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 Solaris build fix for Oracle's LDAP libraries.
10 Bugzilla 1109, patch from Stephen Usher.
12 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
18 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
20 PP/02 Harmonised TLS library version reporting; only show if debugging.
21 Layout now matches that introduced for other libraries in 4.74 PP/03.
23 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
25 PP/04 New "dns_use_edns0" global option.
27 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
30 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
31 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
33 TK/01 Updated PolarSSL code to 0.14.2.
34 Bugzilla 1097. Patch from Andreas Metzler.
36 PP/07 Catch divide-by-zero in ${eval:...}.
39 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
42 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
43 format-string attack -- SECURITY: remote arbitrary code execution.
45 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
46 time unintentionally subject to list matching rules, letting the header
47 cause arbitrary Exim lookups (of items which can occur in lists, *not*
48 arbitrary string expansion). This allowed for information disclosure.
50 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
51 INT_MIN/-1 -- value coerced to INT_MAX.
57 NM/01 Workround for PCRE version dependency in version reporting
60 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
61 This fixes portability to compilers other than gcc, notably
62 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
64 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
65 makefiles for portability to HP-UX and POSIX correctness.
67 PP/01 Permit LOOKUP_foo enabling on the make command-line.
68 Also via indented variable definition in the Makefile.
69 (Debugging by Oliver Heesakkers).
71 PP/02 Restore caching of spamd results with expanded spamd_address.
72 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
74 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
75 Improves build reliability. Fix from: Frank Elsner
77 NM/02 Fix wide character breakage in the rfc2047 coding
78 Fixes bug 1064. Patch from Andrey N. Oktyabrski
80 NM/03 Allow underscore in dnslist lookups
81 Fixes bug 1026. Patch from Graeme Fowler
83 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
84 Code patches from Adam Ciarcinski of NetBSD.
86 NM/04 Fixed exiqgrep to cope with mailq missing size issue
89 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
90 is logged, to avoid truncation. Patch from John Horne.
92 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
93 Patch from Jakob Hirsch.
95 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
96 SQL string expansion failure details.
97 Patch from Andrey Oktyabrski.
99 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
100 Patch from Simon Arlott.
102 PP/09 New lookups functionality failed to compile on old gcc which rejects
103 extern declarations in function scope.
104 Patch from Oliver Fleischmann
106 PP/10 Use sig_atomic_t for flags set from signal handlers.
107 Check getgroups() return and improve debugging.
108 Fixed developed for diagnosis in bug 927 (which turned out to be
111 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
112 Patch from Mark Zealey.
114 PP/12 Bugzilla 1056: Improved spamd server selection.
115 Patch from Mark Zealey.
117 PP/13 Bugzilla 1086: Deal with maildir quota file races.
118 Based on patch from Heiko Schlittermann.
120 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
121 Patch from Uwe Doering, sign-off by Michael Haardt.
123 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
124 variable declaration deep within a block. Bug and patch from
127 PP/15 lookups-Makefile IRIX compatibilty coercion.
129 PP/16 Make DISABLE_DKIM build knob functional.
131 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
132 Patch by Simon Arlott
134 TF/03 Fix valgrind.h portability to C89 compilers that do not support
135 variable argument macros. Our copy now differs from upstream.
141 TF/01 Failure to get a lock on a hints database can have serious
142 consequences so log it to the panic log.
144 TF/02 Log LMTP confirmation messages in the same way as SMTP,
145 controlled using the smtp_confirmation log selector.
147 TF/03 Include the error message when we fail to unlink a spool file.
149 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
150 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
151 for maintaining out-of-tree patches for some time.
153 PP/01 Bugzilla 139: Documentation and portability issues.
154 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
155 Handle per-OS dynamic-module compilation flags.
157 PP/02 Let /dev/null have normal permissions.
158 The 4.73 fixes were a little too stringent and complained about the
159 permissions on /dev/null. Exempt it from some checks.
160 Reported by Andreas M. Kirchwitz.
162 PP/03 Report version information for many libraries, including
163 Exim version information for dynamically loaded libraries. Created
164 version.h, now support a version extension string for distributors
165 who patch heavily. Dynamic module ABI change.
167 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
168 privilege escalation vulnerability whereby the Exim run-time user
169 can cause root to append content of the attacker's choosing to
172 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
175 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
176 If dropping privileges for untrusted macros, we disabled normal logging
177 on the basis that it would fail; for the Exim run-time user, this is not
178 the case, and it resulted in successful deliveries going unlogged.
179 Fixed. Reported by Andreas Metzler.
185 PP/01 Date: & Message-Id: revert to normally being appended to a message,
186 only prepend for the Resent-* case. Fixes regression introduced in
187 Exim 4.70 by NM/22 for Bugzilla 607.
189 PP/02 Include check_rfc2047_length in configure.default because we're seeing
190 increasing numbers of administrators be bitten by this.
192 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
194 PP/03 Bugzilla 994: added openssl_options main configuration option.
196 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
198 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
200 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
202 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
203 without a peer certificate, leading to a segfault because of an
204 assumption that peers always have certificates. Be a little more
205 paranoid. Problem reported by Martin Tscholak.
207 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
208 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
209 NB: ClamAV planning to remove STREAM in "middle of 2010".
210 CL also introduces -bmalware, various -d+acl logging additions and
211 more caution in buffer sizes.
213 PP/09 Implemented reverse_ip expansion operator.
215 PP/10 Bugzilla 937: provide a "debug" ACL control.
217 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
219 PP/12 Bugzilla 973: Implement --version.
221 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
223 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
225 PP/15 Bugzilla 816: support multiple condition rules on Routers.
227 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
228 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
229 ignore trailing whitespace.
231 JJ/02 prevent non-panic DKIM error from being sent to paniclog
233 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
236 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
237 Notification from Dr Andrew Aitchison.
239 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
240 ExtendedDetectionInfo response format.
241 Notification from John Horne.
243 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
246 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
247 XSL and documented dependency on system catalogs, with examples of how
250 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
253 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
254 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
255 configuration file which is writeable by the Exim user or group.
257 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
258 of configuration files to cover files specified with the -C option if
259 they are going to be used with root privileges, not just the default
262 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
263 option (effectively making it always true).
265 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
266 files to be used while preserving root privileges.
268 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
269 that rogue child processes cannot use them.
271 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
272 run-time user, instead of root.
274 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
275 Exim run-time user without dropping privileges.
277 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
278 result string, instead of calling string_vformat() twice with the same
281 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
282 for other users. Others should always drop root privileges if they use
283 -C on the command line, even for a whitelisted configure file.
285 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
287 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
293 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
294 $data_path, and $header_path variables; fixed documentation bugs and
297 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
298 exipick to access non-standard spools, including the "frozen" queue
301 NM/01 Bugzilla 965: Support mysql stored procedures.
302 Patch from Alain Williams
304 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
306 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
307 Patch from Andreas Metzler
309 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
310 Patch from Kirill Miazine
312 NM/05 Bugzilla 671: Added umask to procmail example.
314 JJ/03 installed exipick 20100323.0, fixing doc bug
316 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
317 directory. Notification and patch from Dan Rosenberg.
319 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
321 TK/02 Improve log output when DKIM signing operation fails.
323 MH/01 Treat the transport option dkim_domain as a colon separated
324 list, not as a single string, and sign the message with each element,
325 omitting multiple occurences of the same signer.
327 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
328 Bugzilla 985, 986. Patch by Simon Arlott
330 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
331 Patch by Simon Arlott
333 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
334 MBX locking. Notification from Dan Rosenberg.
340 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
342 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
344 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
346 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
348 NM/04 Bugzilla 915: Flag broken perl installation during build.
354 TK/01 Added patch by Johannes Berg that expands the main option
355 "spamd_address" if it starts with a dollar sign.
357 TK/02 Write list of recipients to X-Envelope-Sender header when building
358 the mbox-format spool file for content scanning (suggested by Jakob
361 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
362 (http://www.dcc-servers.net/) support via dccifd. Activated by
363 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
365 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
366 by Mark Daniel Reidel <mr@df.eu>.
368 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
369 When building exim an external PCRE library is now needed -
370 PCRE is a system library on the majority of modern systems.
371 See entry on PCRE_LIBS in EDITME file.
373 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
374 conversation. Added nologin parameter to request.
375 Patch contributed by Kirill Miazine.
377 TF/01 Do not log submission mode rewrites if they do not change the address.
379 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
381 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
382 log files in place. Contributed by Roberto Lima.
384 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
386 TF/03 Bugzilla 615: When checking the local_parts router precondition
387 after a local_part_suffix or local_part_prefix option, Exim now
388 does not use the address's named list lookup cache, since this
389 contains cached lookups for the whole local part.
391 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
392 Robert Millan. Documentation is in experimental-spec.txt.
394 TF/04 Bugzilla 668: Fix parallel build (make -j).
396 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
398 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
399 Patch provided by Jan Srzednicki.
401 TF/05 Leading white space used to be stripped from $spam_report which
402 wrecked the formatting. Now it is preserved.
404 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
405 that they are available at delivery time.
407 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
409 TF/08 TLS error reporting now respects the incoming_interface and
410 incoming_port log selectors.
412 TF/09 Produce a more useful error message if an SMTP transport's hosts
413 setting expands to an empty string.
415 NM/06 Bugzilla 744: EXPN did not work under TLS.
416 Patch provided by Phil Pennock.
418 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
419 Patch provided by Richard Godbee.
421 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
422 acl_smtp_notquit, added index entry.
424 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
425 Patch provided by Eugene Bujak.
427 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
428 accept(). Patch provided by Maxim Dounin.
430 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
431 Patch provided by Phil Pennock.
433 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
435 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
436 Patch provided by Brad "anomie" Jorsch.
438 NM/14 Bugzilla 622: Added timeout setting to transport filter.
439 Patch provided by Dean Brooks.
441 TK/05 Add native DKIM support (does not depend on external libraries).
443 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
444 Patch provided by Graeme Fowler.
446 NM/16 Bugzilla 851: Documentation example syntax fix.
448 NM/17 Changed NOTICE file to remove references to embedded PCRE.
450 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
453 NM/19 Bugzilla 745: TLS version reporting.
454 Patch provided by Phil Pennock.
456 NM/20 Bugzilla 167: bool: condition support.
457 Patch provided by Phil Pennock.
459 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
460 clients. Patch provided by Phil Pennock.
462 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
463 Patch provided by Brad "anomie" Jorsch.
465 NM/23 Bugzilla 687: Fix misparses in eximstats.
466 Patch provided by Heiko Schlittermann.
468 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
469 Patch provided by Heiko Schlittermann.
471 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
472 plus update to original patch.
474 NM/26 Bugzilla 799: Documentation correction for ratelimit.
476 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
477 Patch provided by David Brownlee.
479 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
481 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
483 NM/30 Bugzilla 888: TLS documentation bugfixes.
485 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
487 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
488 Unlike the original bugzilla I have changed all shell scripts in src tree.
490 NM/33 Bugzilla 898: Transport filter timeout fix.
491 Patch by Todd Rinaldo.
493 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
494 Patch by Serge Demonchaux.
496 NM/35 Bugzilla 39: Base64 decode bug fixes.
497 Patch by Jakob Hirsch.
499 NM/36 Bugzilla 909: Correct connect() call in dcc code.
501 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
503 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
505 NM/39 Bugzilla 911: Fixed MakeLinks build script.
511 TK/01 Add preliminary DKIM support. Currently requires a forked version of
512 ALT-N's libdkim that I have put here:
513 http://duncanthrax.net/exim-experimental/
515 Note to Michael Haardt: I had to rename some vars in sieve.c. They
516 were called 'true' and it seems that C99 defines that as a reserved
517 keyword to be used with 'bool' variable types. That means you could
518 not include C99-style headers which use bools without triggering
519 build errors in sieve.c.
521 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
522 as mailq or other aliases. Changed the --help handling significantly
523 to do whats expected. exim_usage() emits usage/help information.
525 SC/01 Added the -bylocaldomain option to eximstats.
527 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
529 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
531 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
537 PH/01 Another patch from the Sieve maintainer.
539 PH/02 When an IPv6 address is converted to a string for single-key lookup
540 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
541 dots are used instead of colons so that keys in lsearch files need not
542 contain colons. This was done some time before quoting was made available
543 in lsearch files. However, iplsearch files do require colons in IPv6 keys
544 (notated using the quote facility) so as to distinguish them from IPv4
545 keys. This meant that lookups for IP addresses in host lists did not work
546 for iplsearch lookups.
548 This has been fixed by arranging for IPv6 addresses to be expressed with
549 colons if the lookup type is iplsearch. This is not incompatible, because
550 previously such lookups could never work.
552 The situation is now rather anomolous, since one *can* have colons in
553 ordinary lsearch keys. However, making the change in all cases is
554 incompatible and would probably break a number of configurations.
556 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
559 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
560 conversion specification without a maximum field width, thereby enabling
561 a rogue spamd server to cause a buffer overflow. While nobody in their
562 right mind would setup Exim to query an untrusted spamd server, an
563 attacker that gains access to a server running spamd could potentially
564 exploit this vulnerability to run arbitrary code as the Exim user.
566 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
567 $primary_hostname instead of what libspf2 thinks the hosts name is.
569 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
570 a directory entry by the name of the lookup key. Previously, if a
571 symlink pointed to a non-existing file or a file in a directory that
572 Exim lacked permissions to read, a lookup for a key matching that
573 symlink would fail. Now it is enough that a matching directory entry
574 exists, symlink or not. (Bugzilla 503.)
576 PH/03 The body_linecount and body_zerocount variables are now exported in the
579 PH/04 Added the $dnslist_matched variable.
581 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
582 This means they are set thereafter only if the connection becomes
585 PH/06 Added the client_condition to authenticators so that some can be skipped
586 by clients under certain conditions.
588 PH/07 The error message for a badly-placed control=no_multiline_responses left
589 "_responses" off the end of the name.
591 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
593 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
594 (without spaces) instead of just copying the configuration text.
596 PH/10 Added the /noupdate option to the ratelimit ACL condition.
598 PH/11 Added $max_received_linelength.
600 PH/12 Added +ignore_defer and +include_defer to host lists.
602 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
603 way in which PCRE > 7.0 is built.
605 PH/14 Implemented queue_only_load_latch.
607 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
608 MAIL command. The effect was to mangle the value on 64-bit systems.
610 PH/16 Another patch from the Sieve maintainer.
612 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
614 PH/18 If a system quota error occurred while trying to create the file for
615 a maildir delivery, the message "Mailbox is full" was not appended to the
616 bounce if the delivery eventually timed out. Change 4.67/27 below applied
617 only to a quota excession during the actual writing of the file.
619 PH/19 It seems that peer DN values may contain newlines (and other non-printing
620 characters?) which causes problems in log lines. The DN values are now
621 passed through string_printing() before being added to log lines.
623 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
624 and InterBase are left for another time.)
626 PH/21 Added message_body_newlines option.
628 PH/22 Guard against possible overflow in moan_check_errorcopy().
630 PH/23 POSIX allows open() to be a macro; guard against that.
632 PH/24 If the recipient of an error message contained an @ in the local part
633 (suitably quoted, of course), incorrect values were put in $domain and
634 $local_part during the evaluation of errors_copy.
640 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
641 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
644 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
645 issue a MAIL command.
647 PH/02 In an ACL statement such as
649 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
651 if a client was not listed at all, or was listed with a value other than
652 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
653 the condition was not true (as it should be), so access was not denied.
654 The bug was that the ! inversion was incorrectly passed on to the second
655 item. This has been fixed.
657 PH/03 Added additional dnslists conditions == and =& which are different from
658 = and & when the dns lookup returns more than one IP address.
660 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
661 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
663 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
664 FSYNC, which compiles an option called disable_fsync that allows for
665 bypassing fsync(). The documentation is heavily laced with warnings.
667 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
669 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
670 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
671 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
672 including adding "make clean"; (3) Added -fPIC when compiling the test
673 dynamically loaded module, to get rid of a warning.
675 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
676 message fails, move_frozen_messages = true and ignore_bounce_errors_after
677 = 0s. The bug is otherwise harmless.
679 PH/07 There was a bug in the dovecot authenticator such that the value of
680 $auth1 could be overwritten, and so not correctly preserved, after a
681 successful authentication. This usually meant that the value preserved by
682 the server_setid option was incorrect.
684 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
686 PH/09 Installed PCRE release 7.0.
688 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
689 run for batched SMTP input. It is now run at the start of every message
690 in the batch. While fixing this I discovered that the process information
691 (output by running exiwhat) was not always getting set for -bs and -bS
692 input. This is fixed, and it now also says "batched" for BSMTP.
694 PH/11 Added control=no_pipelining.
696 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
697 patch, slightly modified), and move the expansion of helo_data till after
698 the connection is made in the smtp transport (so it can use these
701 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
703 PH/14 Added log_selector = +pid.
705 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
707 PH/16 Add ${if forany and ${if forall.
709 PH/17 Added dsn_from option to vary the From: line in DSNs.
711 PH/18 Flush SMTP output before performing a callout, unless control =
712 no_callout_flush is set.
714 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
715 was true (the default) a successful delivery failed to delete the retry
716 item, thus causing premature timeout of the address. The bug is now
719 PH/20 Added hosts_avoid_pipelining to the smtp transport.
721 PH/21 Long custom messages for fakedefer and fakereject are now split up
722 into multiline reponses in the same way that messages for "deny" and
723 other ACL rejections are.
725 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
726 with slight modification.
728 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
729 draft, changing the syntax and factoring some duplicate code".
731 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
732 for deliveries of the second and subsequent messages over the same SMTP
735 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
736 ${reduce, with only minor "tidies".
738 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
740 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
741 expansion side effects.
743 PH/27 When a message times out after an over-quota error from an Exim-imposed
744 quota, the bounce message says "mailbox is full". This message was not
745 being given when it was a system quota that was exceeded. It now should
748 MH/03 Made $recipients available in local_scan(). local_scan() already has
749 better access to the recipient list through recipients_list[], but
750 $recipients can be useful in postmaster-provided expansion strings.
752 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
753 in the case of a MAIL command with additional options following the
754 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
755 were accidentally chopped off.
757 PH/29 SMTP synchronization checks are implemented when a command is read -
758 there is a check that no more input is waiting when there shouldn't be
759 any. However, for some commands, a delay in an ACL can mean that it is
760 some time before the response is written. In this time, more input might
761 arrive, invalidly. So now there are extra checks after an ACL has run for
762 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
763 pipelining has not been advertised.
765 PH/30 MH's patch to allow iscntrl() characters to be list separators.
767 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
768 returned in the SMTP response when smtp_return_error_details was false.
771 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
772 instead of the C I/O that was originally supplied, because problems were
775 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
776 Exim which did not show up earlier: it was assuming that a call to
777 SSL_CTX_set_info_callback() might give an error value. In fact, there is
778 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
779 was a macro that became an assignment, so it seemed to work. This has
780 changed to a proper function call with a void return, hence the compile
781 error. Exim's code has been fixed.
783 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
786 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
788 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
790 PH/37 If a message is not accepted after it has had an id assigned (e.g.
791 because it turns out to be too big or there is a timeout) there is no
792 "Completed" line in the log. When some messages of this type were
793 selected by exigrep, they were listed as "not completed". Others were
794 picked up by some special patterns. I have improved the selection
795 criteria to be more general.
797 PH/38 The host_find_failed option in the manualroute router can now be set
798 to "ignore", to completely ignore a host whose IP address cannot be
799 found. If all hosts are ignored, the behaviour is controlled by the new
800 host_all_ignored option.
802 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
803 homing or because of multiple MX records with /mx) generated more than
804 one IP address, and the following item turned out to be the local host,
805 all the secondary addresses of the first item were incorrectly removed
806 from the list, along with the local host and any following hosts (which
807 is what is supposed to happen).
809 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
810 whoever called Exim into the -H file. In the case of the daemon it was
811 behaving confusingly. When first started, it used values for whoever
812 started the daemon, but after a SIGHUP it used the Exim user (because it
813 calls itself on a restart). I have changed the code so that it now always
816 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
817 message are rejected with the same error (e.g. no authentication or bad
818 sender address), and a DATA command is nevertheless sent (as can happen
819 with PIPELINING or a stupid MUA), the error message that was given to the
820 RCPT commands is included in the rejection of the DATA command. This is
821 intended to be helpful for MUAs that show only the final error to their
824 PH/42 Another patch from the Sieve maintainer.
826 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
827 Eximstats - Fixed some broken HTML links and added missing column headers
829 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
830 columns for Rejects, Temp Rejects, Ham, and Spam rows.
832 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
834 PH/43 Yet another patch from the Sieve maintainer.
836 PH/44 I found a way to check for a TCP/IP connection going away before sending
837 the response to the final '.' that terminates a message, but only in the
838 case where the client has not sent further data following the '.'
839 (unfortunately, this is allowed). However, in many cases there won't be
840 any further data because there won't be any more messages to send. A call
841 to select() can be used: if it shows that the input is "ready", there is
842 either input waiting, or the socket has been closed. An attempt to read
843 the next input character can distinguish the two cases. Previously, Exim
844 would have sent an OK response which the client would never have see.
845 This could lead to message repetition. This fix should cure that, at
846 least in a lot of common cases.
848 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
849 advertised in response to EHLO.
855 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
856 fixed by 4.65/MH/01 (is this a record?) are fixed:
858 (i) An empty string was always treated as zero by the numeric comparison
859 operators. This behaviour has been restored.
861 (ii) It is documented that the numeric comparison operators always treat
862 their arguments as decimal numbers. This was broken in that numbers
863 starting with 0 were being interpreted as octal.
865 While fixing these problems I realized that there was another issue that
866 hadn't been noticed. Values of message_size_limit (both the global option
867 and the transport option) were treated as octal if they started with 0.
868 The documentation was vague. These values are now always treated as
869 decimal, and I will make that clear in the documentation.
875 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
876 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
879 MH/01 Don't check that the operands of numeric comparison operators are
880 integers when their expansion is in "skipping" mode (fixes bug
881 introduced by 4.64-PH/07).
883 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
884 child addresses, Exim now panics and dies. Previously, because the count
885 is held in a short int, deliveries were likely to be lost. As such a
886 large number of recipients for a single message is ridiculous
887 (performance will be very, very poor), I have chosen to impose a limit
888 rather than extend the field.
894 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
895 leftover -K file (the existence of which was triggered by #402).
896 While we were at it, introduced process PID as part of the -K
897 filename. This should rule out race conditions when creating
900 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
901 processing considerably. Previous code took too long for large mails,
902 triggering a timeout which in turn triggers #401.
904 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
905 in the DK code in transports.c. sendfile() is not really portable,
906 hence the _LINUX specificness.
908 TF/01 In the add_headers option to the mail command in an Exim filter,
909 there was a bug that Exim would claim a syntax error in any
910 header after the first one which had an odd number of characters
913 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
914 callout verification, Exim cached a "reject" for the entire domain. This
915 is correct for most verifications, but it is not correct for a recipient
916 verification with use_sender or use_postmaster set, because in that case
917 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
918 case of MAIL FROM:<> rejection from other early rejections (e.g.
919 rejection of HELO). When verifying a recipient using a non-null MAIL
920 address, the cache is ignored if it shows MAIL FROM:<> rejection.
921 Whatever the result of the callout, the value of the domain cache is
922 left unchanged (for any other kind of callout, getting as far as trying
923 RCPT means that the domain itself is ok).
925 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
928 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
929 manifest itself as EPIPE rather than ECONNECT. When tidying away a
930 session, the daemon ignores ECONNECT errors and logs others; it now
931 ignores EPIPE as well.
933 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
934 (quoted-printable decoding).
936 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
937 later the small subsequent patch to fix an introduced bug.
939 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
941 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
943 PH/08 An error is now given if message_size_limit is specified negative.
945 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
946 to be given (somewhat) arbitrary names.
948 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
951 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
952 miscellaneous code fixes
954 PH/10 Added the log_reject_target ACL modifier to specify where to log
957 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
958 hostname. This is wrong, because it relates to the incoming message (and
959 probably the interface on which it is arriving) and not to the outgoing
960 callout (which could be using a different interface). This has been
961 changed to use the value of the helo_data option from the smtp transport
962 instead - this is what is used when a message is actually being sent. If
963 there is no remote transport (possible with a router that sets up host
964 addresses), $smtp_active_hostname is used.
966 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
967 tweaks were necessary in order to get it to work (see also 21 below):
968 (a) The code assumed that strncpy() returns a negative number on buffer
969 overflow, which isn't the case. Replaced with Exim's string_format()
971 (b) There were several signed/unsigned issues. I just did the minimum
972 hacking in of casts. There is scope for a larger refactoring.
973 (c) The code used strcasecmp() which is not a standard C function.
974 Replaced with Exim's strcmpic() function.
975 (d) The code set only $1; it now sets $auth1 as well.
976 (e) A simple test gave the error "authentication client didn't specify
977 service in request". It would seem that Dovecot has changed its
978 interface. Fortunately there's a specification; I followed it and
979 changed what the client sends and it appears to be working now.
981 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
984 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
985 address A is aliased to B and C, where B exists and C does not. Without
986 -v the output is "A verified" because verification stops after a
987 successful redirection if more than one address is generated. However,
988 with -v the child addresses are also verified. Exim was outputting "A
989 failed to verify" and then showing the successful verification for C,
990 with its parentage. It now outputs "B failed to verify", showing B's
991 parentage before showing the successful verification of C.
993 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
994 look up a TXT record in a specific list after matching in a combined
997 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
998 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
999 they consult the DNS. I had assumed they would set it the way they
1000 wanted; and indeed my experiments on Linux seem to show that in some
1001 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1002 To be on the safe side, however, I have now made the interface to
1003 host_find_byname() similar to host_find_bydns(), with an argument
1004 containing the DNS resolver options. The host_find_byname() function now
1005 sets these options at its start, just as host_find_bydns() does. The smtp
1006 transport options dns_qualify_single and dns_search_parents are passed to
1007 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1008 of host_find_byname() use the default settings of RES_DEFNAMES
1009 (qualify_single) but not RES_DNSRCH (search_parents).
1011 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1012 spool_read_header() do less string testing, by means of a preliminary
1013 switch on the second character of optional "-foo" lines. (This is
1014 overdue, caused by the large number of possibilities that now exist.
1015 Originally there were few.) While I was there, I also converted the
1016 str(n)cmp tests so they don't re-test the leading "-" and the first
1017 character, in the hope this might squeeze out yet more improvement.
1019 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1020 flag allowing group syntax was set by the header_syntax check but not
1021 turned off, possible causing trouble later; (2) The flag was not being
1022 set at all for the header_verify test, causing "group"-style headers to
1023 be rejected. I have now set it in this case, and also caused header_
1024 verify to ignore an empty address taken from a group. While doing this, I
1025 came across some other cases where the code for allowing group syntax
1026 while scanning a header line wasn't quite right (mostly, not resetting
1027 the flag correctly in the right place). These bugs could have caused
1028 trouble for malformed header lines. I hope it is now all correct.
1030 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1031 with the "reply" argument non-NULL. The code, however (which originally
1032 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1033 but it didn't always do it. This confused somebody who was copying the
1034 code for some other use. I have removed all the tests.
1036 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1037 feature that was used to support insecure browsers during the U.S. crypto
1038 embargo. It requires special client support, and Exim is probably the
1039 only MTA that supported it -- and would never use it because real RSA is
1040 always available. This code has been removed, because it had the bad
1041 effect of slowing Exim down by computing (never used) parameters for the
1042 RSA_EXPORT functionality.
1044 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1045 authenticator to fail if there's a tab character in the incoming data
1046 (there should never be unless someone is messing about, as it's supposed
1047 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1048 if the connection is using TLS or if the remote IP is the same as the
1049 local IP, and the "valid-client-cert option" if a client certificate has
1052 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1053 authenticators. This can be used for authorization after authentication
1054 succeeds. (In the case of plaintext, it servers for both authentication
1057 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1058 if any retry times were supplied.
1060 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1061 connection, where there is no client IP address to check. In this
1062 situation, the verify now always succeeds.
1064 PH/25 Applied John Jetmore's -Mset patch.
1066 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1068 PH/27 In a string expansion for a processed (not raw) header when multiple
1069 headers of the same name were present, leading whitespace was being
1070 removed from all of them, but trailing whitespace was being removed only
1071 from the last one. Now trailing whitespace is removed from each header
1072 before concatenation. Completely empty headers in a concatenation (as
1073 before) are ignored.
1075 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1076 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1078 PH/29 [Removed. This was a change that I later backed out, and forgot to
1079 correct the ChangeLog entry (that I had efficiently created) before
1080 committing the later change.]
1082 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1083 address errors (4xx response to RCPT) over the same connection as other
1084 messages routed to the same hosts. Such deliveries are always "forced",
1085 so retry times are not inspected. This resulted in far too many retries
1086 for the affected addresses. The effect occurred only when there were more
1087 hosts than the hosts_max_try setting in the smtp transport when it had
1088 the 4xx errors. Those hosts that it had tried were not added to the list
1089 of hosts for which the message was waiting, so if all were tried, there
1090 was no problem. Two fixes have been applied:
1092 (i) If there are any address or message errors in an SMTP delivery, none
1093 of the hosts (tried or untried) are now added to the list of hosts
1094 for which the message is waiting, so the message should not be a
1095 candidate for sending over the same connection that was used for a
1096 successful delivery of some other message. This seems entirely
1097 reasonable: after all the message is NOT "waiting for some host".
1098 This is so "obvious" that I'm not sure why it wasn't done
1099 previously. Hope I haven't missed anything, but it can't do any
1100 harm, as the worst effect is to miss an optimization.
1102 (ii) If, despite (i), such a delivery is accidentally attempted, the
1103 routing retry time is respected, so at least it doesn't keep
1104 hammering the server.
1106 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1107 in ${readsocket because some servers need this prod.
1109 PH/32 Added some extra debug output when updating a wait-xxx database.
1111 PH/33 The hint "could be header name not terminated by colon", which has been
1112 given for certain expansion errors for a long time, was not being given
1113 for the ${if def:h_colon_omitted{... case.
1115 PH/34 The spec says: "With one important exception, whenever a domain list is
1116 being scanned, $domain contains the subject domain." There was at least
1117 one case where this was not true.
1119 PH/35 The error "getsockname() failed: connection reset by peer" was being
1120 written to the panic log as well as the main log, but it isn't really
1121 panic-worthy as it just means the connection died rather early on. I have
1122 removed the panic log writing for the ECONNRESET error when getsockname()
1125 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1126 runs only) independently of the message's sender address. This meant
1127 that, if the 4xx error was in fact related to the sender, a different
1128 message to the same recipient with a different sender could confuse
1129 things. In particualar, this can happen when sending to a greylisting
1130 server, but other circumstances could also provoke similar problems.
1131 I have changed the default so that the retry time for these errors is now
1132 based a combination of the sender and recipient addresses. This change
1133 can be overridden by setting address_retry_include_sender=false in the
1136 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1137 remote server are returned as part of bounce messages. This was not
1138 happening for LMTP over a pipe (the lmtp transport), but now it is the
1139 same for both kinds of LMTP.
1141 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1142 in header lines that were in fact CNAMEs. This is no longer the case.
1144 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1145 and queue runs started by the daemon processed all messages. This has
1146 been fixed so that -R and -S can now usefully be given with -q<time>.
1148 PH/40 Import PCRE release 6.7 (fixes some bugs).
1150 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1152 PH/42 Give an error if -q is specified more than once.
1154 PH/43 Renamed the variables $interface_address and $interface_port as
1155 $received_ip_address and $received_port, to make it clear that these
1156 values apply to message reception, and not to the outgoing interface when
1157 a message is delivered. (The old names remain recognized, of course.)
1159 PH/44 There was no timeout on the connect() call when using a Unix domain
1160 socket in the ${readsocket expansion. There now is.
1162 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1163 be meaningful with "accept".
1165 SC/01 Eximstats V1.43
1166 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1168 SC/02 Eximstats V1.44
1169 Use a glob alias rather than an array ref in the generated
1170 parser. This improves both readability and performance.
1172 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1173 Collect SpamAssassin and rejection statistics.
1174 Don't display local sender or destination tables unless
1175 there is data to show.
1176 Added average volumes into the top table text output.
1178 SC/04 Eximstats V1.46
1179 Collect data on the number of addresses (recipients)
1180 as well as the number of messages.
1182 SC/05 Eximstats V1.47
1183 Added 'Message too big' to the list of mail rejection
1184 reasons (thanks to Marco Gaiarin).
1186 SC/06 Eximstats V1.48
1187 Mainlog lines which have GMT offsets and are too short to
1188 have a flag are now skipped.
1190 SC/07 Eximstats V1.49 (Alain Williams)
1191 Added the -emptyok flag.
1193 SC/08 Eximstats V1.50
1194 Fixes for obtaining the IP address from reject messages.
1196 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1197 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1198 whitesspace changes from 4.64-PH/27
1200 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1203 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1204 are found, allow negative numbers in numeric criteria)
1206 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1208 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1209 to match changes made in 4.64-PH/43
1211 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1213 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1214 "domain" is actually an IP address, and force a failure. This locks out
1215 those revolvers/nameservers that support "A-for-A" lookups, in
1216 contravention of the specifications.
1218 PH/48 When a host name was looked up from an IP address, and the subsequent
1219 forward lookup of the name timed out, the host name was left in
1220 $sender_host_name, contrary to the specification.
1222 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1223 restricted to single-key lookups, Exim was not diagnosing an error if
1224 * or *@ was used with a query-style lookup.
1226 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1228 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1229 long ago, but noone interested enough thought of it. Let's just say that
1230 the "1.1" means that there are some new functions that weren't there at
1231 some point in the past.
1233 PH/51 Error processing for expansion failure of helo_data from an smtp
1234 transport during callout processing was broken.
1236 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1237 tested/used via the -bh/-bhc/-bs options.
1239 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1240 bug, fixed in subsequent PCRE releases).
1242 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1243 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1245 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1251 SC/01 Use a glob alias rather than an array ref in eximstats generated
1252 parser. This improves both readability and performance.
1254 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1255 Don't display local sender or destination tables in eximstats unless
1256 there is data to show.
1257 Added average volumes into the eximstats top table text output.
1259 SC/03 Collect data on the number of addresses (recipients) as well
1260 as the number of messages in eximstats.
1262 TF/01 Correct an error in the documentation for the redirect router. Exim
1263 does (usually) call initgroups() when daemonizing.
1265 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1266 with consistent privilege compared to when running as a daemon.
1268 TF/03 Note in the spec that $authenticated_id is not set for local
1269 submissions from trusted users.
1271 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1272 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1274 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1275 by adding some example configuration directives to the default
1276 configuration file. A little bit of work is required to uncomment the
1277 directives and define how usernames and passwords are checked, but
1278 there is now a framework to start from.
1280 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1281 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1282 without this. I don't know how relevant this is to other LDAP libraries.
1284 PH/02 Add the verb name to the "unknown ACL verb" error.
1286 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1288 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1290 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1291 directory not expanded when it should be if an expanded home directory
1292 was set for the address (which is overridden by the transport).
1294 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1297 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1298 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1299 because it is too late at that time, and has no effect.
1301 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1302 security issue with \' (bugzilla #107). I could not use the
1303 PQescapeStringConn() function, because it needs a PGconn value as one of
1306 PH/08 When testing addresses using -bt, indicate those final addresses that
1307 are duplicates that would not cause an additional delivery. At least one
1308 person was confused, thinking that -bt output corresponded to deliveries.
1309 (Suppressing duplicates isn't a good idea as you lose the information
1310 about possibly different redirections that led to the duplicates.)
1312 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1313 systems where poll() doesn't work, in particular OS X.
1315 PH/10 Added more information to debugging output for retry time not reached.
1317 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1318 operations in malware.c.
1320 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1323 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1324 syslog_duplication set false, log lines that would normally be written
1325 both the the main log and to the reject log were not written to syslog at
1328 PH/14 In the default configuration, change the use of "message" in ACL warn
1329 statements to "add_header".
1331 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1332 not followed by a command (e.g. "seen endif").
1334 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1335 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1338 PH/17 Added extra conditions to the default value of delay_warning_condition
1342 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1343 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1344 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1347 The Auto-Submitted: and various List- headers are standardised, whereas I
1348 don't think Precedence: ever was.
1350 PH/18 Refactored debugging code in route_finduser() to show more information,
1351 in particular, the error code if getpwnam() issues one.
1353 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1354 This is apparently needed in addition to the PH/07 change above to avoid
1355 any possible encoding problems.
1357 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1358 but not after initializing Perl.
1360 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1361 output them only if debugging. By default they are written stderr,
1362 apparently, which is not desirable.
1364 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1367 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1370 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1372 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1373 authenticated or an ident call has been made. Suppress the default
1374 values for $authenticated_id and $authenticated_sender (but permit -oMai
1375 and -oMas) when testing with -bh.
1377 PH/24 Re-jigged the order of the tests in the default configuration so that the
1378 tests for valid domains and recipients precede the DNS black list and CSA
1379 tests, on the grounds that those ones are more expensive.
1381 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1382 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1383 This bug exists in every version of Exim that I still have, right back to
1386 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1387 However, an attempt to turn on case-sensitivity in a regex key by
1388 including (?-i) didn't work because the subject string was already
1389 lowercased, and the effects were non-intuitive. It turns out that a
1390 one-line patch can be used to allow (?-i) to work as expected.
1396 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1397 other effects) broke the use of negated acl sub-conditions.
1399 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1402 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1403 "Deny" causes Exim to reject the incoming connection with a 554 error.
1404 Unfortunately, if there is a major crisis, such as a disk failure,
1405 tcp-wrappers gives "deny", whereas what one would like would be some
1406 kind of temporary error. A kludge has been added to help with this.
1407 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1408 554 error is used if errno is still zero or contains ENOENT (which occurs
1409 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1412 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1414 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1415 errors. Otherwise a message that provokes a temporary error (when other
1416 messages do not) can cause a whole host to time out.
1418 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1419 addresses were routed directly to files or pipes from a redirect router.
1420 File deliveries just didn't batch; pipe deliveries might have suffered
1423 PH/06 A failure to get a lock for a hints database would erroneously always say
1424 "Failed to get write lock", even when it was really a read lock.
1426 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1427 of 0600. This has been changed to use the value of the lockfile_mode
1428 option (which defaults to 0600).
1430 PH/08 Applied small patch from the Sieve maintainer.
1432 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1433 folder from quota calculations, a direct delivery into this folder messed
1434 up the contents of the maildirsize file. This was because the regex was
1435 used only to exclude .Trash (or whatever) when the size of the mailbox
1436 was calculated. There was no check that a delivery was happening into an
1437 excluded directory. This bug has been fixed by ignoring all quota
1438 processing for deliveries into excluded directories.
1440 PH/10 Added the maildirfolder_create_regex option to appendfile.
1446 PH/01 The code for finding all the local interface addresses on a FreeBSD
1447 system running IPv6 was broken. This may well have applied to all BSD
1448 systems, as well as to others that have similar system calls. The broken
1449 code found IPv4 interfaces correctly, but gave incorrect values for the
1450 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1451 that it would not match correctly against @[] and not recognize the IPv6
1454 PH/02 The ipliteral router was not recognizing addresses of the form user@
1455 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1457 PH/03 Added disable_ipv6.
1459 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1460 decoded form, because it is most often used to construct To: headers
1461 lines in autoreplies, and the decoded form may well be syntactically
1462 invalid. However, $reply_address has leading white space removed, and all
1463 newlines turned into spaces so that the autoreply transport does not
1466 PH/05 If group was specified without a user on a router, and no group or user
1467 was specified on a transport, the group from the router was ignored.
1469 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1470 for visible compile-time settings that can be used to change these
1471 numbers, for those that want even more. Backwards compatibility with old
1472 spool files has been maintained. However, going back to a previous Exim
1473 release will lost any variables that are in spool files.
1475 PH/07 Two small changes when running in the test harness: increase delay when
1476 passing a TCP/IP connection to a new process, in case the original
1477 process has to generate a bounce, and remove special handling of
1478 127.0.0.2 (sic), which is no longer necessary.
1480 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1481 be the same on different OS.
1483 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1486 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1487 whether --show-vars was specified or not
1489 JJ/02 exipick: Added support for new ACL variable spool format introduced
1492 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1493 syntactically invalid From: or Reply-to: line, and a filter used this to
1494 generate an autoreply, and therefore failed to obtain an address for the
1495 autoreply, Exim could try to deliver to a non-existent relative file
1496 name, causing unrelated and misleading errors. What now happens is that
1497 it logs this as a hard delivery error, but does not attempt to create a
1500 PH/11 The exinext utility has a -C option for testing purposes, but although
1501 the given file was scanned by exinext itself; it wasn't being passed on
1502 when Exim was called.
1504 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1505 an end-of-file indication when reading a command response.
1507 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1508 compiled. In many other places in Exim, IPv6 addresses are always
1509 recognized, so I have changed this. It also means that IPv4 domain
1510 literals of the form [IPV4:n.n.n.n] are now always recognized.
1512 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1513 used if the router is not running as root, for example, when verifying at
1514 ACL time, or when using -bh. The debugging output from this situation was
1515 non-existent - all you got was a failure to exec. I have made two
1518 (a) Failures to set uid/gid, the current directory, or a process leader
1519 in a subprocess such as that created by queryprogram now generate
1520 suitable debugging ouput when -d is set.
1522 (b) The queryprogram router detects when it is not running as root,
1523 outputs suitable debugging information if -d is set, and then runs
1524 the subprocess without attempting to change uid/gid.
1526 PH/15 Minor change to Makefile for building test_host (undocumented testing
1529 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1530 additional section of a DNS packet that returns MX or SRV records.
1531 Instead, it always explicitly searches for A/AAAA records. This avoids
1532 major problems that occur when a DNS server includes only records of one
1533 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1534 fixed another bug: if SRV records were looked up and the corresponding
1535 address records were *not* found in the additional section, the port
1536 values from the SRV records were lost.
1538 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1539 using the correct key (the original address) when searching the retry
1540 rules in order to find which one to use for generating the retry hint.
1542 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1543 adding the default one. Similarly, if it contains a Reply-To: header, the
1544 errors_reply_to option, if set, is not used.
1546 PH/19 When calculating a retry time, Exim used to measure the "time since
1547 failure" by looking at the "first failed" field in the retry record. Now
1548 it does not use this if it is later than than the arrival time of the
1549 message. Instead it uses the arrival time. This makes for better
1550 behaviour in cases where some deliveries succeed, thus re-setting the
1551 "first failed" field. An example is a quota failure for a huge message
1552 when small messages continue to be delivered. Without this change, the
1553 "time since failure" will always be short, possible causing more frequent
1554 delivery attempts for the huge message than are intended.
1555 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1557 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1558 $1, $2, $3) because the numerical variables can be reset during some
1559 expansion items (e.g. "match"), thereby losing the authentication data.
1561 PH/21 Make -bV show the size of off_t variables so that the test suite can
1562 decide whether to run tests for quotas > 2G.
1564 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1565 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1566 filecount value is greater than 2G or if a quota value is greater than 2G
1567 on a system where the size of off_t is not greater than 4, a panic error
1570 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1571 never match. The debug and -bh output now contains an explicit error
1572 message indicating a malformed IPv4 address or mask.
1574 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1575 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1576 PH/23 above applies.
1578 PH/25 Do not write to syslog when running in the test harness. The only
1579 occasion when this arises is a failure to open the main or panic logs
1580 (for which there is an explicit test).
1582 PH/26 Added the /no_tell option to "control=freeze".
1584 PH/27 If a host name lookup failed very early in a connection, for example, if
1585 the IP address matched host_lookup and the reverse lookup yielded a name
1586 that did not have a forward lookup, an error message of the form "no IP
1587 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1588 could be logged. Now it outputs the IP address instead of "NULL".
1590 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1591 allows the sender and the authenticated sender to be set when
1592 submitting a message from within Exim. Since child_open_exim() is
1593 documented for local_scan(), the new function should be too.
1595 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1596 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1597 results in an empty string is now treated as unset.
1599 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1601 PH/31 Added sender_verify_fail logging option.
1603 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1604 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1605 tidied the source and removed it altogether.
1607 PH/33 When a queue run was abandoned because the load average was too high, a
1608 log line was always written; now it is written only if the queue_run log
1609 selector is set. In addition, the log line for abandonment now contains
1610 information about the queue run such as the pid. This is always present
1611 in "start" and "stop" lines but was omitted from the "abandon" line.
1613 PH/34 Omit spaces between a header name and the colon in the error message that
1614 is given when verify = headers_syntax fails (if there are lots of them,
1615 the message gets confusing).
1617 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1618 names, as there are now some PTR records that contain slashes. This check
1619 is only to protect against broken name servers that fall over on strange
1620 characters, so the fact that it applies to all lookups doesn't matter.
1622 PH/36 Now that the new test suite is complete, we can remove some of the
1623 special code in Exim that was needed for the old test suite. For example,
1624 sorting DNS records because real resolvers return them in an arbitrary
1625 order. The new test suite's fake resolver always returns records in the
1628 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1629 bounces) except when queue_only is set, to avoid logging races between
1630 the different processes.
1632 PH/38 Panic-die if .include specifies a non-absolute path.
1634 PH/39 A tweak to the "H" retry rule from its user.
1636 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1637 a label. They prevented compilation on older perls.
1639 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1640 a warning to be raised on newish perls.
1642 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1643 on queue. Changes to match documented behaviour of showing count of
1644 messages matching specified criteria.
1646 PH/40 Changed the default ident timeout from 30s to 5s.
1648 PH/41 Added support for the use of login_cap features, on those BSD systems
1649 that have them, for controlling the resources used by pipe deliveries.
1651 PH/42 The content-scanning code uses fopen() to create files in which to put
1652 message data. Previously it was not paying any attention to the mode of
1653 the files. Exim runs with umask(0) because the rest of the code creates
1654 files with open(), and sets the required mode explicitly. Thus, these
1655 files were ending up world-writeable. This was not a big issue, because,
1656 being within the spool directory, they were not world-accessible. I have
1657 created a function called modefopen, which takes an additional mode
1658 argument. It sets umask(777), creates the file, chmods it to the required
1659 mode, then resets the umask. All the relevant calls to fopen() in the
1660 content scanning code have been changed to use this function.
1662 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1663 to 24 hours. This avoids potential overflow problems when processing G
1664 and H retry rules. I suspect nobody ever tinkers with this value.
1666 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1668 PH/45 When the plaintext authenticator is running as a client, the server's
1669 challenges are checked to ensure they are valid base64 strings. By
1670 default, the authentication attempt is cancelled if an invalid string is
1671 received. Setting client_ignore_invalid_base64 true ignores these errors.
1672 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1673 they are received. Thus, the responses can be made to depend on the
1674 challenges. If an invalid string is ignored, an empty string is placed in
1677 PH/46 Messages that are created by the autoreply transport now contains a
1678 References: header, in accordance with RFCs 2822 and 3834.
1680 PH/47 Added authenticated_sender_force to the smtp transport.
1682 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1684 PH/49 Installed latest patch from the Sieve maintainer.
1686 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1687 was also set, the appendfile transport was unnecessarily scanning a
1688 directory of message files (e.g. for maildir delivery) to find the count
1689 of files (along with the size), even though it did not need this
1690 information. It now does the scan only if it needs to find either the
1691 size of the count of files.
1693 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1695 PH/52 Two bugs concerned with error handling when the smtp transport is
1698 (i) Exim was not creating retry information for temporary errors given
1699 for individual recipients after the DATA command when the smtp transport
1700 was used in LMTP mode. This meant that they could be retried too
1701 frequently, and not timed out correctly.
1703 (ii) Exim was setting the flag that allows error details to be returned
1704 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1705 recipients that were returned after the DATA command.
1707 PH/53 This is related to PH/52, but is more general: for any failing address,
1708 when detailed error information was permitted to be returned to the
1709 sender, but the error was temporary, then after the final timeout, only
1710 "retry timeout exceeded" was returned. Now it returns the full error as
1711 well as "retry timeout exceeded".
1713 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1714 do this, and (what is worse) MTAs that accept it.
1716 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1717 will now be deprecated.
1719 PH/56 New os.c-cygwin from the Cygwin maintainer.
1721 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1722 formats (previously only available in exim formats via -bpr, -bpru,
1723 and -bpra. Now also available in native and exiqgrep formats)
1725 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1726 with very large, slow to parse queues
1728 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1730 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1732 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1733 responses to authentication challenges, though it was showing the
1734 challenges; (ii) I've removed the CR characters from the debug output for
1737 PH/58 Allow for the insertion of a newline as well as a space when a string
1738 is turned into more than one encoded-word during RFC 2047 encoding. The
1739 Sieve code now uses this.
1741 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1742 data_4xx, lost_connection, tls_required.
1744 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1745 message was being sent as an SMTP response.
1747 PH/61 Add -l and -k options to exicyclog.
1749 PH/62 When verifying, if an address was redirected to one new address, so that
1750 verification continued, and the new address failed or deferred after
1751 having set something in $address_data, the value of $address_data was not
1752 passed back to the ACL. This was different to the case when no
1753 redirection occurred. The value is now passed back in both cases.
1755 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1756 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1757 use login_cap.h, so on its own it isn't the distinguishing feature. The
1758 new name refers directly to the setclassresources() function.
1760 PH/65 Added configuration files for NetBSD3.
1762 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1764 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1765 is preferred over IPv4.
1767 PH/68 The bounce_return_message and bounce_return_body options were not being
1768 honoured for bounces generated during the reception of non-SMTP messages.
1769 In particular, this applied to messages rejected by the ACL. This bug has
1770 been fixed. However, if bounce_return_message is true and bounce_return_
1771 body is false, the headers that are returned for a non-SMTP message
1772 include only those that have been read before the error was detected.
1773 (In the case of an ACL rejection, they have all been read.)
1775 PH/69 The HTML version of the specification is now built in a directory called
1776 spec_html instead of spec.html, because the latter looks like a path with
1777 a MIME-type, and this confuses some software.
1779 PH/70 Catch two compiler warnings in sieve.c.
1781 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1782 function verify_get_ident() calls ip_connect() to connect a socket, but
1783 if the "connect()" function timed out, ip_connect() used to close the
1784 socket. However, verify_get_ident() also closes the socket later, and in
1785 between Exim writes to the log, which may get opened at this point. When
1786 the socket was closed in ip_connect(), the log could get the same file
1787 descriptor number as the socket. This naturally causes chaos. The fix is
1788 not to close the socket in ip_connect(); the socket should be closed by
1789 the function that creates it. There was only one place in the code where
1790 this was missing, in the iplookup router, which I don't think anybody now
1791 uses, but I've fixed it anyway.
1793 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1794 well as to direct DNS lookups. Otherwise the handling of names in host
1795 lists is inconsistent and therefore confusing.
1801 PH/01 Two changes to the default runtime configuration:
1803 (1) Move the checks for relay_from_hosts and authenticated clients from
1804 after to before the (commented out) DNS black list checks.
1806 (2) Add control=submission to the relay_from_hosts and authenticated
1807 clients checks, on the grounds that messages accepted by these
1808 statements are most likely to be submissions.
1810 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1812 (1) Generate an error if the third argument for the ${prvs expansion is
1815 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1818 (3) Reset the variables that are obtained from the first argument of
1819 ${prvscheck and used in the second argument before leaving the code,
1820 because their memory is reclaimed, so using them afterwards may do
1823 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1824 one (it's much easier than Tom thought :-).
1826 (5) Because of (4), we can now allow for the use of $prvscheck_result
1827 inside the third argument.
1829 PH/03 For some reason, the default setting of PATH when running a command from
1830 a pipe transport was just "/usr/bin". I have changed it to
1833 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1834 anything to be listed in the output from -bV.
1836 PH/05 When a filter generated an autoreply, the entire To: header line was
1837 quoted in the delivery log line, like this:
1839 => >A.N.Other <ano@some.domain> <original@ddress> ...
1841 This has been changed so that it extracts the operative address. There
1842 may be more than one such address. If so, they are comma-separated, like
1845 => >ano@some.domain,ona@other.domain <original@ddress> ...
1847 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1848 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1849 not being looked up in the rDNS to get a host name, Exim was showing the
1850 IP address twice in Received: lines, even though the IP addresses were
1851 identical. For example:
1853 Received: from [1.2.3.4] (helo=[1.2.3.4])
1855 However, if the real host name was known, it was omitting the HELO data
1856 if it matched the actual IP address. This has been tidied up so that it
1857 doesn't show the same IP address twice.
1859 PH/07 When both +timestamp and +memory debugging was on, the value given by
1860 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1861 called by the debug printing, thereby overwriting the timestamp buffer.
1862 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1864 PH/08 When the original message was included in an autoreply transport, it
1865 always said "this is a copy of the message, including all the headers",
1866 even if body_only or headers_only was set. It now gives an appropriate
1869 PH/09 Applied a patch from the Sieve maintainer which:
1871 o fixes some comments
1872 o adds the (disabled) notify extension core
1873 o adds some debug output for the result of if/elsif tests
1874 o points to the current vacation draft in the documentation
1875 and documents the missing references header update
1879 o fixes a bug in processing the envelope test (when testing
1880 multiple envelope elements, the last element determinted the
1883 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1884 Electronic Mail") by including:
1886 Auto-submitted: auto-generated
1888 in the messages that it generates (bounce messages and others, such as
1889 warnings). In the case of bounce messages for non-SMTP mesages, there was
1890 also a typo: it was using "Auto_submitted" (underscore instead of
1891 hyphen). Since every message generated by Exim is necessarily in response
1892 to another message, thes have all been changed to:
1894 Auto-Submitted: auto-replied
1896 in accordance with these statements in the RFC:
1898 The auto-replied keyword:
1900 - SHOULD be used on messages sent in direct response to another
1901 message by an automatic process,
1903 - MUST NOT be used on manually-generated messages,
1905 - MAY be used on Delivery Status Notifications (DSNs) and Message
1906 Disposition Notifications (MDNs),
1908 - MUST NOT be used on messages generated by automatic or periodic
1909 processes, except for messages which are automatic responses to
1912 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1913 to the default Received: header definition.
1915 PH/12 Added log selector acl_warn_skipped (default on).
1917 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1918 variables because (a) they are in the wrong storage pool and (b) even if
1919 they were copied, it wouldn't work properly because of the caching.
1921 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1922 checking when decoding. Apparently there are clients that generate
1923 overlong encoded strings. Why am I not surprised?
1925 PH/15 If the first argument of "${if match_address" was not empty, but did not
1926 contain an "@" character, Exim crashed. Now it writes a panic log message
1927 and treats the condition as false.
1929 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1931 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1932 extension "envelope-auth". The code is finished and in agreement with
1933 other implementations, but there is no documentation so far and in fact,
1934 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1935 not changing the active code.
1937 Print executed "if" and "elsif" statements when debugging is used. This
1938 helps a great deal to understand what a filter does.
1940 Document more things not specified clearly in RFC3028. I had all this
1941 sorted out, when out of a sudden new issues came to my mind. Oops."
1943 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1946 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1947 canonical form (as documented). However, after a host name lookup from
1948 the IP address, check_host() was doing a simple string comparison with
1949 addresses acquired from the DNS when checking that the found name did
1950 have the original IP as one of its addresses. Since any found IPv6
1951 addresses are likely to be in abbreviated form, the comparison could
1952 fail. Luckily, there already exists a function for doing the comparison
1953 by converting both addresses to binary, so now that is used instead of
1954 the text comparison.
1956 PH/20 There was another similar case to PH/19, when a complete host name was
1957 given in a host list; looking up its IP address could give an abbreviated
1958 form, whereas the current host's name might or might not be abbreviated.
1959 The same fix has been applied.
1965 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1966 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1969 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1970 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1972 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1974 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1975 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1976 Also, for "Auto-Submitted", treat anything other than "no" as
1977 non-personal, in accordance with RFC 3834. (Previously it treated
1978 anything starting "auto-" as non-personal.)
1980 TF/01 The control=submission/name=... option had a problem with syntax
1981 errors if the name included a slash character. The /name= option
1982 now slurps the rest of the string, so it can include any characters
1983 but it must come last in the list of options (after /sender_retain
1986 PH/05 Some modifications to the interface to the fake nameserver for the new
1994 TK/01 Added the "success_on_redirect" address verification option. See
1995 NewStuff for rationale and an example.
1997 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1999 PH/02 Patch to exigrep to allow it to work on syslog lines.
2001 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2002 fread() to skip over the body file's header line, because in Cygwin the
2003 header line is locked and is inaccessible.
2005 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2006 co-exist for some time) to make it clear that it is the Exim ID that is
2007 referenced, not the Message-ID: header line.
2009 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2010 string_format() function, because snprintf() does not exist on all
2013 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2014 ${stat: expansion item.
2016 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2017 protocol synchronization error", to keep the pedants happy.
2019 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2020 well as for IRIX systems, when gcc is being used. See the host.c source
2023 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2025 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2028 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2029 transport and to the smtp transport in LMTP mode.
2031 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2033 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2034 run a filter in a subprocess. This could lead to confusion in subsequent
2035 lookups in the parent process. There should also be a search_tidyup() at
2036 the end of the subprocess.
2038 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2039 only if the host matched helo_try_verify_hosts, which caused the
2040 verification to occur when the EHLO/HELO command was issued. The ACL just
2041 tested the remembered result. Now, if a previous verification attempt has
2042 not happened, "verify = helo" does it there and then.
2044 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2046 TK/03 Fix log output including CR from clamd.
2048 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2049 exist provoked a memory error which could cause a segfault.
2051 PH/15 Installed PCRE 6.2
2053 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2055 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2056 of the problem. Specifically, suggested +O2 rather than +O1 for the
2059 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2061 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2062 record, the port was not correctly logged when the outgoing_port log
2063 selector was set (it logged the transort's default port).
2065 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2066 fallback_hosts, and "hosts" in the smtp transport.
2068 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2069 host errors such as "Connection refused".
2071 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2072 authentication with radiusclient 0.4.9:
2074 - Error returned from rc_read_config was caught wrongly
2075 - Username/password not passed on to radius server due to wrong length.
2077 The presumption is that some radiusclient API changes for 4.51/PH/17
2078 were not taken care of correctly. The code is still untested by me (my
2079 Linux distribution still has 0.3.2 of radiusclient), but it was
2080 contributed by a Radius user.
2082 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2083 expanding the "port" option of the smtp transport.
2085 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2086 while reading a MIME header. Thanks to Tom Hughes for a patch.
2088 PH/24 Include config.h inside local_scan.h so that configuration settings are
2091 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2092 that in an ACL for RCPT (for example), you can examine exactly what was
2095 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2096 commands, but it was not correctly comparing the address with the actual
2097 client host address. Thus, it would show the EHLO address in Received:
2098 header lines when this was not necessary.
2100 PH/27 Added the % operator to ${eval:}.
2102 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2103 it should be ignoring failures (because with -C, for example, it has lost
2104 privilege). It wasn't ignoring creation failures other than "already
2107 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2110 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2111 because an input line was too long, either on its own, or by virtue of
2112 too many continuations, the temporary file was not being removed, and the
2113 return code was incorrect.
2115 PH/31 Missing "BOOL" in function definition in filtertest.c.
2117 PH/32 Applied Sieve patches from the maintainer.
2119 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2121 PH/33 Added "verify = not_blind".
2123 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2124 Local/Makefile (with some defaults set). These are used in built scripts
2125 such as exicyclog, but they have never been used in the exim_install
2126 script (though there are many overriding facilities there). I have
2127 arranged that the exim_install script now takes note of these two
2130 PH/35 Installed configuration files for Dragonfly.
2132 PH/36 When a locally submitted message by a trusted user did not contain a
2133 From: header, and the sender address was obtained from -f or from an SMTP
2134 MAIL command, and the trusted user did not use -F to supply a sender
2135 name, $originator_name was incorrectly used when constructing a From:
2136 header. Furthermore, $originator_name was used for submission mode
2137 messages from external hosts without From: headers in a similar way,
2138 which is clearly wrong.
2140 PH/37 Added control=suppress_local_fixups.
2142 PH/38 When log_selector = +received_sender was set, and the addition of the
2143 sender made the log line's construction buffer exactly full, or one byte
2144 less than full, an overflow happened when the terminating "\n" was
2147 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2148 when the result of a list match is failure because a DNS lookup failed.
2150 PH/40 RM_COMMAND is now used in the building process.
2152 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2153 the "build-* directories that it finds.
2155 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2156 address in a domain literal was a prefix of an interface address.
2158 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2159 when verifying a sender address, unless rewrite_headers is false.
2161 PH/44 Wrote a long comment about why errors_to addresses are verified as
2162 recipients, not senders.
2164 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2165 the ratelimit ACL was added.
2167 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2169 PH/46 Added extra information about PostgreSQL errors to the error string.
2171 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2172 avoiding the need to install special zones in a real server. This is
2173 backwards compatible; if it can't find the fake resolver, it drops back.
2174 Thus, both old and new test suites can be run.
2176 TF/02 Added util/ratelimit.pl
2178 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2179 clock is set back in time.
2181 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2182 Candler <B.Candler@pobox.com>.
2184 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2185 OK for addresses that are the result of redirections.
2187 PH/48 A number of further additions for the benefit of the new test suite,
2188 including a fake gethostbyname() that interfaces to the fake DNS resolver
2191 TF/06 The fix for widen_domains has also been applied to qualify_single and
2192 search_parents which are the other dnslookup options that can cause
2195 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2198 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2200 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2201 service" if it encountered an MX record with an empty target hostname.
2202 The message is now "an MX or SRV record indicated no SMTP service".
2204 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2205 if the DNS of the sending site is misconfigured. This is quite a
2206 common situation. This change restores the behaviour of treating a
2207 helo verification defer as a failure.
2209 PH/51 If self=fail was set on a router, the bounce message did not include the
2210 actual error message.
2216 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2218 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2219 command itself ended in error, the underlying message about the transport
2220 filter timeout was being overwritten with the pipe command error. Now the
2221 underlying error message should be appended to the second error message.
2223 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2225 PH/02 Reduce the amount of output that "make" produces by default. Full output
2226 can still be requested.
2228 PH/03 The warning log line about a condition test deferring for a "warn" verb
2229 was being output only once per connection, rather than after each
2230 occurrence (because it was using the same function as for successful
2231 "warn" verbs). This seems wrong, so I have changed it.
2233 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2234 it should not have, which might have caused a crash in the right
2235 circumstances, but probably never did.
2237 PH/04 Installed a modified version of Tony Finch's patch to make submission
2238 mode fix the return path as well as the Sender: header line, and to
2239 add a /name= option so that you can make the user's friendly name appear
2242 TF/03 Added the control = fakedefer ACL modifier.
2244 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2245 Mark Lowes for thorough testing.
2247 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2249 TK/03 Merged latest SRS patch from Miles Wilton.
2251 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2252 with the definition in sysexits.h (which is #included earlier).
2253 Fortunately, Exim does not actually use EX_OK. The code used to try to
2254 preserve the sysexits.h value, by assumimg that macro definitions were
2255 scanned for macro replacements. I have been disabused of this notion,
2256 so now the code just undefines EX_OK before #including unistd.h.
2258 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2259 in the smtp transport. When a block could not be written in a single
2260 write() function, the timeout was being re-applied to each part-write.
2261 This seems wrong - if the receiver was accepting one byte at a time it
2262 would take for ever. The timeout is now adjusted when this happens. It
2263 doesn't have to be particularly precise.
2265 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2266 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2268 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2269 without a domain if the check to <postmaster@domain> fails.
2271 SC/01 Eximstats: added -xls and the ability to specify output files
2272 (patch written by Frank Heydlauf).
2274 SC/02 Eximstats: use FileHandles for outputing results.
2276 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2278 SC/04 Eximstats: fixed display of large numbers with -nvr option
2280 SC/05 Eximstats: fixed merging of reports with empty tables.
2282 SC/06 Eximstats: added the -include_original_destination flag
2284 SC/07 Eximstats: removed tabs and trailing whitespace.
2286 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2288 TK/06 MBOX spool code: Add real "From " MBOX separator line
2289 so the .eml file is really in mbox format (even though
2290 most programs do not really care). Patch from Alex Miller.
2292 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2293 The latter is generated from $received_to and is only set if the
2294 message has one envelope recipient. SA can use these headers,
2295 obviously out-of-the-box. Patch from Alex Miller.
2297 PH/08 The ${def test on a variable was returning false if the variable's
2298 value was "0", contrary to what the specification has always said!
2299 The result should be true unless the variable is empty.
2301 PH/09 The syntax error of a character other than { following "${if
2302 def:variable_name" (after optional whitespace) was not being diagnosed.
2303 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2304 accidental colon was present, for example, could give incorrect results.
2306 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2307 result is used (not including appendfile, where other changes are about
2310 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2311 This involved changing a lot of size variables from int to off_t. It
2312 should work with maildirs and everything.
2314 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2315 spamd dying while we are connected to it.
2317 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2320 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2321 function for BDB 4.3.
2323 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2325 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2326 the sources for PCRE have been re-arranged and more files are now
2329 PH/15 The code I had for printing potentially long long variables in PH/11
2330 above was not the best (it lost precision). The length of off_t variables
2331 is now inspected at build time, and an appropriate printing format (%ld
2332 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2333 to be "long long int" or "long int". This is needed for the internal
2334 formatting function string_vformat().
2336 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2337 the configuration file to be ":syslog", then the script "guesses" where
2338 the logs files are, rather than using the compiled in default. In our
2339 case the guess is not the same as the compiled default, so the script
2340 suddenly stopped working when I started to use syslog. The patch checks
2341 to see if log_file_path is "". If so, it attempts to read it from exim
2342 with no configuration file to get the compiled in version, before it
2343 falls back to the previous guessing code."
2345 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2346 implementing BATV in an Exim configuration. See NewStuff for the gory
2349 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2350 Makefile that are specific to HP-UX.
2352 PH/18 If the "use_postmaster" option was set for a recipient callout together
2353 with the "random" option, the postmaster address was used as the MAIL
2354 FROM address for the random test, but not for the subsequent recipient
2355 test. It is now used for both.
2357 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2358 patch removes a few documentation additions to RFC 3028, because the
2359 latest draft now contains them. It adds the new en;ascii-case comparator
2360 and a new error check for 8bit text in MIME parts. Comparator and
2361 require names are now matched exactly. I enabled the subaddress
2362 extension, but it is not well tested yet (read: it works for me)."
2364 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2365 rework some of the code of TK/09 above to avoid the hardwired use of
2366 "%lld" and "long long". Replaced the call to snprintf() with a call to
2369 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2370 records point to non-existent hosts", "retry timeout exceeded", and
2371 "retry time not reached for any host after a long failure period".
2373 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2374 experimental DomainKeys support:
2376 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2377 (2) On an error such as an illegally used "control", the wrong name for
2378 the control was given.
2380 These problems did NOT occur unless DomainKeys support was compiled.
2382 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2384 PH/24 Added ${if match_ip condition.
2386 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2387 databases so that it will be absolutely obvious if a crash occurs in the
2388 DB library. This is a regular occurrence (often caused by mis-matched
2391 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2392 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2393 picked up on a user's system that detects such things. There doesn't seem
2394 to be a gcc warning option for this - only an attribute that has to be
2395 put on the function's prototype. It seems that in Fedora Core 4 they have
2396 set this on a number of new functions. No doubt there will be more in due
2399 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2400 specify a transport. However, if an address that was verified by such a
2401 router was the subject of a callout, Exim crashed because it tried to
2402 read the rcpt_include_affixes from the non-existent transport. Now it
2403 just assumes that the setting of that option is false. This bug was
2404 introduced by 4.51/PH/31.
2406 PH/28 Changed -d+all to exclude +memory, because that information is very
2407 rarely of interest, but it makes the output a lot bigger. People tend to
2408 do -d+all out of habit.
2410 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2411 code in os-type was giving problems when libc.so lives in lib64, like on
2414 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2415 aren't the modern standard, and it seems that some systems' include files
2416 don't always have them. Exim was already checking for some of the newer
2417 ones like T_AAAA, and defining it itself. I've added checks for all the
2418 record types that Exim uses.
2420 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2421 not automatically generating a new one, as it is supposed to. This
2422 prevented TLS from working. If the file did exist, but contained invalid
2423 data, a new version was generated, as expected. It was only the case of a
2424 non-existent file that was broken.
2426 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2427 with a change in libdomainkeys > 0.64.
2429 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2430 from DNS. If the selector record carries the flag, it now has
2431 precedence over the domain-wide flag.
2433 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2435 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2436 the use of a port name that isn't defined in /etc/services) occurred, the
2437 message was deferred as in a normal delivery, and thus remained on the
2438 spool, instead of being failed because of the mua_wrapper setting. This
2439 is now fixed, and I tidied up some of the mua_wrapper messages at the
2442 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2443 the messages in a hash of arrays rather than using individual hashes.
2444 This is a bit cleaner and results in dramatic memory savings, albeit
2445 at a slight CPU cost.
2447 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2448 as requested by Marc Sherman.
2450 SC/10 Eximstats: added histograms for user specified patterns as requested
2453 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2455 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2456 fopen() in the content-scanning modules that did not already have it.
2462 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2463 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2465 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2467 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2469 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2470 to test Sieve filters that use "vacation".
2472 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2473 that changes the way the GnuTLS parameters are stored in the cache file.
2474 The new format can be generated externally. For backward compatibility,
2475 if the data in the cache doesn't make sense, Exim assumes it has read an
2476 old-format file, and it generates new data and writes a new file. This
2477 means that you can't go back to an older release without removing the
2480 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2481 work if there are any delivery delays because "one_time" forces the
2482 parent to be marked "delivered", so its unseen clone is never tried
2483 again. For this reason, Exim now forbids the simultaneous setting of
2486 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2487 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2488 entry if you want to know the details. The fix, however, neglected to
2489 consider the case when local delivery batching is involved. The test for
2490 "previously delivered" was not happening when checking to see if an
2491 address could be batched with a previous (undelivered) one; under
2492 certain circumstances this could lead to multiple deliveries to the same
2495 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2496 in its include files, and this causes problems building Exim.
2498 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2499 header_syntax) but Exim was just ignoring anything given after a slash.
2500 In particular, this caused confusion with an attempt to use "verify =
2501 reverse_host_lookup/defer_ok". An error is now given when options are
2502 supplied for verify items that do not have them. (Maybe reverse_host_
2503 lookup should have a defer_ok option, but that's a different point.)
2505 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2506 defined by RFC 821) to 2048, because there were problems with some AUTH
2507 commands, and RFC 1869 says the size should be increased for extended
2508 SMTP commands that take arguments.
2510 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2513 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2514 "unknown" error; now it says that the functionality isn't in the binary.
2516 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2517 an address' error message when a string expansion fails (syntax or
2518 whatever). Otherwise the password may appear in the log. Following change
2519 PH/42 below, there is no longer a chance of it appearing in a bounce
2522 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2524 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2525 of its addresses was ever tried. (Bugzilla bug #2.)
2527 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2528 the result incorrectly in the debug output. (It correctly added a newline
2529 to what was transported.)
2531 TF/01 Added $received_time.
2533 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2534 commented out examples of how to interface to a virus scanner and to
2535 SpamAssassin. Also added commented examples of av_scanner and
2536 spamd_address settings.
2538 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2539 and controls are allowed in which ACLs. There were a couple of minor
2540 errors. Some of the entries in the conditions table (which is a table of
2541 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2542 negation of where the condition IS allowed.
2544 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2546 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2547 header file does not have a version number, so I've had to invent a new
2548 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2549 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2550 radiusclient), but it was contributed by a Radius user.
2552 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2553 files or directories, for OpenSSL.
2555 PH/19 When an Exim process that is running as root has to create an Exim log
2556 file, it does so in a subprocess that runs as exim:exim so as to get the
2557 ownership right at creation (otherwise, other Exim processes might see
2558 the file with the wrong ownership). There was no test for failure of this
2559 fork() call, which would lead to the process getting stuck as it waited
2560 for a non-existent subprocess. Forks do occasionally fail when resources
2561 run out. I reviewed all the other calls to fork(); they all seem to check
2564 PH/20 When checking for unexpected SMTP input at connect time (before writing
2565 the banner), Exim was not dealing correctly with a non-positive return
2566 from the read() function. If the client had disconnected by this time,
2567 the result was a log entry for a synchronization error with an empty
2568 string after "input=" when read() returned zero. If read() returned -1
2569 (an event I could not check), uninitialized data bytes were printed.
2570 There were reports of junk text (parts of files, etc) appearing after
2573 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2575 PH/22 Added support for macro redefinition, and (re)definition in between
2576 driver and ACL definitions.
2578 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2579 forgetting to use the resulting value; it was using the unexpanded value.
2581 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2582 hadn't been configured. The fix is from Juergen Kreileder, who
2583 understands it better than I do:
2585 "Here's what I see happening with three configured cyrus_sasl
2586 authenticators configured (plain, login, cram-md5):
2588 On startup auth_cyrus_sasl_init() gets called for each of these.
2589 This means three calls to sasl_listmech() without a specified mech_list.
2590 => SASL tests which mechs of all available mechs actually work
2591 => three warnings about OTP not working
2592 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2594 With the patch, sasl_listmech() also gets called three times. But now
2595 SASL's mech_list option is set to the server_mech specified in the the
2596 authenticator. Or in other words, the answer from sasl_listmech()
2597 gets limited to just the mech you're testing for (which is different
2599 => the return list contains just 'plain' or 'login', 'cram-md5' or
2600 nothing depending on the value of ob->server_mech.
2602 I've just tested the patch: Authentication still works fine,
2603 unavailable mechs specified in the exim configuration are still
2604 caught, and the auth.log warnings about OTP are gone."
2606 PH/25 When debugging is enabled, the contents of the command line are added
2607 to the debugging output, even when log_selector=+arguments is not
2610 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2611 answer is "GNU", and only if the return is "GNU/something" is the answer
2614 PH/27 $acl_verify_message is now set immediately after the failure of a
2615 verification in an ACL, and so is available in subsequent modifiers. In
2616 particular, the message can be preserved by coding like this:
2618 warn !verify = sender
2619 set acl_m0 = $acl_verify_message
2621 Previously, $acl_verify_message was set only while expanding "message"
2622 and "log_message" when a very denied access.
2624 PH/28 Modified OS/os.c-Linux with
2626 -#ifndef OS_LOAD_AVERAGE
2627 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2629 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2630 nomenclature these days.)
2632 PH/29 Installed patch from the Sieve maintainer that adds the options
2633 sieve_useraddress and sieve_subaddress to the redirect router.
2635 PH/30 In these circumstances:
2636 . Two addresses routed to the same list of hosts;
2637 . First host does not offer TLS;
2638 . First host accepts first address;
2639 . First host gives temporary error to second address;
2640 . Second host offers TLS and a TLS session is established;
2641 . Second host accepts second address.
2642 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2643 and peerdn, if requested) that were in fact used only for the second
2646 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2647 attention to any local part prefix or suffix that was matched by the
2648 router that accepted the address. It now behaves in the same way as it
2649 does for delivery: the affixes are removed from the local part unless
2650 rcpt_include_affixes is set on the transport.
2652 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2653 timeout during the DATA phase of an incoming message.
2655 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2656 applied a patch sanctioned by the Sieve maintainer.
2658 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2659 the uid or gid is negative. A case of a negative gid caused this to be
2660 noticed. The fix allows for either to be negative.
2662 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2663 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2666 PH/36 The change PH/12 above was broken. Fixed it.
2668 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2669 the grounds that routing the same address twice would always produce the
2670 same answer. This might have been true once, but it is certainly no
2671 longer true now. Routing a child address may depend on the previous
2672 routing that produced that child. Some complicated redirection strategies
2673 went wrong when messages had multiple recipients, and made Exim's
2674 behaviour dependent on the order in which the addresses were given.
2676 I have moved the duplicate checking until after the routing is complete.
2677 Exim scans the addresses that are assigned to local and remote
2678 transports, and removes any duplicates. This means that more work will be
2679 done, as duplicates will always all be routed, but duplicates are
2680 presumably rare, so I don't expect this is of any significance.
2682 For deliveries to pipes, files, and autoreplies, the duplicate checking
2683 still happens during the routing process, since they are not going to be
2686 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2687 It corrects a timeout issue with spamd. This is Ian's comment: "The
2688 background is that sometimes spamd either never reads data from a
2689 connection it has accepted, or it never writes response data. The exiscan
2690 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2691 blindly assumes that writes won't block so it may never time out."
2693 PH/39 Allow G after quota size as well as K and M.
2695 PH/40 The value set for $authenticated_id in an authenticator may not contain
2696 binary zeroes or newlines because the value is written to log lines and
2697 to spool files. There was no check on this. Now the value is run through
2698 the string_printing() function so that such characters are converted to
2699 printable escape sequences.
2701 PH/41 $message_linecount is a new variable that contains the total number of
2702 lines in the message. Compare $body_linecount, which is the count for the
2705 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2706 bounce and delay warning messages, except in certain special cases, which
2709 (a) An SMTP error message from a remote host;
2710 (b) A message specified in a :fail: redirection;
2711 (c) A message specified in a "fail" command in a system filter;
2712 (d) A message specified in a FAIL return from the queryprogram router;
2713 (e) A message specified by the cannot_route_message router option.
2715 In these cases only, Exim does include the error details in bounce and
2716 warning messages. There are also a few cases where bland messages such
2717 as "unrouteable address" or "local delivery error" are given.
2719 PH/43 $value is now also set for the "else" part of a ${run expansion.
2721 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2722 being worked on, but at least Exim now implements the latest version to
2725 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2726 process to complete was treated as a delivery failure, a timeout while
2727 writing the message to the pipe was logged, but erroneously treated as a
2728 successful delivery. Such timeouts include transport filter timeouts. For
2729 consistency with the overall process timeout, these timeouts are now
2730 treated as errors, giving rise to delivery failures by default. However,
2731 there is now a new Boolean option for the pipe transport called
2732 timeout_defer, which, if set TRUE, converts the failures into defers for
2733 both kinds of timeout. A transport filter timeout is now identified in
2736 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2737 systems where "make" and "gmake" are different, calling "gmake" at top
2738 level broke things. I've arranged for the value of $(MAKE) to be passed
2739 from the Makefile to this script so that it can call the same version of
2743 A note about Exim versions 4.44 and 4.50
2744 ----------------------------------------
2746 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2747 changes of various kinds. As a consequence, a big documentation update was
2748 needed. This delayed the release for rather longer than seemed good, especially
2749 in the light of a couple of (minor) security issues. Therefore, the changes
2750 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2751 release. So 4.44 and 4.50 are in effect two different branches that both start
2754 I have left the 4.50 change log unchanged; it contains all the changes since
2755 4.43. The change log for 4.44 is below; many of its items are identical to
2756 those for 4.50. This seems to be the most sensible way to preserve the
2757 historical information.
2763 1. Minor wording change to the doc/README.SIEVE file.
2765 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2766 computation of the current number of files was incorrect.
2768 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2769 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2772 4. Give more explanation in the error message when the command for a transport
2773 filter fails to execute.
2775 5. There are several places where Exim runs a non-Exim command in a
2776 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2777 was being done only for the command run by the queryprogram router. It is
2778 now done for all such subprocesses. The other cases are: ${run, transport
2779 filters, and the commands run by the lmtp and pipe transports.
2781 6. Added CONFIGURE_GROUP build-time option.
2783 7. Some older OS have a limit of 256 on the maximum number of file
2784 descriptors. Exim was using setrlimit() to set 1000 as a large value
2785 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2786 systems. I've change it so that if it can't get 1000, it tries for 256.
2788 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2789 was an oversight, and furthermore, ever since the addition of extra
2790 controls (e.g. 4.43/32), the checks on when to allow different forms of
2791 "control" were broken. There should now be diagnostics for all cases when a
2792 control that does not make sense is encountered.
2794 9. Added the /retain_sender option to "control=submission".
2796 10. $recipients is now available in the predata ACL (oversight).
2798 11. Tidy the search cache before the fork to do a delivery from a message
2799 received from the command line. Otherwise the child will trigger a lookup
2800 failure and thereby defer the delivery if it tries to use (for example) a
2801 cached ldap connection that the parent has called unbind on.
2803 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2804 of $address_data from the recipient verification was clobbered by the
2805 sender verification.
2807 13. The value of address_data from a sender verification is now available in
2808 $sender_address_data in subsequent conditions in the ACL statement.
2810 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2812 15. Added a new option "connect=<time>" to callout options, to set a different
2815 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2816 was its contents. (It was OK if the option was not defined at all.)
2818 17. A "Completed" log line is now written for messages that are removed from
2819 the spool by the -Mrm option.
2821 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2822 information about exactly what failed.
2824 19. Added -dd to debug only the daemon process.
2826 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2827 handles timeouts, both on the server side and network timeouts. Renamed the
2828 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2830 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2831 It is now set to "smtps".
2833 22. $host_address is now set to the target address during the checking of
2834 ignore_target_hosts.
2836 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2837 being passed; this would have caused $sender_host_name to have been used if
2838 matching the list had actually called for a host name (not very likely,
2839 since this list is usually IP addresses). A host name is now passed as
2842 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2843 code that specifies a non-restarting handler (typically sigaction() in
2844 modern systems) in an attempt to fix a rare and obscure crash bug.
2846 25. Narrowed the window for a race in the daemon that could cause it to ignore
2847 SIGCHLD signals. This is not a major problem, because they are used only to
2848 wake it up if nothing else does.
2850 26. A malformed maildirsize file could cause Exim to calculate negative values
2851 for the mailbox size or file count. Odd effects could occur as a result.
2852 The maildirsize information is now recalculated if the size or filecount
2855 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2856 support for a long time. Removed HAVE_SYS_VFS_H.
2858 28. Installed the latest version of exipick from John Jetmore.
2860 29. In an address list, if the pattern was not a regular expression, an empty
2861 subject address (from a bounce message) matched only if the pattern was an
2862 empty string. Non-empty patterns were not even tested. This was the wrong
2863 because it is perfectly reasonable to use an empty address as part of a
2864 database query. An empty address is now tested by patterns that are
2865 lookups. However, all the other forms of pattern expect the subject to
2866 contain a local part and a domain, and therefore, for them, an empty
2867 address still always fails if the pattern is not itself empty.
2869 30. Exim went into a mad DNS loop when attempting to do a callout where the
2870 host was specified on an smtp transport, and looking it up yielded more
2871 than one IP address.
2873 31. Re-factored the code for checking spool and log partition space into a
2874 function that finds that data and another that does the check. The former
2875 is then used to implement four new variables: $spool_space, $log_space,
2876 $spool_inodes, and $log_inodes.
2878 32. The RFC2047 encoding function was originally intended for short strings
2879 such as real names; it was not keeping to the 75-character limit for
2880 encoded words that the RFC imposes. It now respects the limit, and
2881 generates multiple encoded words if necessary. To be on the safe side, I
2882 have increased the buffer size for the ${rfc2047: expansion operator from
2885 33. It is now permitted to omit both strings after an "if" condition; if the
2886 condition is true, the result is "true". As before, when the second string
2887 is omitted, a false condition yields an empty string. This makes it less
2888 cumbersome to write custom ACL and router conditions.
2890 34. Failure to deliver a bounce message always caused it to be frozen, even if
2891 there was an errors_to setting on the router. The errors_to setting is now
2894 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2895 canonical form (fully expanded) before being placed in
2896 $sender_host_address.
2898 36. The table in the code that translates DNS record types into text (T_A to
2899 "A" for instance) was missing entries for NS and CNAME. It is just possible
2900 that this could have caused confusion if both these types were looked up
2901 for the same domain, because the text type is used as part of Exim's
2902 per-process caching. But the chance of anyone hitting this buglet seems
2905 37. The dnsdb lookup has been extended in a number of ways.
2907 (1) There is a new type, "zns", which walks up the domain tree until it
2908 finds some nameserver records. It should be used with care.
2910 (2) There is a new type, "mxh", which is like "mx" except that it returns
2911 just the host names, not the priorities.
2913 (3) It is now possible to give a list of domains (or IP addresses) to be
2914 looked up. The behaviour when one of the lookups defers can be
2915 controlled by a keyword.
2917 (4) It is now possible to specify the separator character for use when
2918 multiple records are returned.
2920 38. The dnslists ACL condition has been extended: it is now possible to supply
2921 a list of IP addresses and/or domains to be looked up in a particular DNS
2924 39. Added log_selector=+queue_time_overall.
2926 40. When running the queue in the test harness, wait just a tad after forking a
2927 delivery process, to get repeatability of debugging output.
2929 41. Include certificate and key file names in error message when GnuTLS fails
2930 to set them up, because the GnuTLS error message doesn't include the name
2931 of the failing file when there is a problem reading it.
2933 42. Allow both -bf and -bF in the same test run.
2935 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2937 44. The "Exiscan patch" is now merged into the mainline Exim source.
2939 45. Sometimes the final signoff response after QUIT could fail to get
2940 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2941 < 0 before doing a fflush(). This bug looks as though it goes back to the
2942 introduction of TLS in release 3.20, but "sometimes" must have been rare
2943 because the tests only now provoked it.
2945 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2946 (this can affect the format of dates).
2948 47. exim_tidydb, when checking for the continued existence of a message for
2949 which it has found a message-specific retry record, was not finding
2950 messages that were in split spool directories. Consequently, it was
2951 deleting retry records that should have stayed in existence.
2953 48. Steve fixed some bugs in eximstats.
2955 49. The SPA authentication driver was not abandoning authentication and moving
2956 on to the next authenticator when an expansion was forced to fail,
2957 contradicting the general specification for all authenticators. Instead it
2958 was generating a temporary error. It now behaves as specified.
2960 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2961 (the order specifies the preference for clients). The order is now AES256,
2962 AES128, 3DES, ARCFOUR128.
2964 51. Small patch to Sieve code - explicitly set From: when generating an
2967 52. Exim crashed if a remote delivery caused a very long error message to be
2968 recorded - for instance if somebody sent an entire SpamAssassin report back
2969 as a large number of 550 error lines. This bug was coincidentally fixed by
2970 increasing the size of one of Exim's internal buffers (big_buffer) that
2971 happened as part of the Exiscan merge. However, to be on the safe side, I
2972 have made the code more robust (and fixed the comments that describe what
2975 53. Now that there can be additional text after "Completed" in log lines (if
2976 the queue_time_overall log selector is set), a one-byte patch to exigrep
2977 was needed to allow it to recognize "Completed" as not the last thing in
2980 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2981 patch that reportedly fixes this has been added. I am not expert enough to
2982 create a test for it. This is what the patch creator wrote:
2984 "I found a little strange behaviour of ldap code when working with
2985 Windows 2003 AD Domain, where users was placed in more than one
2986 Organization Units. When I tried to give exim partial DN, the exit code
2987 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2988 But simultaneously result of request was absolutely normal ldap result,
2989 so I produce this patch..."
2991 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2992 REFERENCE, so I have modified the code to exclude the patch when that macro
2995 55. Some experimental protocols are using DNS PTR records for new purposes. The
2996 keys for these records are domain names, not reversed IP addresses. The
2997 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2998 leaves it alone. Component reversal etc. now happens only for IP addresses.
3001 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3003 57. Double the size of the debug message buffer (to 2048) so that more of very
3004 long debug lines gets shown.
3006 58. The exicyclog utility now does better if the number of log files to keep
3007 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3009 59. Two changes related to the smtp_active_hostname option:
3011 (1) $smtp_active_hostname is now available as a variable.
3012 (2) The default for smtp_banner uses $smtp_active_hostname instead
3013 of $primary_hostname.
3015 60. The host_aton() function is supposed to be passed a string that is known
3016 to be a valid IP address. However, in the case of IPv6 addresses, it was
3017 not checking this. This is a hostage to fortune. Exim now panics and dies
3018 if the condition is not met. A case was found where this could be provoked
3019 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3020 components; fortuitously, this particular loophole had already been fixed
3021 by change 4.50/55 above.
3023 If there are any other similar loopholes, the new check in host_aton()
3024 itself should stop them being exploited. The report I received stated that
3025 data on the command line could provoke the exploit when Exim was running as
3026 exim, but did not say which command line option was involved. All I could
3027 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3028 running as the user.
3031 61. There was a buffer overflow vulnerability in the SPA authentication code
3032 (which came originally from the Samba project). I have added a test to the
3033 spa_base64_to_bits() function which I hope fixes it.
3036 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3037 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3039 63. The daemon start-up calls getloadavg() while still root for those OS that
3040 need the first call to be done as root, but it missed one case: when
3041 deliver_queue_load_max is set with deliver_drop_privilege. This is
3042 necessary for the benefit of the queue runner, because there is no re-exec
3043 when deliver_drop_privilege is set.
3045 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3046 This has been fixed.
3048 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3049 was in use, was not putting the data itself into the right store pool;
3050 consequently, it could be overwritten for a subsequent message in the same
3051 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3054 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3056 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3057 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3058 the function were treating the return as a boolean value, which happened to
3059 work because 0=false and not-0=true, but is not correct code.
3061 68. The host_aton() function was not handling scoped IPv6 addresses (those
3062 with, for example, "%eth0" on the end) correctly.
3064 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3065 negated items (that is, ~something) in unsigned ints. Some compilers
3066 apparently mutter when there is no cast.
3068 70. If an address verification called from an ACL failed, and did not produce a
3069 user-specific message (i.e. there was only a "system" message), nothing was
3070 put in $acl_verify_message. In this situation, it now puts the system
3073 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3074 session; change 4.31/43 added the unwanted input to the log line - except
3075 that it did not do this in the start of session case. It now does.
3077 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3078 This is wrong and can cause the other end to generate a synchronization
3079 error if it is another Exim or anything else that does the synchronization
3080 check. A QUIT command is no longer sent after a timeout.
3082 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3083 during host lookups.
3085 74. The defer_ok option of callout verification was not working if it was used
3086 when verifying addresses in header lines, that is, for this case:
3088 verify = header_sender/callout=defer_ok
3090 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3091 those file descriptors could be used for SMTP connections. If anything
3092 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3093 could be sent to the SMTP client, causing chaos. The daemon now opens
3094 stdin, stdout, and stderr to /dev/null when it puts itself into the
3097 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3098 log by default. The user can override this with suitable Perl magic.
3100 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3101 the log message when discard triggers, was not working for the DATA ACL or
3102 for the non-SMTP ACL.
3104 78. Error message wording change in sieve.c.
3106 79. If smtp_accept_max_per_host was set, the number of connections could be
3107 restricted to fewer than expected, because the daemon was trying to set up
3108 a new connection before checking whether the processes handling previous
3109 connections had finished. The check for completed processes is now done
3110 earlier. On busy systems, this bug wouldn't be noticed because something
3111 else would have woken the daemon, and it would have reaped the completed
3114 80. If a message was submitted locally by a user whose login name contained one
3115 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3116 It caused a spool format error. I have fixed the spool reading code. A
3117 related problem was that the "from" clause in the Received: line became
3118 illegal because of the space(s). It is now covered by ${quote_local_part.
3120 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3123 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3125 83. Patch from Sieve maintainer to fix unterminated string problem in
3126 "vacation" handling.
3128 84. Some minor changes to the Linux configuration files to help with other
3129 OS variants using glibc.
3131 85. One more patch for Sieve to update vacation handling to latest spec.
3134 ----------------------------------------------------
3135 See the note above about the 4.44 and 4.50 releases.
3136 ----------------------------------------------------
3142 1. Change 4.43/35 introduced a bug that caused file counts to be
3143 incorrectly computed when quota_filecount was set in an appendfile
3146 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3147 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3150 3. Give more explanation in the error message when the command for a transport
3151 filter fails to execute.
3153 4. There are several places where Exim runs a non-Exim command in a
3154 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3155 was being done only for the command run by the queryprogram router. It is
3156 now done for all such subprocesses. The other cases are: ${run, transport
3157 filters, and the commands run by the lmtp and pipe transports.
3159 5. Some older OS have a limit of 256 on the maximum number of file
3160 descriptors. Exim was using setrlimit() to set 1000 as a large value
3161 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3162 systems. I've change it so that if it can't get 1000, it tries for 256.
3164 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3165 was an oversight, and furthermore, ever since the addition of extra
3166 controls (e.g. 4.43/32), the checks on when to allow different forms of
3167 "control" were broken. There should now be diagnostics for all cases when a
3168 control that does not make sense is encountered.
3170 7. $recipients is now available in the predata ACL (oversight).
3172 8. Tidy the search cache before the fork to do a delivery from a message
3173 received from the command line. Otherwise the child will trigger a lookup
3174 failure and thereby defer the delivery if it tries to use (for example) a
3175 cached ldap connection that the parent has called unbind on.
3177 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3178 of $address_data from the recipient verification was clobbered by the
3179 sender verification.
3181 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3182 was its contents. (It was OK if the option was not defined at all.)
3184 11. A "Completed" log line is now written for messages that are removed from
3185 the spool by the -Mrm option.
3187 12. $host_address is now set to the target address during the checking of
3188 ignore_target_hosts.
3190 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3191 being passed; this would have caused $sender_host_name to have been used if
3192 matching the list had actually called for a host name (not very likely,
3193 since this list is usually IP addresses). A host name is now passed as
3196 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3197 code that specifies a non-restarting handler (typically sigaction() in
3198 modern systems) in an attempt to fix a rare and obscure crash bug.
3200 15. Narrowed the window for a race in the daemon that could cause it to ignore
3201 SIGCHLD signals. This is not a major problem, because they are used only to
3202 wake it up if nothing else does.
3204 16. A malformed maildirsize file could cause Exim to calculate negative values
3205 for the mailbox size or file count. Odd effects could occur as a result.
3206 The maildirsize information is now recalculated if the size or filecount
3209 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3210 support for a long time. Removed HAVE_SYS_VFS_H.
3212 18. Updated exipick to current release from John Jetmore.
3214 19. Allow an empty sender to be matched against a lookup in an address list.
3215 Previously the only cases considered were a regular expression, or an
3218 20. Exim went into a mad DNS lookup loop when doing a callout where the
3219 host was specified on the transport, if the DNS lookup yielded more than
3222 21. The RFC2047 encoding function was originally intended for short strings
3223 such as real names; it was not keeping to the 75-character limit for
3224 encoded words that the RFC imposes. It now respects the limit, and
3225 generates multiple encoded words if necessary. To be on the safe side, I
3226 have increased the buffer size for the ${rfc2047: expansion operator from
3229 22. Failure to deliver a bounce message always caused it to be frozen, even if
3230 there was an errors_to setting on the router. The errors_to setting is now
3233 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3234 canonical form (fully expanded) before being placed in
3235 $sender_host_address.
3237 24. Updated eximstats to version 1.33
3239 25. Include certificate and key file names in error message when GnuTLS fails
3240 to set them up, because the GnuTLS error message doesn't include the name
3241 of the failing file when there is a problem reading it.
3243 26. Expand error message when OpenSSL has problems setting up cert/key files.
3246 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3247 (this can affect the format of dates).
3249 28. exim_tidydb, when checking for the continued existence of a message for
3250 which it has found a message-specific retry record, was not finding
3251 messages that were in split spool directories. Consequently, it was
3252 deleting retry records that should have stayed in existence.
3254 29. eximstats updated to version 1.35
3255 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3256 1.35 - bugfix such that pie charts by volume are generated correctly
3258 30. The SPA authentication driver was not abandoning authentication and moving
3259 on to the next authenticator when an expansion was forced to fail,
3260 contradicting the general specification for all authenticators. Instead it
3261 was generating a temporary error. It now behaves as specified.
3263 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3264 (the order specifies the preference for clients). The order is now AES256,
3265 AES128, 3DES, ARCFOUR128.
3267 31. Small patch to Sieve code - explicitly set From: when generating an
3270 32. Exim crashed if a remote delivery caused a very long error message to be
3271 recorded - for instance if somebody sent an entire SpamAssassin report back
3272 as a large number of 550 error lines. This bug was coincidentally fixed by
3273 increasing the size of one of Exim's internal buffers (big_buffer) that
3274 happened as part of the Exiscan merge. However, to be on the safe side, I
3275 have made the code more robust (and fixed the comments that describe what
3278 33. Some experimental protocols are using DNS PTR records for new purposes. The
3279 keys for these records are domain names, not reversed IP addresses. The
3280 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3281 leaves it alone. Component reversal etc. now happens only for IP addresses.
3284 34. The host_aton() function is supposed to be passed a string that is known
3285 to be a valid IP address. However, in the case of IPv6 addresses, it was
3286 not checking this. This is a hostage to fortune. Exim now panics and dies
3287 if the condition is not met. A case was found where this could be provoked
3288 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3289 components; fortuitously, this particular loophole had already been fixed
3290 by change 4.50/55 or 4.44/33 above.
3292 If there are any other similar loopholes, the new check in host_aton()
3293 itself should stop them being exploited. The report I received stated that
3294 data on the command line could provoke the exploit when Exim was running as
3295 exim, but did not say which command line option was involved. All I could
3296 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3297 running as the user.
3300 35. There was a buffer overflow vulnerability in the SPA authentication code
3301 (which came originally from the Samba project). I have added a test to the
3302 spa_base64_to_bits() function which I hope fixes it.
3305 36. The daemon start-up calls getloadavg() while still root for those OS that
3306 need the first call to be done as root, but it missed one case: when
3307 deliver_queue_load_max is set with deliver_drop_privilege. This is
3308 necessary for the benefit of the queue runner, because there is no re-exec
3309 when deliver_drop_privilege is set.
3311 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3312 was in use, was not putting the data itself into the right store pool;
3313 consequently, it could be overwritten for a subsequent message in the same
3314 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3317 38. Sometimes the final signoff response after QUIT could fail to get
3318 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3319 < 0 before doing a fflush(). This bug looks as though it goes back to the
3320 introduction of TLS in release 3.20, but "sometimes" must have been rare
3321 because the tests only now provoked it.
3327 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3328 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3329 Now it returns an integer. A number of calls were still expecting a T/F
3330 return. Fortuitously, in all cases, the tests worked in OK situations,
3331 which is the norm. However, things would have gone wrong on any write
3332 failures on the smtp file descriptor. This function is used when sending
3333 messages over SMTP and also when doing verify callouts.
3335 2. When Exim is called to do synchronous delivery of a locally submitted
3336 message (the -odf or -odi options), it no longer closes stderr before doing
3339 3. Implemented the mua_wrapper option.
3341 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3343 5. Implemented the functions header_remove(), header_testname(),
3344 header_add_at_position(), and receive_remove_recipient(), and exported them
3347 6. If an ACL "warn" statement specified the addition of headers, Exim already
3348 inserted X-ACL-Warn: at the start if there was no header name. However, it
3349 was not making this test for the second and subsequent header lines if
3350 there were newlines in the string. This meant that an invalid header could
3351 be inserted if Exim was badly configured.
3353 7. Allow an ACL "warn" statement to add header lines at the start or after all
3354 the Received: headers, as well as at the end.
3356 8. Added the rcpt_4xx retry error code.
3358 9. Added postmaster_mailfrom=xxx to callout verification option.
3360 10. Added mailfrom=xxxx to the callout verification option, for verify=
3363 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3364 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3366 12. Inserted some casts to stop certain compilers warning when using pointer
3367 differences as field lengths or precisions in printf-type calls (mostly
3368 affecting debugging statements).
3370 13. Added optional readline() support for -be (dynamically loaded).
3372 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3373 same clock tick as a message's arrival, so that its received time was the
3374 same as the "first fail" time on the retry record, and that message
3375 remained on the queue past the ultimate address timeout, every queue runner
3376 would try a delivery (because it was past the ultimate address timeout) but
3377 after another failure, the ultimate address timeout, which should have then
3378 bounced the address, did not kick in. This was a "< instead of <=" error;
3379 in most cases the first failure would have been in the next clock tick
3380 after the received time, and all would be well.
3382 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3383 being recognized when the domain list was tested by the match_domain
3384 condition in an expansion string.
3386 16. Added the ${str2b64: operator.
3388 17. Exim was always calling setrlimit() to set a large limit for the number of
3389 processes, without checking whether the existing limit was already
3390 adequate. (It did check for the limit on file descriptors.) Furthermore,
3391 errors from getrlimit() and setrlimit() were being ignored. Now they are
3392 logged to the main and panic logs, but Exim does carry on, to try to do its
3393 job under whatever limits there are.
3395 18. Imported PCRE 5.0.
3397 19. Trivial typo in log message " temporarily refused connection" (the leading
3400 20. If the log selector return_path_on_delivery was set and an address was
3401 redirected to /dev/null, the delivery process crashed because it assumed
3402 that a return path would always be set for a "successful" delivery. In this
3403 case, the whole delivery is bypassed as an optimization, and therefore no
3406 21. Internal re-arrangement: the function for sending a challenge and reading
3407 a response while authentication was assuming a zero-terminated challenge
3408 string. It's now changed to take a pointer and a length, to allow for
3409 binary data in such strings.
3411 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3413 23. Exim was not respecting finduser_retries when seeking the login of the
3414 uid under which it was called; it was always trying 10 times. (The default
3415 setting of finduser_retries is zero.) Also, it was sleeping after the final
3416 failure, which is pointless.
3418 24. Implemented tls_on_connect_ports.
3420 25. Implemented acl_smtp_predata.
3422 26. If the domain in control=submission is set empty, Exim assumes that the
3423 authenticated id is a complete email address when it generates From: or
3424 Sender: header lines.
3426 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3427 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3428 chown and chgrp in /bin and hostname in /usr/bin.
3430 28. Exim was keeping the "process log" file open after each use, just as it
3431 does for the main log. This opens the possibility of it remaining open for
3432 long periods when the USR1 signal hits a daemon. Occasional processlog
3433 errors were reported, that could have been caused by this. Anyway, it seems
3434 much more sensible not to leave this file open at all, so that is what now
3437 29. The long-running daemon process does not normally write to the log once it
3438 has entered its main loop, and it closes the log before doing so. This is
3439 so that log files can straightforwardly be renamed and moved. However,
3440 there are a couple of unusual error situations where the daemon does write
3441 log entries, and I had neglected to close the log afterwards.
3443 30. The text of an SMTP error response that was received during a remote
3444 delivery was being truncated at 512 bytes. This is too short for some of
3445 the long messages that one sometimes sees. I've increased the limit to
3448 31. It is now possible to make retry rules that apply only when a message has a
3449 specific sender, in particular, an empty sender.
3451 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3452 it possible to be selective about when SMTP synchronization is enforced.
3454 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3456 32. Implemented hosts_connection_nolog.
3458 33. Added an ACL for QUIT.
3460 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3463 35. Added mailbox_size and mailbox_filecount to appendfile.
3465 36. Added control = no_multiline_responses to ACLs.
3467 37. There was a bug in the logic of the code that waits for the clock to tick
3468 in the case where the clock went backwards by a substantial amount such
3469 that the microsecond fraction of "now" was more than the microsecond
3470 fraction of "then" (but the whole seconds number was less).
3472 38. Added support for the libradius Radius client library this is found on
3473 FreeBSD (previously only the radiusclient library was supported).
3479 1. When certain lookups returned multiple values in the form name=value, the
3480 quoting of the values was not always being done properly. Specifically:
3481 (a) If the value started with a double quote, but contained no whitespace,
3483 (b) If the value contained whitespace other than a space character (i.e.
3484 tabs or newlines or carriage returns) it was not quoted.
3485 This fix has been applied to the mysql and pgsql lookups by writing a
3486 separate quoting function and calling it from the lookup code. The fix
3487 should probably also be applied to nisplus, ibase and oracle lookups, but
3488 since I cannot test any of those, I have not disturbed their existing code.
3490 2. A hit in the callout cache for a specific address caused a log line with no
3491 reason for rejecting RCPT. Now it says "Previous (cached) callout
3492 verification failure".
3494 3. There was an off-by-one bug in the queryprogram router. An over-long
3495 return line was truncated at 256 instead of 255 characters, thereby
3496 overflowing its buffer with the terminating zero. As well as fixing this, I
3497 have increased the buffer size to 1024 (and made a note to document this).
3499 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3500 when Exim is waiting for an SMTP response from a remote server, Exim
3501 restarts its select() call on the socket, thereby resetting its timeout.
3502 This is not a problem when such interrupts are rare. Somebody set up a cron
3503 job to run exiwhat every 2 minutes, which is less than the normal select()
3504 timeout (5 or 10 minutes). This meant that the select() timeout never
3505 kicked in because it was always reset. I have fixed this by comparing the
3506 time when an interrupt arrives with the time at the start of the first call
3507 to select(). If more time than the timeout has elapsed, the interrupt is
3508 treated as a timeout.
3510 5. Some internal re-factoring in preparation for the addition of Sieve
3511 extensions (by MH). In particular, the "personal" test is moved to a
3512 separate function, and given an option for scanning Cc: and Bcc: (which is
3513 not set for Exim filters).
3515 6. When Exim created an email address using the login of the caller as the
3516 local part (e.g. when creating a From: or Sender: header line), it was not
3517 quoting the local part when it contained special characters such as @.
3519 7. Installed new OpenBSD configuration files.
3521 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3522 try to make them clearer.
3524 9. Callout options, other than the timeout value, were being ignored when
3525 verifying sender addresses in header lines. For example, when using
3527 verify = header_sender/callout=no_cache
3529 the cache was (incorrectly) being used.
3531 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3532 only the Cygwin environment.
3534 11. When return_path_on_delivery was set as a log selector, if different remote
3535 addresses in the same message used different return paths and parallel
3536 remote delivery occurred, the wrong values would sometimes be logged.
3537 (Whenever a remote delivery process finished, the return path value from
3538 the most recently started remote delivery process was logged.)
3540 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3541 lines when AUTH and/or TLS are in use. This is the "received protocol"
3542 field. Exim used to use "asmtp" for authenticated SMTP, without any
3543 indication (in the protocol name) for TLS use. Now it follows the RFC and
3544 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3545 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3546 names appear in log lines as well as in Received: header lines.
3548 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3549 extensions, and comparison tests, and to fix some bugs.
3551 14. Changes to the "personal" filter test:
3553 (1) The test was buggy in that it was just doing the equivalent of
3554 "contains" tests on header lines. For example, if a user's address was
3555 anne@some.where, the "personal" test would incorrectly be true for
3557 To: susanne@some.where
3559 This test is now done by extracting each address from the header in turn,
3560 and checking the entire address. Other tests that are part of "personal"
3561 are now done using regular expressions (for example, to check local parts
3562 of addresses in From: header lines).
3564 (2) The list of non-personal local parts in From: addresses has been
3565 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3566 taken from the Sieve specification recommendations.
3568 (3) If the message contains any header line starting with "List-" it is
3569 treated as non-personal.
3571 (4) The test for "circular" in the Subject: header line has been removed
3572 because it now seems ill-conceived.
3574 15. Minor typos in src/EDITME comments corrected.
3576 16. Installed latest exipick from John Jetmore.
3578 17. If headers_add on a router specified a text string that was too long for
3579 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3580 of string_sprintf() is now avoided.
3582 18. $message_body_size was not set (it was always zero) when running the DATA
3583 ACL and the local_scan() function.
3585 19. For the "mail" command in an Exim filter, no default was being set for
3586 the once_repeat time, causing a random time value to be used if "once" was
3587 specified. (If the value happened to be <= 0, no repeat happened.) The
3588 default is now 0s, meaning "never repeat". The "vacation" command was OK
3589 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3590 (I found it when inspecting the code).
3592 20. There is now an overall timeout for performing a callout verification. It
3593 defaults to 4 times the callout timeout, which applies to individual SMTP
3594 commands during the callout. The overall timeout applies when there is more
3595 than one host that can be tried. The timeout is checked before trying the
3596 next host. This prevents very long delays if there are a large number of
3597 hosts and all are timing out (e.g. when the network connections are timing
3598 out). The value of the overall timeout can be changed by specifying an
3599 additional sub-option for "callout", called "maxwait". For example:
3601 verify = sender/callout=5s,maxwait=20s
3603 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3604 to the end before writing, but this should make it even safer).
3606 22. Exim was forgetting that it had advertised PIPELINING for the second and
3607 subsequent messages on an SMTP connection. It was also not resetting its
3608 memory on STARTTLS and an internal HELO.
3610 23. When Exim logs an SMTP synchronization error within a session, it now
3611 records whether PIPELINING has been advertised or not.
3613 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3614 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3615 rather than long int.
3617 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3619 26. Added the never_mail option to autoreply.
3625 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3626 crash if the getsockname() call failed; this can happen if a connection is
3627 closed very soon after it is established. The problem was simply in the
3628 order in which certain operations were done, causing Exim to try to write
3629 to the SMTP stream before it had set up the file descriptor. The bug has
3630 been fixed by making things happen in the correct order.
3636 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3637 before the connection was closed, thus losing the rejection response.
3639 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3640 some early Solaris releases, but causes trouble in current releases where
3641 socklen_t is defined.
3643 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3646 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3649 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3650 that releasing the top of it at the end releases what was used for sub-
3651 expansions (unless the block got too big). However, discard this block if
3652 the first thing is a variable or header, so that we can use its block when
3653 it is dynamic (useful for very large $message_headers, for example).
3655 6. Lookups now cache *every* query, not just the most recent. A new, separate
3656 store pool is used for this. It can be recovered when all lookup caches are
3657 flushed. Lookups now release memory at the end of their result strings.
3658 This has involved some general refactoring of the lookup sources.
3660 7. Some code has been added to the store_xxx() functions to reduce the amount
3661 of flapping under certain conditions.
3663 8. log_incoming_interface used to affect only the <= reception log lines. Now
3664 it causes the local interface and port to be added to several more SMTP log
3665 lines, for example "SMTP connection from", and rejection lines.
3667 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3669 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3671 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3672 could be overwritten at the end of the current message (or the start of a
3673 new message if it was set in a HELO ACL). The value is now preserved for
3674 the duration of the SMTP connection.
3676 12. If a transport had a headers_rewrite setting, and a matching header line
3677 contained an unqualified address, that address was qualified, even if it
3678 did not match any rewriting rules. The underlying bug was that the values
3679 of the flags that permit the existence of unqualified sender and recipient
3680 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3681 non-local messages, and by -bnq for local messages) were not being
3682 preserved with the message after it was received.
3684 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3685 "next input=" as part of the text comprising the host identity instead of
3686 the correct text. The code was using the same buffer for two different
3687 strings. However, depending on which order the printing function evaluated
3688 its arguments, the bug did not always show up. Under Linux, for example, my
3689 test suite worked just fine.
3691 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3692 doesn't work with some older versions of Perl. It has been changed to "my",
3693 which in any case is probably the better facility to use.
3695 15. A really picky compiler found some instances of statements for creating
3696 error messages that either had too many or two few arguments for the format
3699 16. The size of the buffer for calls to the DNS resolver has been increased
3700 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3701 for addresses that have a lot of PTR records. This alleviates a problem; it
3702 does not fully solve it.
3704 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3705 buffer now truncates the list and logs the incident, which is the same
3706 action as happens when Exim is looking up a host name and its aliases.
3707 Previously in this situation something unpredictable would happen;
3708 sometimes it was "internal error: store_reset failed".
3710 18. If a server dropped the connection unexpectedly when an Exim client was
3711 using GnuTLS and trying to read a response, the client delivery process
3712 crashed while trying to generate an error log message.
3714 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3715 string, for example:
3717 warn message = H1: something\nH2: something
3719 the text was added as a single header line from Exim's point of view
3720 though it ended up OK in the delivered message. However, searching for the
3721 second and subsequent header lines using $h_h2: did not work. This has been
3722 fixed. Similarly, if a system filter added multiple headers in this way,
3723 the routers could not see them.
3725 20. Expanded the error message when iplsearch is called with an invalid key to
3726 suggest using net-iplsearch in a host list.
3728 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3729 ACLs are no longer actually imposed (and a message to that effect is
3732 22. If a "gecos" field in a passwd entry contained escaped characters, in
3733 particular, if it contained a \" sequence, Exim got it wrong when building
3734 a From: or a Sender: header from that name. A second bug also caused
3735 incorrect handling when an unquoted " was present following a character
3736 that needed quoting.
3738 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3739 was not being matched caselessly.
3741 24. Arranged for all hyphens in the exim.8 source to be escaped with
3744 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3745 properly. Recipient callout cache records were still being keyed to include
3746 the sender, even when use_sender was set false. This led to far more
3747 callouts that were necessary. The sender is no longer included in the key
3748 when use_sender is false.
3750 26. Added "control = submission" modifier to ACLs.
3752 27. Added the ${base62d: operator to decode base 62 numbers.
3754 28. dnsdb lookups can now access SRV records.
3756 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3757 the configuration file.
3759 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3760 (-v) mode. This makes the output for a verbose queue run more intelligible.
3762 31. Added a use_postmaster feature to recipient callouts.
3764 32. Added the $body_zerocount variable, containing the number of binary zero
3765 bytes in the message body.
3767 33. The time of last modification of the "new" subdirectory is now used as the
3768 "mailbox time last read" when there is a quota error for a maildir
3771 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3773 35. Added +ignore_unknown as a special item in host lists.
3775 36. Code for decoding IPv6 addresses in host lists is now included, even if
3776 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3777 address was recognized as an IP address, but was then not correctly decoded
3778 into binary, causing unexpected and incorrect effects when compared with
3785 1. Very minor rewording of debugging text in manualroute to say "list of
3786 hosts" instead of "hostlist".
3788 2. If verify=header_syntax was set, and a header line with an unqualified
3789 address (no domain) and a large number of spaces between the end of the
3790 name and the colon was received, the reception process suffered a buffer
3791 overflow, and (when I tested it) crashed. This was caused by some obsolete
3792 code that should have been removed. The fix is to remove it!
3794 3. When running in the test harness, delay a bit after writing a bounce
3795 message to get a bit more predictability in the log output.
3797 4. Added a call to search_tidyup() just before forking a reception process. In
3798 theory, someone could use a lookup in the expansion of smtp_accept_max_
3799 per_host which, without the tidyup, could leave open a database connection.
3801 5. Added the variables $recipient_data and $sender_data which get set from a
3802 lookup success in an ACL "recipients" or "senders" condition, or a router
3803 "senders" option, similar to $domain_data and $local_part_data.
3805 6. Moved the writing of debug_print from before to after the "senders" test
3808 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3809 problems for message scanning, either using a data ACL, or using
3810 local_scan() because the Received: header was not generated till after they
3811 were called (in order to set the time as the time of reception completion).
3812 I have revised the way this works. The header is now generated after the
3813 body is received, but before the ACL or local_scan() are called. After they
3814 are run, the timestamp in the header is updated.
3820 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3821 before starting a queue runner without re-exec. This happened only when
3822 deliver_drop_privilege was set or when the Exim user was set to root. The
3823 effect of the bug was that timeouts during subsequent deliveries caused
3824 crashes instead of being properly handled. The handler is now left at its
3825 default (and expected) setting.
3827 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3828 message, again when deliver_drop_privilege is set or Exim is run as root.
3829 The bug described in (1) was not present in this case, but the tidying up
3830 of the other signals was missing. I have made the two cases consistent.
3832 3. The ignore_target_hosts setting on a manualroute router was being ignored
3833 for hosts that were looked up using the /MX notation.
3835 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3838 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3839 operated on the sender address. After changing the $sender_address to <>
3840 for the sender address verify, Exim was re-instated it as the original
3841 (before rewriting) address, but remembering that it had rewritten it, so it
3842 wasn't rewriting it again. This bug also had the effect of breaking the
3843 sender address verification caching when the sender address was rewritten.
3845 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3846 This has been changed so that if the ip literal address matches
3847 ignore_target_hosts, the router declines.
3849 7. Added expansion conditions match_domain, match_address, and match_local_
3850 part (NOT match_host).
3852 8. The placeholder for the Received: header didn't have a length field set.
3854 9. Added code to Exim itself and to exim_lock to test for a specific race
3855 condition that could lead to file corruption when using MBX delivery. The
3856 issue is with the lockfile that is created in /tmp. If this file is removed
3857 after a process has opened it but before that process has acquired a lock,
3858 there is the potential for a second process to recreate the file and also
3859 acquire a lock. This could lead to two Exim processes writing to the file
3860 at the same time. The added code performs the same test as UW imapd; it
3861 checks after acquiring the lock that its file descriptor still refers to
3862 the same named file.
3864 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3865 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3868 11. Added the smtp_active_hostname option. If used, this will typically be made
3869 to depend on the incoming interface address. Because $interface_address is
3870 not set up until the daemon has forked a reception process, error responses
3871 that can happen earlier (such as "too many connections") no longer contain
3874 12. If an expansion in a condition on a "warn" statement fails because a lookup
3875 defers, the "warn" statement is abandoned, and the next ACL statement is
3876 processed. Previously this caused the whole ACL to be aborted.
3878 13. Added the iplsearch lookup type.
3880 14. Added ident_timeout as a log selector.
3882 15. Added tls_certificate_verified as a log selector.
3884 16. Added a global option tls_require_ciphers (compare the smtp transport
3885 option of the same name). This controls incoming TLS connections.
3887 17. I finally figured out how to make tls_require_ciphers do a similar thing
3888 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3889 before starting the TLS session.
3891 18. Tabs are now shown as \t in -bP output.
3893 19. If the log selector return_path_on_delivery was set, Exim crashed when
3894 bouncing a message because it had too many Received: header lines.
3896 20. If two routers both had headers_remove settings, and the first one included
3897 a superfluous trailing colon, the final name in the first list and the
3898 first name in the second list were incorrectly joined into one item (with a
3899 colon in the middle).
3905 1. Added -C and -D options to the exinext utility, mainly to make it easier
3906 to include in the automated testing, but these could be helpful when
3907 multiple configurations are in use.
3909 2. The exinext utility was not formatting the output nicely when there was
3910 an alternate port involved in the retry record key, nor when there was a
3911 message id as well (for retries that were specific to a specific message
3912 and a specific host). It was also confused by IPv6 addresses, because of
3913 the additional colons they contain. I have fixed the IPv4 problem, and
3914 patched it up to do a reasonable job for IPv6.
3916 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3917 delivery, the log line now contains "pipelined" if PIPELINING was used.
3919 4. An SMTP transport process used to panic and die if the bind() call to set
3920 an explicit outgoing interface failed. This has been changed; it is now
3921 treated in the same way as a connect() failure.
3923 5. A reference to $sender_host_name in the part of a conditional expansion
3924 that was being skipped was still causing a DNS lookup. This no longer
3927 6. The def: expansion condition was not recognizing references to header lines
3928 that used bh_ and bheader_.
3930 7. Added the _cache feature to named lists.
3932 8. The code for checking quota_filecount in the appendfile transport was
3933 allowing one more file than it should have been.
3935 9. For compatibility with Sendmail, the command line option
3943 and sets the incoming protocol and host name (for trusted callers). The
3944 host name and its colon can be omitted when only the protocol is to be set.
3945 Note the Exim already has two private options, -pd and -ps, that refer to
3946 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3947 "s", but I don't think that's a major issue.
3949 10. A number of refactoring changes to the code, none of which should affect
3952 (a) The number of logging options was getting close to filling up the
3953 32-bit word that was used as a bit map. I have split them into two classes:
3954 those that are passed in the argument to log_write(), and those that are
3955 only ever tested independently outside of that function. These are now in
3956 separate 32-bit words, so there is plenty of room for expansion again.
3957 There is no change in the user interface or the logging behaviour.
3959 (b) When building, for example, log lines, the code previously used a
3960 macro that called string_cat() twice, in order to add two strings. This is
3961 not really sufficiently general. Furthermore, there was one instance where
3962 it was actually wrong because one of the argument was used twice, and in
3963 one call a function was used. (As it happened, calling the function twice
3964 did not affect the overall behaviour.) The macro has been replaced by a
3965 function that can join an arbitrary number of extra strings onto a growing
3968 (c) The code for expansion conditions now uses a table and a binary chop
3969 instead of a serial search (which was left over from when there were very
3970 few conditions). Also, it now recognizes conditions like "pam" even when
3971 the relevant support is not compiled in: a suitably worded error message is
3972 given if an attempt is made to use such a condition.
3974 11. Added ${time_interval:xxxxx}.
3976 12. A bug was causing one of the ddress fields not to be passed back correctly
3977 from remote delivery subprocesses. The field in question was not being
3978 subsequently used, so this caused to problems in practice.
3980 13. Added new log selectors queue_time and deliver_time.
3982 14. Might have fixed a bug in maildirsizefile handling that threw up
3983 "unexpected character" debug warnings, and recalculated the data
3984 unnecessarily. In any case, I expanded the warning message to give more
3987 15. Added the message "Restricted characters in address" to the statements in
3988 the default ACL that block characters like @ and % in local parts.
3990 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3991 Three changes have been made:
3993 (a) There was a serious bug; a negative response to MAIL caused the whole
3994 recipient domain to be cached as invalid, thereby blocking all messages
3995 to all local parts at the same domain, from all senders. This bug has
3996 been fixed. The domain is no longer cached after a negative response to
3997 MAIL if the sender used is not empty.
3999 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4002 (c) A new callout option, "use_sender" has been added for people who want
4003 the modified behaviour.
4009 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4012 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4013 indeed breaks things for older releases.
4015 3. Added additional logging to the case where there is a problem reading data
4016 from a filter that is running in a subprocess using a pipe, in order to
4017 try to track down a specific problem.
4019 4. Testing facility fudge: when running in the test harness and attempting
4020 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4021 getting "No route to host". Convert this to a timeout.
4023 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4026 6. Some OS don't have socklen_t but use size_t instead. This affects the
4027 fifth argument of getsockopt() amongst other things. This is now
4028 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4029 can be set for individual OS. I have set it for SunOS5, OSF1, and
4030 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4031 some earlier ones do not.
4033 7. Change 4.30/15 was not doing the test caselessly.
4035 8. The standard form for an IPv6 address literal was being rejected by address
4036 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4037 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4038 this, as well as the form without the "IPv6" on the front (but only when
4039 address literals are enabled, of course).
4041 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4043 10. Exim crashed if a message with an empty sender address specified by -f
4044 encountered a router with an errors_to setting. This could be provoked only
4045 by a command such as
4049 where an empty string was supplied; "<>" did not hit this bug.
4051 11. Installed PCRE release 4.5.
4053 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4054 remained set. It is now erased.
4056 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4057 times from message ids (which are base 36 rather than the normal 62).
4059 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4060 were being counted as actual protocol errors, and logged if the log
4061 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4062 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4063 and DATA following a set of rejected RCPTs do not count as protocol errors.
4064 In other words, Exim assumes they were pipelined, though this may not
4065 actually be the case. Of course, in all cases the client gets an
4066 appropriate error code.
4068 15. If a lookup fails in an ACL condition, a message about the failure may
4069 be available; it is used if testing the ACL cannot continue, because most
4070 such messages specify what the cause of the deferral is. However, some
4071 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4072 that caused an old message to be retained and used if a later statement
4073 caused a defer, replacing the real cause of the deferral.
4075 16. If an IP address had so many PTR records that the DNS lookup buffer
4076 was not large enough to hold them, Exim could crash while trying to process
4077 the truncated data. It now detects and logs this case.
4079 17. Further to 4.21/58, another change has been made: if (and only if) the
4080 first line of a message (the first header line) ends with CRLF, a bare LF
4081 in a subsequent header line has a space inserted after it, so as not to
4082 terminate the header.
4084 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4085 unnecessarily, used atoi() instead of strtol(), and didn't check the
4086 termination when getting file sizes from file names by regex.
4088 19. Completely re-implemented the support for maildirsize files, in the light
4089 of a number of problems with the previous contributed implementation
4090 (4.30/29). In particular:
4092 . If the quota is zero, the maildirsize file is maintained, but no quota is
4095 . If the maildir directory does not exist, it is created before any attempt
4096 to write a maildirsize file.
4098 . The quota value in the file is just a cache; if the quota is changed in
4099 the transport, the new value overrides.
4101 . A regular expression is available for excluding directories from the
4104 20. The autoreply transport checks the characters in options that define the
4105 message's headers; it allows continued headers, but it was checking with
4106 isspace() after an embedded newline instead of explicitly looking for a
4109 21. If all the "regular" hosts to which an address was routed had passed their
4110 expiry times, and had not reached their retry times, the address was
4111 bounced, even if fallback hosts were defined. Now Exim should go on to try
4114 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4115 equivalent code in the SMTP transport. Some hosts send humungous responses
4116 to HELO/EHLO, more than 1024 it seems.
4118 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4119 gives compiler warnings in some environments. I've now done it "properly",
4122 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4123 (because of problems with the built-in one) was declared to return uschar *
4124 instead of char *, causing compiler failure.
4126 25. Fixed a file descriptor leak when processing alias/forward files.
4128 26. Fixed a minor format string issue in dbfn.c.
4130 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4132 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4133 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4136 29. When a sender address is verified, it is cached, to save repeating the test
4137 when there is more than one recipient in a message. However, when the
4138 verification involves a callout, it is possible for different callout
4139 options to be set for different recipients. It is too complicated to keep
4140 track of this in the cache, so now Exim always runs a verification when a
4141 callout is required, relying on the callout cache for the optimization.
4142 The overhead is duplication of the address routing, but this should not be
4145 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4146 to be verified with callout=postmaster, and the main callout worked but the
4147 postmaster check failed, the verification correctly failed. However, if a
4148 subsequent RCPT command asked for sender verification *without* the
4149 postmaster check, incorrect caching caused this verification also to fail,
4152 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4153 it was not caching the DNS options (qualify_single, search_parents) that
4154 were used when the lookup failed. A subsequent lookup with different
4155 options therefore always gave the same answer, though there were cases
4156 where it should not have. (Example: a "domains = !$mx_any" option on a
4157 dnslookup router: the "domains" option is always processed without any
4158 widening, but the router might have qualify_single set.) Now Exim uses the
4159 cached value only when the same options are set.
4161 32. Added John Jetmore's "exipick" utility to the distribution.
4163 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4164 than a timeout (e.g. a certificate is required, and is not provided), an
4165 Exim server now closes the connection immediately. Previously it waited for
4166 the client to close - but if the client is SSL, it seems that they each
4167 wait for each other, leading to a delay before one of them times out.
4169 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4170 maintained 0.8.x compatibility because I don't think many are using it, and
4171 it is clearly obsolete.
4173 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4176 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4177 client certificate was expired. A simple patch fixes this, though I don't
4178 understand the full logic of why the verify callback is called multiple
4181 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4182 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4183 which causes problems with some clients (such as the Certicom SSL Plus
4184 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4185 disables the coutermeasure allowing Eudora to connect."
4187 38. Exim was not checking that a write() to a log file succeeded. This could
4188 lead to Bad Things if a log got too big, in particular if it hit a file
4189 size limit. Exim now panics and dies if it cannot write to a log file, just
4190 as it does if it cannot open a log file.
4192 39. Modified OS/Makefile-Linux so that it now contains
4194 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4196 The two -D definitions ensure that Exim is compiled with large file
4197 support, which makes it possible to handle log files that are bigger than
4200 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4201 instance) a domain was checked against a named list that involved a lookup,
4202 causing $domain_data to be set, then another domain was checked against the
4203 same list, then the first domain was re-checked, the value of $domain_data
4204 after the final check could be wrong. In particular, if the second check
4205 failed, it could be set empty. This bug probably also applied to
4208 41. The strip_trailing_dot option was not being applied to the address given
4209 with the -f command-line option.
4211 42. The code for reading a message's header from the spool was incrementing
4212 $received_count, but never initializing it. This meant that the value was
4213 incorrect (doubled) while delivering a message in the same process in which
4214 it was received. In the most common configuration of Exim, this never
4215 happens - a fresh exec is done - but it can happen when
4216 deliver_drop_privilege is set.
4218 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4219 it now includes up to 150 characters of the unexpected data in the log
4222 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4223 and building data strings. The size of both of these buffers was 10 000
4224 bytes - far larger than anybody would *ever* want, thought I. Needless to
4225 say, somebody hit the limit. I have increased the maximum line length to
4226 20 000 and the maximum data length of concatenated lines to 100 000. I have
4227 also fixed two bugs, because there was no checking on these buffers. Tsk,
4228 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4229 buffer is too small.
4231 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4232 lsearch lookups. Now it does.
4234 46. When parsing a route_list item in a manualroute router, a fixed-length
4235 buffer was used for the list of hosts. I made this 1024 bytes long,
4236 thinking that nobody would ever have a list of hosts that long. Wrong.
4237 Somebody had a whole pile of complicated expansion conditions, and the
4238 string was silently truncated, leading to an expansion error. It turns out
4239 that it is easier to change to an unlimited length (owing to other changes
4240 that have happened since this code was originally written) than to build
4241 structure for giving a limitation error. The length of the item that
4242 expands into the list of hosts is now unlimited.
4244 47. The lsearch lookup could not handle data where the length of text line was
4245 more than 4095 characters. Such lines were truncated, leading to shortened
4246 data being returned. It should now handle lines of any length.
4248 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4249 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4252 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4253 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4254 no longer happens in comment lines. A list of replacements is now placed
4255 at the head of all of the source files, except those whose only change is
4256 to replace PERL_COMMAND in the very first #! line.
4258 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4259 messages on the queue, with a bottom-up merge sort, using code contributed
4260 by Michael Haardt. This should make operations like -bp somewhat faster on
4261 large queues. It won't affect queue runners, except when queue_run_in_order
4264 51. Installed eximstats 1.31 in the distribution.
4266 52. Added support for SRV lookups to the dnslookup router.
4268 53. If an ACL referred to $message_body or $message_body_end, the value was not
4269 reset for any messages that followed in the same SMTP session.
4271 54. The store-handling optimization for building very long strings was not
4272 differentiating between the different store pools. I don't think this
4273 actually made any difference in practice, but I've tidied it.
4275 55. While running the routers to verify a sender address, $sender_address
4276 was still set to the sender address. This is wrong, because when routing to
4277 send a bounce to the sender, it would be empty. Therefore, I have changed
4278 it so that, while verifying a sender address, $sender_address is set to <>.
4279 (There is no change to what happens when verifying a recipient address.)
4281 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4282 target A or AAAA records (if not already returned) without resetting the
4283 qualify_single or search_parents options of the DNS resolver. These are
4284 inappropriate in this case because the targets of MX and SRV records must
4285 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4286 target that, when qualified, matched something in the local domain. These
4287 two options are now turned off when doing these lookups.
4289 57. It seems that at least some releases of Reiserfs (which does not have the
4290 concept of a fixed number of inodes) returns zero and not -1 for the
4291 number of available inodes. This interacted badly with check_spool_inodes,
4292 which assumed that -1 was the "no such thing" setting. What I have done is
4293 to check that the total number of inodes is greater than zero before doing
4294 the test of how many are available.
4296 58. When a "warn" ACL statement has a log_message modifier, the message is
4297 remembered, and not repeated. This is to avoid a lot of repetition when a
4298 message has many recipients that cause the same warning to be written.
4299 Howewer, Exim was preserving the list of already written lines for an
4300 entire SMTP session, which doesn't seem right. The memory is now reset if a
4301 new message is started.
4303 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4304 debugging output unless log_rewrite was also set.
4306 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4307 of (int)(handle) when we know that handle contains (void *)(-1).
4309 61. The Exim daemon panic-logs an error return when it closes the incoming
4310 connection. However "connection reset by peer" seems to be common, and
4311 isn't really an error worthy of noting specially, so that particular error
4314 62. When Exim is trying to find all the local interfaces, it used to panic and
4315 die if the ioctl to get the interface flags failed. However, it seems that
4316 on at least one OS (Solaris 9) it is possible to have an interface that is
4317 included in the list of interfaces, but for which you get a failure error
4318 for this call. This happens when the interface is not "plumbed" into a
4319 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4320 failure of the "get flags" call assumes that the interface is down.
4322 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4323 makes life easier for people who are doing arithmetic on fields extracted
4324 from dates, where you often get leading zeros that should not be
4325 interpreted as octal.
4327 64. Added qualify_domain to the redirect router, to override the global
4330 65. If a pathologically long header line contained very many addresses (the
4331 report of this problem mentioned 10 000) and each of them was rewritten,
4332 Exim could use up a very large amount of memory. (It kept on making new
4333 copies of the header line as it rewrote, and never released the old ones.)
4334 At the expense of a bit more processing, the header rewriting function has
4335 been changed so that it no longer eats memory in this way.
4337 66. The generation of the Received: header has been moved from the time that a
4338 message starts to be received, to the time that it finishes. The timestamp
4339 in the Received: header should now be very close to that of the <= log
4340 line. There are two side-effects of this change:
4342 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4343 logged header lines no longer include the local Received: line, because
4344 it has not yet been created. The same applies to a copy of the message
4345 that is returned to a non-SMTP sender when a message is rejected.
4347 (b) When a filter file is tested using -bf, no additional Received: header
4348 is added to the test message. After some thought, I decided that this
4351 This change does not affect the value of $received_for. It is still set
4352 after address rewriting, but before local_scan() is called.
4354 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4356 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4357 gave an unhelpful panic error message, and a defer error. I have managed to
4358 change this behaviour so that it now rejects any supplied certificate,
4359 which seems right, as the list of acceptable certificates is empty.
4361 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4362 gave an unhelpful defer error. I have not managed to make this reject any
4363 supplied certificates, but the error message it gives is "no certificate
4364 supplied", which is not helpful.
4366 70. exigrep's output now also includes lines that are not associated with any
4367 message, but which match the given pattern. Implemented by a patch from
4368 Martin Sluka, which also tidied up the Perl a bit.
4370 71. Recipient callout verification, like sender verification, was using <> in
4371 the MAIL FROM command. This isn't really the right thing, since the actual
4372 sender may affect whether the remote host accepts the recipient or not. I
4373 have changed it to use the actual sender in the callout; this means that
4374 the cache record is now keyed on a recipient/sender pair, not just the
4375 recipient address. There doesn't seem to be a real danger of callout loops,
4376 since a callout by the remote host to check the sender would use <>.
4377 [SEE ABOVE: changed after hitting problems.]
4379 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4380 temporary errors. However, in the case of such a code being given after
4381 the end of a data transmission (i.e. after ".") Exim was failing to write
4382 a retry record for the message. (Yes, there was some broken host that was
4383 actually sending 8xx at this point.)
4385 73. An unknown lookup type in a host list could cause Exim to panic-die when
4386 the list was checked. (An example that provoked this was putting <; in the
4387 middle of a list instead of at the start.) If this happened during a DATA
4388 ACL check, a -D file could be left lying around. This kind of configuration
4389 error no longer causes Exim to die; instead it causes a defer errror. The
4390 incident is still logged to the main and panic logs.
4392 74. Buglet left over from Exim 3 conversion. The message "too many messages
4393 in one connection" was written to the rejectlog but not the mainlog, except
4394 when address rewriting (yes!) was being logged.
4396 75. Added write_rejectlog option.
4398 76. When a system filter was run not as root (that is, when system_filter_user
4399 was set), the values of the $n variables were not being returned to the
4400 main process; thus, they were not subsequently available in the $sn
4403 77. Added +return_path_on_delivery log selector.
4405 78. A connection timeout was being treated differently from recipients deferred
4406 when testing hosts_max_try with a message that was older than the host's
4407 retry timeout. (The host should not be counted, thus allowing all hosts to
4408 be tried at least once before bouncing.) This may have been the cause of an
4409 occasionally reported bug whereby a message would remain on the queue
4410 longer than the retry timeout, but would be bounced if a delivery was
4411 forced. I say "may" because I never totally pinned down the problem;
4412 setting up timeout/retry tests is difficult. See also the next item.
4414 79. The ultimate address timeout was not being applied to errors that involved
4415 a combination of host plus message (for example, a timeout on a MAIL
4416 command). When an address resolved to a number of possible hosts, and they
4417 were not all tried for each delivery (e.g. because of hosts_max_try), a
4418 message could remain on the queue longer than the retry timeout.
4420 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4423 81. Fixed an obscure SMTP outgoing bug which required at least the following
4424 conditions: (a) there was another message waiting for the same server;
4425 (b) the server returned 5xx to all RCPT commands in the first message so
4426 that the message was not completed; (c) the server dropped the connection
4427 or gave a negative response to the RSET that Exim sends to abort the
4428 transaction. The observed case was a dropped connection after DATA that had
4429 been sent in pipelining mode. That is, the server had advertised PIPELINING
4430 but was not implementing it correctly. The effect of the bug was incorrect
4431 behaviour, such as trying another host, and this could lead to a crash.
4437 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4438 and daemon.c were passed as pointers to ints; they should have been
4439 pointers to socklen_t variables (which are typically unsigned ints).
4441 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4444 3. Fixed a really odd bug that affected only the testing scheme; patching a
4445 certain fixed string in the binary changed the value of another string that
4446 happened to be identical to the end of the original first string.
4448 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4449 name", it returns that address as the IP address. On some operating
4450 systems (e.g. Solaris), it also passes back the IP address string as the
4451 "host name". However, on others (e.g. Linux), it passes back an empty
4452 string. Exim wasn't checking for this, and was changing the host name to an
4453 empty string, assuming it had been canonicized.
4455 5. Although rare, it is permitted to have more than one PTR record for a given
4456 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4457 all the names associated with an address, because they do in Solaris.
4458 However, it seems that they do not in Linux for data that comes from the
4459 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4460 I found this out when I moved to a new Linux workstation and tried to run
4461 the Exim test suite.
4463 To get round this problem I have changed the code so that it now does its
4464 own call to the DNS to look up PTR records when searching for a host name.
4465 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4466 addresses that are only in /etc/hosts are still found.
4468 This behaviour is, however, controlled by an option called host_lookup_
4469 order, which defaults to "bydns:byaddr". If people want to use the other
4470 order, or indeed, just use one or the other means of lookup, they can
4471 specify it in this variable.
4473 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4474 some operating systems, this comes back from gethostbyaddr() as an empty
4475 string, and this is what Exim used to test for. However, it seems that in
4476 other systems, "." is yielded. Exim now tests for this case too.
4478 7. The values of check_spool_space and check_log_space are now held internally
4479 as a number of kilobytes instead of an absolute number of bytes. If a
4480 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4481 kilobyte. This means that much larger values can be stored.
4483 8. Exim monitor: an attempt to get the action menu when not actually pointing
4484 at a message produces an empty menu entitled "No message selected". This
4485 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4486 no entries in it ("Shell widget menu has zero width and/or height"). So I
4487 have added a single, blank menu entry in this case.
4489 9. Added ${quote_local_part.
4491 10. MIME decoding is now applied to the contents of Subject: header lines when
4494 11. Now that a reference to $sender_host_address automatically causes a reverse
4495 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4496 host lookup before query-style lookups in lists that might use this
4497 variable. This has therefore been abolished, and the "net-" prefix is no
4498 longer necessary for query-style lookups.
4500 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4501 have been a typo for LFLAGS, so it has been changed.
4503 13. The install script calls Exim with "-C /dev/null" in order to find the
4504 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4505 to be output. Howeve, since Exim outputs its version number before the
4506 error, it didn't break the script. It just looked ugly. I fixed this by
4507 always allowing "-C /dev/null" if the caller is root.
4509 14. Ignore overlarge ACL variable number when reading spool file - insurance
4510 against a later release with more variables having written the file.
4512 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4513 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4514 this, as well as the form without the "IPv6" on the front.
4516 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4517 OS/Makefile-Darwin file.
4519 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4520 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4522 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4523 variable. However, some OS use ints for inodes. I've added cast to long int
4524 to get rid of the compiler warning.
4526 19. I had forgotten to lock out "/../" in configuration file names when
4527 ALT_CONFIG_PREFIX was set.
4529 20. Routers used for verification do not need to specify transports. However,
4530 if such a router generated a host list, and callout was configured, Exim
4531 crashed, because it could not find a port number from the (non-existent)
4532 transport. It now assumes port 25 in this circumstance.
4534 21. Added the -t option to exigrep.
4536 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4537 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4538 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4539 them is set, LOOKUP_LSEARCH is forced.
4541 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4543 24. Added sender and host information to the "rejected by local_scan()" log
4544 line; previously there was no indication of these.
4546 25. Added .include_if_exists.
4548 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4549 Linux. It turns out that not all file systems support this. Apparently some
4550 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4551 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4552 not supported on the file descriptor, is now ignored when Exim is trying to
4553 sync a directory. This applies only to Linux.
4555 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4557 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4558 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4561 29. Added support for maildirsize files from supplied patch (modified a bit).
4563 30. The use of :fail: followed by an empty string could lead Exim to respond to
4564 sender verification failures with (e.g.):
4566 550 Verification failed for <xxx>
4567 550 Sender verify failed
4569 where the first response line was missing the '-' that indicates it is not
4570 the final line of the response.
4572 31. The loop for finding the name of the user that called Exim had a hardwired
4573 limit of 10; it now uses the value of finduser_retries, which is used for
4574 all other user lookups.
4576 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4579 33. Exim was neglecting to zero errno before one call of strtol() when
4580 expanding a string and expecting an integer value. On some systems this
4581 resulted in spurious "integer overflow" errors. Also, it was casting the
4582 result into an int without checking.
4584 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4585 did not work. The code looks as if it has *never* worked, though it appears
4586 to have been documented since at least releast 1.62. I have made it work.
4588 35. The "timeout_DNS" error in retry rules, also documented since at least
4589 1.62, also never worked. As it isn't clear exactly what this means, and
4590 clearly it isn't a major issue, I have abolished the feature by treating it
4591 as "timeout", and writing a warning to the main and panic logs.
4593 36. The display of retry rules for -brt wasn't always showing the error code
4596 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4597 timeout_connect_A, timeout_connect_MX.
4599 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4600 to the empty sender.
4602 39. The daemon was not analysing the content of -oX till after it had closed
4603 stderr and disconnected from the controlling terminal. This meant that any
4604 syntax errors were only noted on the panic log, and the return code from
4605 the command was 0. By re-arranging the code a little, I've made the
4606 decoding happen first, so such errors now appear on stderr, and the return
4607 code is 1. However, the actual setting up of the sockets still happens in
4608 the disconnected process, so errors there are still only recorded on the
4611 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4612 connections (as happens on some IP stacks) was logged at start up time as
4613 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4614 it from "IPv6 and IPv4", which means that two separate sockets are being
4617 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4618 says whether AF_INET or AF_INET6 was passed as an argument.
4620 42. Exiwhat output was messed up when time zones were included in log
4623 43. Exiwhat now gives more information about the daemon's listening ports,
4624 and whether -tls-on-connect was used.
4626 44. The "port" option of the smtp transport is now expanded.
4628 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4629 silently ignored. Now an error message is written to the main and panic
4632 46. There's a new ACL modifier called "logwrite" which writes to a log file
4633 as soon as it is encountered.
4635 47. Added $local_user_uid and $local_user_gid at routing time.
4637 48. Exim crashed when trying to verify a sender address that was being
4640 49. Exim was recognizing only a space character after ".include". It now also
4641 recognizes a tab character.
4643 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4644 extracting the relevant information from the specification. The man page no
4645 longer contains scrambled data for the -d option, and I've added a section
4646 at the front about calling Exim under different names.
4648 51. Added "extra_headers" argument to the "mail" command in filter files.
4650 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4653 53. Installed eximstats 1.29.
4655 54. Added transport_filter_timeout as a generic transport option.
4657 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4658 Cc: header lines. This was required by RFC 822, but it not required by RFC
4661 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4662 incoming messages that did not have them. Now it does so only if the
4663 message originates locally, that is, if there is no associated remote host
4664 address. When Resent- header lines are present, this applies to the Resent-
4665 lines rather than the non-Resent- lines.
4667 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4668 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4670 58. Messages for configuration errors now include the name of the main
4671 configuration file - useful now that there may be more than one file in a
4672 list (.included file names were always shown).
4674 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4675 for those rare installations that do not start the daemon as root or run it
4676 setuid root. I've cut out the call to initgroups() if the daemon is not
4679 60. The Exim user and group can now be bound into the binary as text strings
4680 that are looked up at the start of Exim's processing.
4682 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4684 62. Added $mailstore_basename variable.
4686 63. Installed patch to sieve.c from Michael Haardt.
4688 64. When Exim failed to open the panic log after failing to open the main log,
4689 the original message it was trying to log was written to stderr and debug
4690 output, but if they were not available (the usual case in production), it
4691 was lost. Now it is written to syslog before the two lines that record the
4692 failures to open the logs.
4694 65. Users' Exim filters run in subprocesses under the user's uid. It is
4695 possible for a "deliver" command or an alias in a "personal" command to
4696 provoke an address rewrite. If logging of address rewriting is configured,
4697 this fails because the process is not running as root or exim. There may be
4698 a better way of dealing with this, but for the moment (because 4.30 needs
4699 to be released), I have disabled address rewrite logging when running a
4700 filter in a non-root, non-exim process.
4706 1. The buildconfig auxiliary program wasn't quoting the value set for
4707 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4708 not defined. This bug was masked in 4.22 by the effect that was fixed in
4711 2. Some messages that were rejected after a message id was allocated were
4712 shown as "incomplete" by exigrep. It no longer does this for messages that
4713 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4715 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4716 have allow_domain_literals set, the ID did not get logged in the <= line.
4717 Domain literals are now always recognized in Message-ID: header lines.
4719 4. The first argument for a ${extract expansion item is the key name or field
4720 number. Leading and trailing spaces in this item were not being ignored,
4721 causing some misleading effects.
4723 5. When deliver_drop_privilege was set, single queue runner processes started
4724 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4725 same command in the process it spins off) were not dropping privilege.
4727 6. When the daemon running as "exim" started a queue runner, it always
4728 re-executed Exim in the spun-off process. This is a waste of effort when
4729 deliver_drop_privilege is set. The new process now just calls the
4730 queue-runner function directly.
4736 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4739 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4740 ignored. Though the use of -oP was forcing the writing of a pid file, it
4741 was always written to the default place.
4743 3. If the message "no IP address found for host xxxx" is generated during
4744 incoming verification, it is now followed by identification of the incoming
4745 connection (so you can more easily find what provoked it).
4747 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4749 5. Added some features to "harden" Exim a bit more against certain attacks:
4751 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4752 be put in Local/Makefile. This is like the never_users runtime option,
4753 but it cannot be overridden. The default setting is "root".
4755 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4756 prefix string with which any file named in a -C command line option
4759 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4760 is retained for -C and -D only if the caller of Exim is root. Without
4761 it, the exim user may also use -C and -D and retain privilege.
4763 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4764 command line option is disabled.
4766 6. Macro names set by the -D option must start with an upper case letter, just
4767 like macro names defined in the configuration file.
4769 7. Added "dereference=" facility to LDAP.
4771 8. Two instances of the typo "uknown" in the source files are fixed.
4773 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4774 the Configure-Makefile script screwed up while processing it.
4776 10. Incorporated PCRE 4.4.
4778 11. The SMTP synchronization check was not operating right at the start of an
4779 SMTP session. For example, it could not catch a HELO sent before the client
4780 waited for the greeting. There is now a check for outstanding input at the
4781 point when the greeting is written. Because of the duplex, asynchronous
4782 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4783 way, but not yet received, when the check is performed.
4785 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4786 on TCP/IP sockets, because this apparently causes some broken clients to
4789 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4790 unchanged) from the Cygwin maintainer.
4792 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4793 was supported instead of testing for mbx. Effectively a typo.
4795 15. The spa authenticator server code was not checking that the input it
4796 received was valid base64.
4798 16. The debug output line for the "set" modifier in ACLs was not showing the
4799 name of the variable that was being set.
4801 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4803 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4804 lookup on its own. Something else was needed to trigger the lookup. For
4805 example, a match in host_lookup or the need for a host name in a host list.
4806 Now, if $sender_host_name is referenced and the host name has not yet been
4807 looked up, a lookup is performed. If the lookup fails, the variable remains
4808 empty, and $host_lookup_failed is set to "1".
4810 19. Added "eqi" as a case-independent comparison operator.
4812 20. The saslauthd authentication condition could segfault if neither service
4813 nor realm was specified.
4815 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4816 error message that was logged was misleading, and incoming SMTP
4817 connections were dropped. The message is now more accurate, and temporary
4818 errors are given to SMTP connections.
4820 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4821 (except RSET) with a 421 error, until QUIT is received. However, it was
4822 failing to send a response to QUIT.
4824 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4825 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4826 ACL is now run after the helo_try_verify_hosts code.
4828 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4829 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4830 case-independent, so other case variants are also recognized). Apparently
4831 some systems use these upper case variants.
4833 25. If more than two messages were waiting for the same host, and a transport
4834 filter was specified for the transport, Exim sent two messages over the
4835 same TCP/IP connection, and then failed with "socket operation on non-
4836 socket" when it tried to send the third.
4838 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4840 27. The extern definition of crypt16() in expand.c was not being excluded when
4841 the OS had its own crypt16() function.
4843 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4844 as a preferred synonym for return_size_limit, both as an option and as an
4847 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4849 30. Changed the default configuration ACL to relax the local part checking rule
4850 for addresses that are not in any local domains. For these addresses,
4851 slashes and pipe symbols are allowed within local parts, but the sequence
4852 /../ is explicitly forbidden.
4854 31. SPA server authentication was not clearing the challenge buffer before
4857 32. log_message in a "warn" ACL statement was writing to the reject log as
4858 well as to the main log, which contradicts the documentation and doesn't
4859 seem right (because no rejection is happening). So I have stopped it.
4861 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4862 However, I am unable to do any testing of this.
4864 34. Fixed an infelicity in the appendfile transport. When checking directories
4865 for a mailbox, to see if any needed to be created, it was accidentally
4866 using path names with one or more superfluous leading slashes; tracing
4867 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4869 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4870 contents are added to the log line that is written for every discarded
4871 recipient. (Previously a log_message setting was ignored.)
4873 36. The ${quote: operator now quotes the string if it is empty.
4875 37. The install script runs exim in order to find its version number. If for
4876 some reason other than non-existence or emptiness, which it checks, it
4877 could not run './exim', it was installing it with an empty version number,
4878 i.e. as "exim-". This error state is now caught, and the installation is
4881 38. An argument was missing from the function that creates an error message
4882 when Exim fails to connect to the socket for saslauthd authentication.
4883 This could cause Exim to crash, or give a corrupted message.
4885 39. Added isip, isip4, and isip6 to ${if conditions.
4887 40. The ACL variables $acl_xx are now saved with the message, and can be
4888 accessed later in routers, transports, and filters.
4890 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4891 strings in the file are not string-expanded.
4893 42. If a MAIL command specified a SIZE value that was too large to fit into an
4894 int variable, the check against message_size_limit failed. Such values are
4895 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4896 day this will have to be increased, but I don't think I want to be around
4897 when emails are that large.
4904 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4905 iconv() is not standard in FreeBSD.
4907 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4908 IPv6 enabled. The observed symptom was a segmentation fault on return from
4909 the function os_common_find_running_interfaces() in src/os.c.
4911 3. In the check_special_case() function in daemon.c I had used "errno" as an
4912 argument name, which causes warnings on some systems. This was basically a
4913 typo, since it was named "eno" in the comments!
4915 4. The code that waits for the clock to tick (at a resolution of some fraction
4916 of a second) so as to ensure message-id uniqueness was always waiting for
4917 at least one whole tick, when it could have waited for less. [This is
4918 almost certainly not relevant at current processor speeds, where it is
4919 unlikely to ever wait at all. But we try to future-proof.]
4921 5. The function that sleeps for a time interval that includes fractions of a
4922 second contained a race. It did not block SIGALRM between setting the
4923 timer, and suspending (a couple of lines later). If the interval was short
4924 and the sigsuspend() was delayed until after it had expired, the suspension
4925 never ended. On busy systems this could lead to processes getting stuck for
4928 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4929 process, before it forks any delivery processes. The open lookup caching
4930 mechanism meant that the open file or database connection was passed into
4931 the delivery process. The problem was that delivery processes always tidy
4932 up cached lookup data. This could cause a problem for the next delivery
4933 process started by the queue runner, because the external queue runner
4934 process does not know about the closure. So the next delivery process
4935 still has data in the lookup cache. In the case of a file lookup, there was
4936 no problem because closing a file descriptor in a subprocess doesn't affect
4937 the parent. However, if the lookup was caching a connection to a database,
4938 the connection was closed, and the second delivery process was likely to
4939 see errors such as "PGSQL: query failed: server closed the connection
4940 unexpectedly". The problem has been fixed by closing all cached lookups
4941 in a queue runner before running a delivery process.
4943 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4944 seem to have the "const" qualifier which it has on other OS. I've
4947 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4948 *of the same type* (client or server) with the same public name that an
4949 error should be diagnosed.
4951 9. When Exim looked up a host name for an IP address, but failed to find the
4952 original IP address when looking up the host name (a safety check), it
4953 output the message "<ip address> does not match any IP for NULL", which was
4954 confusing, to say the least. The bug was that the host name should have
4955 appeared instead of "NULL".
4957 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4958 user that is built into the binary, and the -C or -D options is used, root
4959 privilege is dropped before the configuration file is read. In addition,
4960 logging is switched to stderr instead of the normal log files. If the
4961 configuration then re-defines the Exim user, the unprivileged environment
4962 is probably not what is expected, so Exim logs a panic warning message (but
4965 However, if deliver_drop_privilege is set, the unprivileged state may well
4966 be exactly what is intended, so the warning has been cut out in that case,
4967 and Exim is allowed to try to write to its normal log files.
4973 1. smtp_return_error_details was not giving details for temporary sender
4974 or receiver verification errors.
4976 2. Diagnose a configuration error if two authenticators have the same public
4979 3. Exim used not to create the message log file for a message until the first
4980 delivery attempt. This could be confusing when incoming messages were held
4981 for policy or load reasons. The message log file is now created at the time
4982 the message is received, and an initial "Received" line is written to it.
4984 4. The automatically generated man page for command line options had a minor
4985 bug that caused no ill effects; however, a more serious problem was that
4986 the procedure for building the man page automatically didn't always
4987 operate. Consequently, release 4.20 contains an out-of-date version. This
4988 shouldn't happen again.
4990 5. When building Exim with embedded Perl support, the script that builds the
4991 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4992 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4994 6. The freeze_tell option was not being used for messages that were frozen on
4995 arrival, either by an ACL or by local_scan().
4997 7. Added the smtp_incomplete_transaction log selector.
4999 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5000 was accepting AUTH without a new EHLO.
5002 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5003 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5006 10. Logging of TCP/IP connections (when configured) now happens in the main
5007 daemon process instead of the child process, so that the TCP/IP connection
5008 count is more accurate (but it can never be perfect).
5010 11. The use of "drop" in a nested ACL was not being handled correctly in the
5011 outer ACL. Now, if condition failure induced by the nested "drop" causes
5012 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5013 or "require"), the connection is dropped.
5015 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5016 that yield "discard" can now be used with an "accept" or a "discard" verb,
5017 but an error is generated for any others (because I can't see a useful way
5018 to define what should happen).
5020 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5021 are now processed in the same way as lines in the Exim configuration file.
5022 In particular, continuation lines are supported.
5024 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5026 15. Added -ti meaning -t -i.
5028 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5029 domains, and warn by logging if others are found.
5031 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5032 returned from GIFCONF when Exim is trying to find the list of interfaces on
5033 a host. The code in os.c has been modified to copy each ifreq to an aligned
5034 structure in all cases.
5036 Also, in some cases, the returned ifreq's were being copied to a 'struct
5037 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5038 means the last couple of bytes of an IPv6 address could be chopped if the
5039 ifreq contained only a normal sockaddr (14 bytes storage).
5041 18. Named domain lists were not supported in the hosts_treat_as_local option.
5042 An entry such as +xxxx was not recognized, and was treated as a literal
5045 19. Ensure that header lines added by a DATA ACL are included in the reject log
5046 if the ACL subsequently rejects the message.
5048 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5049 MD5 (which is deprecated).
5051 21. When testing a filter file using -bf, Exim was writing a message when it
5052 took the sender from a "From " line in the message, but it was not doing so
5053 when it took $return_path from a Return-Path: header line. It now does.
5055 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5056 with a valid header line field name (a series of printing characters
5057 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5059 23. Changed "disc" in the source to "disk" to conform to the documentation and
5060 the book and for uniformity.
5062 24. Ignore Sendmail's -Ooption=value command line item.
5064 25. When execve() failed while trying to run a command in a pipe transport,
5065 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5066 could be confused with a return value of 69 from the command itself. This
5067 has been changed to 127, the value the shell returns if it is asked to run
5068 a non-existent command. The wording for the related log line suggests a
5069 non-existent command as the problem.
5071 26. If received_header_text expands to an empty string, do not add a Received:
5072 header line to the message. (Well, it adds a token one on the spool, but
5073 marks it "old" so that it doesn't get used or transmitted.)
5075 27. Installed eximstats 1.28 (addition of -nt option).
5077 28. There was no check for failure on the call to getsockname() in the daemon
5078 code. This can fail if there is a shortage of resources on the system, with
5079 ENOMEM, for example. A temporary error is now given on failure.
5081 29. Contrary to the C standard, it seems that in some environments, the
5082 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5083 program. Exim now does this explicitly; it affects the formatting of
5084 timestamps using strftime().
5086 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5087 complaints. I've now initialized all the variables, to avoid this.
5089 32. Header lines added by a system filter were not being "seen" during
5090 transport-time rewrites.
5092 33. The info_callback() function passed to OpenSSL is set up with type void
5093 (*)(SSL *, int, int), as described somewhere. However, when calling the
5094 function (actually a macro) that sets it up, the type void(*)() is
5095 expected. I've put in a cast to prevent warnings from picky compilers.
5097 34. If a DNS black list lookup found a CNAME record, but there were no A
5098 records associated with the domain it pointed at, Exim crashed.
5100 35. If a DNS black list lookup returned more than one A record, Exim ignored
5101 all but the first. It now scans all returned addresses if a particular IP
5102 value is being sought. In this situation, the contents of the
5103 $dnslist_value variable are a list of all the addresses, separated by a
5106 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5107 to accept a host name and all its aliases if the forward lookup for any of
5108 them yielded the IP address of the incoming connection. Now it accepts only
5109 those names whose forward lookup yields the correct IP address. Any other
5110 names are discarded. This closes a loophole whereby a rogue DNS
5111 administrator could create reverse DNS records to break through a
5112 wildcarded host restriction in an ACL.
5114 37. If a user filter or a system filter that ran in a subprocess used any of
5115 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5116 the wrong values were passed to the pipe command ($thisaddress had the
5117 value of $0, $0 had the value of $1, etc). This bug was introduced by
5118 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5120 38. Improved the line breaking for long SMTP error messages from ACLs.
5121 Previously, if there was no break point between 40 and 75 characters, Exim
5122 left the rest of the message alone. Two changes have been made: (a) I've
5123 reduced the minimum length to 35 characters; (b) if it can't find a break
5124 point between 35 and 75 characters, it looks ahead and uses the first one
5125 that it finds. This may give the occasional overlong line, but at least the
5126 remaining text gets split now.
5128 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5129 file descriptors might be low, and that setting 1000 would always raise it.
5130 It turns out that in some environments, the limit is already over 1000 and
5131 that lowering it causes trouble. So now Exim takes care not to decrease it.
5133 40. When delivering a message, the value of $return_path is set to $sender_
5134 address at the start of routing (routers may change the value). By an
5135 oversight, this default was not being set up when an address was tested by
5136 -bt or -bv, which affected the outcome if any router or filter referred to
5139 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5140 log only when "message" or "log_message" are set. However, if one of the
5141 conditions was an address verification, or a call to a nested ACL, the
5142 messages generated by the underlying test were being passed through. This
5143 no longer happens. The underlying message is available in $acl_verify_
5144 message for both "message" and "log_message" expansions, so it can be
5145 passed through if needed.
5147 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5148 new expansion $bh_ to give the encoded byte string without charset
5149 translation. Translation happens only if iconv() is available; HAVE_ICONV
5150 indicates this at build time. HEADERS_CHARSET gives the charset to
5151 translate to; headers_charset can change it in the configuration, and
5152 "headers charset" can change it in an individual filter file.
5154 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5155 that creates RFC 2047 encoded "words" labels them as that charset instead
5156 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5157 expansion operator; (ii) when Exim creates a From: line for a local
5158 message; (iii) when a header line is rewritten to include a "phrase" part.
5160 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5161 buggy, causing it to skip the first lines of messages whose message ID
5162 ended in 'D'. This would not have bitten before Exim release 4.14, because
5163 message IDs were unlikely to end in 'D' before then. The effect was to have
5164 incorrect size information for certain domains.
5166 45. #include "config.h" was missing at the start of the crypt16.c module. This
5167 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5170 46. If there was a timeout during a "random" callout check, Exim treated it as
5171 a failure of the random address, and carried on sending RSET and the real
5172 address. If the delay was just some slowness somewhere, the response to the
5173 original RCPT would be taken as a response to RSET and so on, causing
5174 mayhem of various kinds.
5176 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5177 when I implemented it. It didn't allow for the fact that some option values
5178 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5179 the right test for positive values.
5181 48. Domain names in DNS records are case-independent. Exim always looks them up
5182 in lower case. Some resolvers return domain names in exactly the case they
5183 appear in the zone file, that is, they may contain uppercase letters. Not
5184 all resolvers do this - some return always lower case. Exim was treating a
5185 change of case by a resolver as a change of domain, similar to a widening
5186 of a domain abbreviation. This triggered its re-routing code and so it was
5187 trying to route what was effectively the same domain again. This normally
5188 caused routing to fail (because the router wouldn't handle the domain
5189 twice). Now Exim checks for this case specially, and just changes the
5190 casing of the domain that it ultimately uses when it transmits the message
5193 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5196 50. If a filter generated a file delivery with a non-absolute name (possible if
5197 no home directory exists for the router), the forbid_file option was not
5200 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5201 the existing equality matching.
5203 52. Exim was using ints instead of ino_t variables in some places where it was
5204 dealing with inode numbers.
5206 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5207 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5208 called TMPDIR, and if it finds it is different, it changes its value.
5210 54. The smtp_printf() function is now made available to local_scan() so
5211 additional output lines can be written before returning. There is also an
5212 smtp_fflush() function to enable the detection of a dropped connection.
5213 The variables smtp_input and smtp_batched_input are exported to
5216 55. Changed the default runtime configuration: the message "Unknown user"
5217 has been removed from the ACL, and instead placed on the localuser router,
5218 using the cannot_route_message feature. This means that any verification
5219 failures that generate their own messages won't get overridden. Similarly,
5220 the "Unrouteable address" message that was in the ACL for unverifiable
5221 relay addresses has also been removed.
5223 56. Added hosts_avoid_esmtp to the smtp transport.
5225 57. The exicyclog script was not checking for the esoteric option
5226 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5227 will work only if exicyclog is run under the appropriate euid.
5229 58. Following a discussion on the list, the rules by which Exim recognises line
5230 endings on incoming messages have been changed. The -dropcr and drop_cr
5231 options are now no-ops, retained only for backwards compatibility. The
5232 following line terminators are recognized: LF CRLF CR. However, special
5233 processing applies to CR:
5235 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5236 nor a local message in the state where . is a terminator.
5238 (ii) If a bare CR is encountered in a header line, an extra space is added
5239 after the line terminator so as not to end the header. The reasoning
5240 behind this is that bare CRs in header lines are most likely either
5241 to be mistakes, or people trying to play silly games.
5243 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5244 was being incorrectly given as 18 bytes larger than it should have been.
5245 This is a VOB (very old bug).
5247 60. This may never have affected anything current, but just in case it has:
5248 When the local host is found other than at the start of a list of hosts,
5249 the local host, those with the same MX, and any that follow, are discarded.
5250 When the list in question was part of a longer list of hosts, the following
5251 hosts (not currently being processed) were also being discarded. This no
5252 longer happens. I'm not sure if this situation could ever has previously
5255 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5258 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5259 header. This is something that is recommended in a new Internet Draft, and
5260 is something that is documented as being done by Sendmail. There are two
5261 possible values. For messages generated by the autoreply transport, Exim
5264 Auto-Submitted: auto-replied
5266 whereas for all other generated messages (e.g. bounces) it adds
5268 Auto-Submitted: auto-generated
5270 63. The "personal" condition in filters now includes a test for the
5271 Auto-Submitted: header. If it contains the string "auto-" the message it
5272 not considered personal.
5274 64. Added rcpt_include_affixes as a generic transport option.
5276 65. Added queue_only_override (default true).
5278 66. Added the syslog_duplication option.
5280 67. If what should have been the first header line of a message consisted of
5281 a space followed by a colon, Exim was mis-interpreting it as a header line.
5282 It isn't of course - it is syntactically invalid and should therefore be
5283 treated as the start of the message body. The misbehaviour could have
5284 caused a number of strange effects, including loss of data in subsequent
5285 header lines, and spool format errors.
5287 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5288 client host had authenticated. This control can now be exercised by an ACL
5289 for more flexibility.
5291 69. By default, callouts do not happen when testing with -bh. There is now a
5292 variant, -bhc, which does actually run the callout code, including
5293 consulting and updating the callout cache.
5295 70. Added support for saslauthd authentication, courtesy of Alexander
5298 71. If statvfs() failed on the spool or log directories while checking their
5299 size for availability, Exim confusingly gave the error "space shortage".
5300 Furthermore, in debugging mode it crashed with a floating point exception.
5301 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5302 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5303 serious problem, Exim now writes to the main and panic logs when this
5304 happens, with details of the failure. It then refuses to accept the
5305 incoming message, giving the message "spool directory problem" or "log
5306 directory problem" with a 421 code for SMTP messages.
5308 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5309 0, 1, and 2 exist, because some OS complain for execs without them (see
5310 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5311 descriptors. However, the code omitted to check that the open succeeded,
5312 causing mysterious errors if for some reason the permissions on /dev/null
5313 got screwed. Now Exim writes a message to the main and panic logs, and
5314 bombs out if it can't open /dev/null.
5316 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5317 interact so that it is all more flexible. It is supposed to remain
5318 backwards compatible. Also added extra_local_interfaces.
5320 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5321 to bomb out with an assertion failure - to the client this appears as a
5322 connection drop. This problem occurs in the part of the code that was taken
5323 from the Samba project. Fortunately, the assertion is in a very simple
5324 function, so I have fixed this by reproducing the function inline in the
5325 one place where it is called, and arranging for authentication to fail
5326 instead of killing the process with assert().
5328 75. The SPA client code was not working when the server requested OEM rather
5329 than Unicode encoding.
5331 76. Added code to make require_files with a specific uid setting more usable in
5332 the case where statting the file as root fails - usually a non-root-mounted
5333 NFS file system. When this happens and the failure is EACCES, Exim now
5334 forks a subprocess and does the per-uid checking as the relevant uid.
5336 77. Added process_log_path.
5338 78. If log_file_path was not explicitly set, a setting of check_log_space or
5339 check_log_inodes was ignored.
5341 79. If a space check for the spool or log partitions fails, the incident is now
5342 logged. Of course, in the latter case the data may get lost...
5344 80. Added the %p formatting code to string_format() so that it can be used to
5345 print addresses in debug_print(). Adjusted all the address printing in the
5346 debugging in store.c to use %p rather than %d.
5348 81. There was a concern that a line of code in smtp_in.c could overflow a
5349 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5350 initially expressed, the concern was not well-founded, because trailing
5351 spaces are removed early. However, if the trailing spaces were followed by
5352 a NULL, they did not get removed, so the overflow was possible. Two fixes
5355 (a) I re-wrote the offending code in a cleaner fashion.
5356 (b) If an incoming SMTP command contains a NULL character, it is rejected
5359 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5360 runs initgroups(), so that if the Exim user is in any additional groups,
5361 they will be used during message reception.
5367 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob