've added a section
4637 at the front about calling Exim under different names.
4639 51. Added "extra_headers" argument to the "mail" command in filter files.
4641 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4644 53. Installed eximstats 1.29.
4646 54. Added transport_filter_timeout as a generic transport option.
4648 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4649 Cc: header lines. This was required by RFC 822, but it not required by RFC
4652 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4653 incoming messages that did not have them. Now it does so only if the
4654 message originates locally, that is, if there is no associated remote host
4655 address. When Resent- header lines are present, this applies to the Resent-
4656 lines rather than the non-Resent- lines.
4658 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4659 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4661 58. Messages for configuration errors now include the name of the main
4662 configuration file - useful now that there may be more than one file in a
4663 list (.included file names were always shown).
4665 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4666 for those rare installations that do not start the daemon as root or run it
4667 setuid root. I've cut out the call to initgroups() if the daemon is not
4670 60. The Exim user and group can now be bound into the binary as text strings
4671 that are looked up at the start of Exim's processing.
4673 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4675 62. Added $mailstore_basename variable.
4677 63. Installed patch to sieve.c from Michael Haardt.
4679 64. When Exim failed to open the panic log after failing to open the main log,
4680 the original message it was trying to log was written to stderr and debug
4681 output, but if they were not available (the usual case in production), it
4682 was lost. Now it is written to syslog before the two lines that record the
4683 failures to open the logs.
4685 65. Users' Exim filters run in subprocesses under the user's uid. It is
4686 possible for a "deliver" command or an alias in a "personal" command to
4687 provoke an address rewrite. If logging of address rewriting is configured,
4688 this fails because the process is not running as root or exim. There may be
4689 a better way of dealing with this, but for the moment (because 4.30 needs
4690 to be released), I have disabled address rewrite logging when running a
4691 filter in a non-root, non-exim process.
4697 1. The buildconfig auxiliary program wasn't quoting the value set for
4698 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4699 not defined. This bug was masked in 4.22 by the effect that was fixed in
4702 2. Some messages that were rejected after a message id was allocated were
4703 shown as "incomplete" by exigrep. It no longer does this for messages that
4704 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4706 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4707 have allow_domain_literals set, the ID did not get logged in the <= line.
4708 Domain literals are now always recognized in Message-ID: header lines.
4710 4. The first argument for a ${extract expansion item is the key name or field
4711 number. Leading and trailing spaces in this item were not being ignored,
4712 causing some misleading effects.
4714 5. When deliver_drop_privilege was set, single queue runner processes started
4715 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4716 same command in the process it spins off) were not dropping privilege.
4718 6. When the daemon running as "exim" started a queue runner, it always
4719 re-executed Exim in the spun-off process. This is a waste of effort when
4720 deliver_drop_privilege is set. The new process now just calls the
4721 queue-runner function directly.
4727 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4730 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4731 ignored. Though the use of -oP was forcing the writing of a pid file, it
4732 was always written to the default place.
4734 3. If the message "no IP address found for host xxxx" is generated during
4735 incoming verification, it is now followed by identification of the incoming
4736 connection (so you can more easily find what provoked it).
4738 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4740 5. Added some features to "harden" Exim a bit more against certain attacks:
4742 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4743 be put in Local/Makefile. This is like the never_users runtime option,
4744 but it cannot be overridden. The default setting is "root".
4746 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4747 prefix string with which any file named in a -C command line option
4750 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4751 is retained for -C and -D only if the caller of Exim is root. Without
4752 it, the exim user may also use -C and -D and retain privilege.
4754 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4755 command line option is disabled.
4757 6. Macro names set by the -D option must start with an upper case letter, just
4758 like macro names defined in the configuration file.
4760 7. Added "dereference=" facility to LDAP.
4762 8. Two instances of the typo "uknown" in the source files are fixed.
4764 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4765 the Configure-Makefile script screwed up while processing it.
4767 10. Incorporated PCRE 4.4.
4769 11. The SMTP synchronization check was not operating right at the start of an
4770 SMTP session. For example, it could not catch a HELO sent before the client
4771 waited for the greeting. There is now a check for outstanding input at the
4772 point when the greeting is written. Because of the duplex, asynchronous
4773 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4774 way, but not yet received, when the check is performed.
4776 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4777 on TCP/IP sockets, because this apparently causes some broken clients to
4780 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4781 unchanged) from the Cygwin maintainer.
4783 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4784 was supported instead of testing for mbx. Effectively a typo.
4786 15. The spa authenticator server code was not checking that the input it
4787 received was valid base64.
4789 16. The debug output line for the "set" modifier in ACLs was not showing the
4790 name of the variable that was being set.
4792 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4794 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4795 lookup on its own. Something else was needed to trigger the lookup. For
4796 example, a match in host_lookup or the need for a host name in a host list.
4797 Now, if $sender_host_name is referenced and the host name has not yet been
4798 looked up, a lookup is performed. If the lookup fails, the variable remains
4799 empty, and $host_lookup_failed is set to "1".
4801 19. Added "eqi" as a case-independent comparison operator.
4803 20. The saslauthd authentication condition could segfault if neither service
4804 nor realm was specified.
4806 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4807 error message that was logged was misleading, and incoming SMTP
4808 connections were dropped. The message is now more accurate, and temporary
4809 errors are given to SMTP connections.
4811 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4812 (except RSET) with a 421 error, until QUIT is received. However, it was
4813 failing to send a response to QUIT.
4815 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4816 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4817 ACL is now run after the helo_try_verify_hosts code.
4819 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4820 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4821 case-independent, so other case variants are also recognized). Apparently
4822 some systems use these upper case variants.
4824 25. If more than two messages were waiting for the same host, and a transport
4825 filter was specified for the transport, Exim sent two messages over the
4826 same TCP/IP connection, and then failed with "socket operation on non-
4827 socket" when it tried to send the third.
4829 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4831 27. The extern definition of crypt16() in expand.c was not being excluded when
4832 the OS had its own crypt16() function.
4834 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4835 as a preferred synonym for return_size_limit, both as an option and as an
4838 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4840 30. Changed the default configuration ACL to relax the local part checking rule
4841 for addresses that are not in any local domains. For these addresses,
4842 slashes and pipe symbols are allowed within local parts, but the sequence
4843 /../ is explicitly forbidden.
4845 31. SPA server authentication was not clearing the challenge buffer before
4848 32. log_message in a "warn" ACL statement was writing to the reject log as
4849 well as to the main log, which contradicts the documentation and doesn't
4850 seem right (because no rejection is happening). So I have stopped it.
4852 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4853 However, I am unable to do any testing of this.
4855 34. Fixed an infelicity in the appendfile transport. When checking directories
4856 for a mailbox, to see if any needed to be created, it was accidentally
4857 using path names with one or more superfluous leading slashes; tracing
4858 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4860 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4861 contents are added to the log line that is written for every discarded
4862 recipient. (Previously a log_message setting was ignored.)
4864 36. The ${quote: operator now quotes the string if it is empty.
4866 37. The install script runs exim in order to find its version number. If for
4867 some reason other than non-existence or emptiness, which it checks, it
4868 could not run './exim', it was installing it with an empty version number,
4869 i.e. as "exim-". This error state is now caught, and the installation is
4872 38. An argument was missing from the function that creates an error message
4873 when Exim fails to connect to the socket for saslauthd authentication.
4874 This could cause Exim to crash, or give a corrupted message.
4876 39. Added isip, isip4, and isip6 to ${if conditions.
4878 40. The ACL variables $acl_xx are now saved with the message, and can be
4879 accessed later in routers, transports, and filters.
4881 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4882 strings in the file are not string-expanded.
4884 42. If a MAIL command specified a SIZE value that was too large to fit into an
4885 int variable, the check against message_size_limit failed. Such values are
4886 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4887 day this will have to be increased, but I don't think I want to be around
4888 when emails are that large.
4895 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4896 iconv() is not standard in FreeBSD.
4898 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4899 IPv6 enabled. The observed symptom was a segmentation fault on return from
4900 the function os_common_find_running_interfaces() in src/os.c.
4902 3. In the check_special_case() function in daemon.c I had used "errno" as an
4903 argument name, which causes warnings on some systems. This was basically a
4904 typo, since it was named "eno" in the comments!
4906 4. The code that waits for the clock to tick (at a resolution of some fraction
4907 of a second) so as to ensure message-id uniqueness was always waiting for
4908 at least one whole tick, when it could have waited for less. [This is
4909 almost certainly not relevant at current processor speeds, where it is
4910 unlikely to ever wait at all. But we try to future-proof.]
4912 5. The function that sleeps for a time interval that includes fractions of a
4913 second contained a race. It did not block SIGALRM between setting the
4914 timer, and suspending (a couple of lines later). If the interval was short
4915 and the sigsuspend() was delayed until after it had expired, the suspension
4916 never ended. On busy systems this could lead to processes getting stuck for
4919 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4920 process, before it forks any delivery processes. The open lookup caching
4921 mechanism meant that the open file or database connection was passed into
4922 the delivery process. The problem was that delivery processes always tidy
4923 up cached lookup data. This could cause a problem for the next delivery
4924 process started by the queue runner, because the external queue runner
4925 process does not know about the closure. So the next delivery process
4926 still has data in the lookup cache. In the case of a file lookup, there was
4927 no problem because closing a file descriptor in a subprocess doesn't affect
4928 the parent. However, if the lookup was caching a connection to a database,
4929 the connection was closed, and the second delivery process was likely to
4930 see errors such as "PGSQL: query failed: server closed the connection
4931 unexpectedly". The problem has been fixed by closing all cached lookups
4932 in a queue runner before running a delivery process.
4934 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4935 seem to have the "const" qualifier which it has on other OS. I've
4938 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4939 *of the same type* (client or server) with the same public name that an
4940 error should be diagnosed.
4942 9. When Exim looked up a host name for an IP address, but failed to find the
4943 original IP address when looking up the host name (a safety check), it
4944 output the message "<ip address> does not match any IP for NULL", which was
4945 confusing, to say the least. The bug was that the host name should have
4946 appeared instead of "NULL".
4948 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4949 user that is built into the binary, and the -C or -D options is used, root
4950 privilege is dropped before the configuration file is read. In addition,
4951 logging is switched to stderr instead of the normal log files. If the
4952 configuration then re-defines the Exim user, the unprivileged environment
4953 is probably not what is expected, so Exim logs a panic warning message (but
4956 However, if deliver_drop_privilege is set, the unprivileged state may well
4957 be exactly what is intended, so the warning has been cut out in that case,
4958 and Exim is allowed to try to write to its normal log files.
4964 1. smtp_return_error_details was not giving details for temporary sender
4965 or receiver verification errors.
4967 2. Diagnose a configuration error if two authenticators have the same public
4970 3. Exim used not to create the message log file for a message until the first
4971 delivery attempt. This could be confusing when incoming messages were held
4972 for policy or load reasons. The message log file is now created at the time
4973 the message is received, and an initial "Received" line is written to it.
4975 4. The automatically generated man page for command line options had a minor
4976 bug that caused no ill effects; however, a more serious problem was that
4977 the procedure for building the man page automatically didn't always
4978 operate. Consequently, release 4.20 contains an out-of-date version. This
4979 shouldn't happen again.
4981 5. When building Exim with embedded Perl support, the script that builds the
4982 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4983 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4985 6. The freeze_tell option was not being used for messages that were frozen on
4986 arrival, either by an ACL or by local_scan().
4988 7. Added the smtp_incomplete_transaction log selector.
4990 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4991 was accepting AUTH without a new EHLO.
4993 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4994 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4997 10. Logging of TCP/IP connections (when configured) now happens in the main
4998 daemon process instead of the child process, so that the TCP/IP connection
4999 count is more accurate (but it can never be perfect).
5001 11. The use of "drop" in a nested ACL was not being handled correctly in the
5002 outer ACL. Now, if condition failure induced by the nested "drop" causes
5003 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5004 or "require"), the connection is dropped.
5006 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5007 that yield "discard" can now be used with an "accept" or a "discard" verb,
5008 but an error is generated for any others (because I can't see a useful way
5009 to define what should happen).
5011 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5012 are now processed in the same way as lines in the Exim configuration file.
5013 In particular, continuation lines are supported.
5015 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5017 15. Added -ti meaning -t -i.
5019 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5020 domains, and warn by logging if others are found.
5022 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5023 returned from GIFCONF when Exim is trying to find the list of interfaces on
5024 a host. The code in os.c has been modified to copy each ifreq to an aligned
5025 structure in all cases.
5027 Also, in some cases, the returned ifreq's were being copied to a 'struct
5028 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5029 means the last couple of bytes of an IPv6 address could be chopped if the
5030 ifreq contained only a normal sockaddr (14 bytes storage).
5032 18. Named domain lists were not supported in the hosts_treat_as_local option.
5033 An entry such as +xxxx was not recognized, and was treated as a literal
5036 19. Ensure that header lines added by a DATA ACL are included in the reject log
5037 if the ACL subsequently rejects the message.
5039 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5040 MD5 (which is deprecated).
5042 21. When testing a filter file using -bf, Exim was writing a message when it
5043 took the sender from a "From " line in the message, but it was not doing so
5044 when it took $return_path from a Return-Path: header line. It now does.
5046 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5047 with a valid header line field name (a series of printing characters
5048 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5050 23. Changed "disc" in the source to "disk" to conform to the documentation and
5051 the book and for uniformity.
5053 24. Ignore Sendmail's -Ooption=value command line item.
5055 25. When execve() failed while trying to run a command in a pipe transport,
5056 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5057 could be confused with a return value of 69 from the command itself. This
5058 has been changed to 127, the value the shell returns if it is asked to run
5059 a non-existent command. The wording for the related log line suggests a
5060 non-existent command as the problem.
5062 26. If received_header_text expands to an empty string, do not add a Received:
5063 header line to the message. (Well, it adds a token one on the spool, but
5064 marks it "old" so that it doesn't get used or transmitted.)
5066 27. Installed eximstats 1.28 (addition of -nt option).
5068 28. There was no check for failure on the call to getsockname() in the daemon
5069 code. This can fail if there is a shortage of resources on the system, with
5070 ENOMEM, for example. A temporary error is now given on failure.
5072 29. Contrary to the C standard, it seems that in some environments, the
5073 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5074 program. Exim now does this explicitly; it affects the formatting of
5075 timestamps using strftime().
5077 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5078 complaints. I've now initialized all the variables, to avoid this.
5080 32. Header lines added by a system filter were not being "seen" during
5081 transport-time rewrites.
5083 33. The info_callback() function passed to OpenSSL is set up with type void
5084 (*)(SSL *, int, int), as described somewhere. However, when calling the
5085 function (actually a macro) that sets it up, the type void(*)() is
5086 expected. I've put in a cast to prevent warnings from picky compilers.
5088 34. If a DNS black list lookup found a CNAME record, but there were no A
5089 records associated with the domain it pointed at, Exim crashed.
5091 35. If a DNS black list lookup returned more than one A record, Exim ignored
5092 all but the first. It now scans all returned addresses if a particular IP
5093 value is being sought. In this situation, the contents of the
5094 $dnslist_value variable are a list of all the addresses, separated by a
5097 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5098 to accept a host name and all its aliases if the forward lookup for any of
5099 them yielded the IP address of the incoming connection. Now it accepts only
5100 those names whose forward lookup yields the correct IP address. Any other
5101 names are discarded. This closes a loophole whereby a rogue DNS
5102 administrator could create reverse DNS records to break through a
5103 wildcarded host restriction in an ACL.
5105 37. If a user filter or a system filter that ran in a subprocess used any of
5106 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5107 the wrong values were passed to the pipe command ($thisaddress had the
5108 value of $0, $0 had the value of $1, etc). This bug was introduced by
5109 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5111 38. Improved the line breaking for long SMTP error messages from ACLs.
5112 Previously, if there was no break point between 40 and 75 characters, Exim
5113 left the rest of the message alone. Two changes have been made: (a) I've
5114 reduced the minimum length to 35 characters; (b) if it can't find a break
5115 point between 35 and 75 characters, it looks ahead and uses the first one
5116 that it finds. This may give the occasional overlong line, but at least the
5117 remaining text gets split now.
5119 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5120 file descriptors might be low, and that setting 1000 would always raise it.
5121 It turns out that in some environments, the limit is already over 1000 and
5122 that lowering it causes trouble. So now Exim takes care not to decrease it.
5124 40. When delivering a message, the value of $return_path is set to $sender_
5125 address at the start of routing (routers may change the value). By an
5126 oversight, this default was not being set up when an address was tested by
5127 -bt or -bv, which affected the outcome if any router or filter referred to
5130 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5131 log only when "message" or "log_message" are set. However, if one of the
5132 conditions was an address verification, or a call to a nested ACL, the
5133 messages generated by the underlying test were being passed through. This
5134 no longer happens. The underlying message is available in $acl_verify_
5135 message for both "message" and "log_message" expansions, so it can be
5136 passed through if needed.
5138 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5139 new expansion $bh_ to give the encoded byte string without charset
5140 translation. Translation happens only if iconv() is available; HAVE_ICONV
5141 indicates this at build time. HEADERS_CHARSET gives the charset to
5142 translate to; headers_charset can change it in the configuration, and
5143 "headers charset" can change it in an individual filter file.
5145 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5146 that creates RFC 2047 encoded "words" labels them as that charset instead
5147 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5148 expansion operator; (ii) when Exim creates a From: line for a local
5149 message; (iii) when a header line is rewritten to include a "phrase" part.
5151 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5152 buggy, causing it to skip the first lines of messages whose message ID
5153 ended in 'D'. This would not have bitten before Exim release 4.14, because
5154 message IDs were unlikely to end in 'D' before then. The effect was to have
5155 incorrect size information for certain domains.
5157 45. #include "config.h" was missing at the start of the crypt16.c module. This
5158 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5161 46. If there was a timeout during a "random" callout check, Exim treated it as
5162 a failure of the random address, and carried on sending RSET and the real
5163 address. If the delay was just some slowness somewhere, the response to the
5164 original RCPT would be taken as a response to RSET and so on, causing
5165 mayhem of various kinds.
5167 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5168 when I implemented it. It didn't allow for the fact that some option values
5169 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5170 the right test for positive values.
5172 48. Domain names in DNS records are case-independent. Exim always looks them up
5173 in lower case. Some resolvers return domain names in exactly the case they
5174 appear in the zone file, that is, they may contain uppercase letters. Not
5175 all resolvers do this - some return always lower case. Exim was treating a
5176 change of case by a resolver as a change of domain, similar to a widening
5177 of a domain abbreviation. This triggered its re-routing code and so it was
5178 trying to route what was effectively the same domain again. This normally
5179 caused routing to fail (because the router wouldn't handle the domain
5180 twice). Now Exim checks for this case specially, and just changes the
5181 casing of the domain that it ultimately uses when it transmits the message
5184 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5187 50. If a filter generated a file delivery with a non-absolute name (possible if
5188 no home directory exists for the router), the forbid_file option was not
5191 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5192 the existing equality matching.
5194 52. Exim was using ints instead of ino_t variables in some places where it was
5195 dealing with inode numbers.
5197 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5198 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5199 called TMPDIR, and if it finds it is different, it changes its value.
5201 54. The smtp_printf() function is now made available to local_scan() so
5202 additional output lines can be written before returning. There is also an
5203 smtp_fflush() function to enable the detection of a dropped connection.
5204 The variables smtp_input and smtp_batched_input are exported to
5207 55. Changed the default runtime configuration: the message "Unknown user"
5208 has been removed from the ACL, and instead placed on the localuser router,
5209 using the cannot_route_message feature. This means that any verification
5210 failures that generate their own messages won't get overridden. Similarly,
5211 the "Unrouteable address" message that was in the ACL for unverifiable
5212 relay addresses has also been removed.
5214 56. Added hosts_avoid_esmtp to the smtp transport.
5216 57. The exicyclog script was not checking for the esoteric option
5217 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5218 will work only if exicyclog is run under the appropriate euid.
5220 58. Following a discussion on the list, the rules by which Exim recognises line
5221 endings on incoming messages have been changed. The -dropcr and drop_cr
5222 options are now no-ops, retained only for backwards compatibility. The
5223 following line terminators are recognized: LF CRLF CR. However, special
5224 processing applies to CR:
5226 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5227 nor a local message in the state where . is a terminator.
5229 (ii) If a bare CR is encountered in a header line, an extra space is added
5230 after the line terminator so as not to end the header. The reasoning
5231 behind this is that bare CRs in header lines are most likely either
5232 to be mistakes, or people trying to play silly games.
5234 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5235 was being incorrectly given as 18 bytes larger than it should have been.
5236 This is a VOB (very old bug).
5238 60. This may never have affected anything current, but just in case it has:
5239 When the local host is found other than at the start of a list of hosts,
5240 the local host, those with the same MX, and any that follow, are discarded.
5241 When the list in question was part of a longer list of hosts, the following
5242 hosts (not currently being processed) were also being discarded. This no
5243 longer happens. I'm not sure if this situation could ever has previously
5246 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5249 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5250 header. This is something that is recommended in a new Internet Draft, and
5251 is something that is documented as being done by Sendmail. There are two
5252 possible values. For messages generated by the autoreply transport, Exim
5255 Auto-Submitted: auto-replied
5257 whereas for all other generated messages (e.g. bounces) it adds
5259 Auto-Submitted: auto-generated
5261 63. The "personal" condition in filters now includes a test for the
5262 Auto-Submitted: header. If it contains the string "auto-" the message it
5263 not considered personal.
5265 64. Added rcpt_include_affixes as a generic transport option.
5267 65. Added queue_only_override (default true).
5269 66. Added the syslog_duplication option.
5271 67. If what should have been the first header line of a message consisted of
5272 a space followed by a colon, Exim was mis-interpreting it as a header line.
5273 It isn't of course - it is syntactically invalid and should therefore be
5274 treated as the start of the message body. The misbehaviour could have
5275 caused a number of strange effects, including loss of data in subsequent
5276 header lines, and spool format errors.
5278 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5279 client host had authenticated. This control can now be exercised by an ACL
5280 for more flexibility.
5282 69. By default, callouts do not happen when testing with -bh. There is now a
5283 variant, -bhc, which does actually run the callout code, including
5284 consulting and updating the callout cache.
5286 70. Added support for saslauthd authentication, courtesy of Alexander
5289 71. If statvfs() failed on the spool or log directories while checking their
5290 size for availability, Exim confusingly gave the error "space shortage".
5291 Furthermore, in debugging mode it crashed with a floating point exception.
5292 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5293 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5294 serious problem, Exim now writes to the main and panic logs when this
5295 happens, with details of the failure. It then refuses to accept the
5296 incoming message, giving the message "spool directory problem" or "log
5297 directory problem" with a 421 code for SMTP messages.
5299 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5300 0, 1, and 2 exist, because some OS complain for execs without them (see
5301 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5302 descriptors. However, the code omitted to check that the open succeeded,
5303 causing mysterious errors if for some reason the permissions on /dev/null
5304 got screwed. Now Exim writes a message to the main and panic logs, and
5305 bombs out if it can't open /dev/null.
5307 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5308 interact so that it is all more flexible. It is supposed to remain
5309 backwards compatible. Also added extra_local_interfaces.
5311 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5312 to bomb out with an assertion failure - to the client this appears as a
5313 connection drop. This problem occurs in the part of the code that was taken
5314 from the Samba project. Fortunately, the assertion is in a very simple
5315 function, so I have fixed this by reproducing the function inline in the
5316 one place where it is called, and arranging for authentication to fail
5317 instead of killing the process with assert().
5319 75. The SPA client code was not working when the server requested OEM rather
5320 than Unicode encoding.
5322 76. Added code to make require_files with a specific uid setting more usable in
5323 the case where statting the file as root fails - usually a non-root-mounted
5324 NFS file system. When this happens and the failure is EACCES, Exim now
5325 forks a subprocess and does the per-uid checking as the relevant uid.
5327 77. Added process_log_path.
5329 78. If log_file_path was not explicitly set, a setting of check_log_space or
5330 check_log_inodes was ignored.
5332 79. If a space check for the spool or log partitions fails, the incident is now
5333 logged. Of course, in the latter case the data may get lost...
5335 80. Added the %p formatting code to string_format() so that it can be used to
5336 print addresses in debug_print(). Adjusted all the address printing in the
5337 debugging in store.c to use %p rather than %d.
5339 81. There was a concern that a line of code in smtp_in.c could overflow a
5340 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5341 initially expressed, the concern was not well-founded, because trailing
5342 spaces are removed early. However, if the trailing spaces were followed by
5343 a NULL, they did not get removed, so the overflow was possible. Two fixes
5346 (a) I re-wrote the offending code in a cleaner fashion.
5347 (b) If an incoming SMTP command contains a NULL character, it is rejected
5350 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5351 runs initgroups(), so that if the Exim user is in any additional groups,
5352 they will be used during message reception.
5358 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob