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.
14 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
17 TF/02 Fix a couple more cases where we did not log the error message
18 when unlink() failed. See also change 4.74-TF/03.
24 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
26 PP/02 Harmonised TLS library version reporting; only show if debugging.
27 Layout now matches that introduced for other libraries in 4.74 PP/03.
29 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
31 PP/04 New "dns_use_edns0" global option.
33 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
36 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
37 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
39 TK/01 Updated PolarSSL code to 0.14.2.
40 Bugzilla 1097. Patch from Andreas Metzler.
42 PP/07 Catch divide-by-zero in ${eval:...}.
45 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
48 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
49 format-string attack -- SECURITY: remote arbitrary code execution.
51 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
52 time unintentionally subject to list matching rules, letting the header
53 cause arbitrary Exim lookups (of items which can occur in lists, *not*
54 arbitrary string expansion). This allowed for information disclosure.
56 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
57 INT_MIN/-1 -- value coerced to INT_MAX.
63 NM/01 Workround for PCRE version dependency in version reporting
66 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
67 This fixes portability to compilers other than gcc, notably
68 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
70 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
71 makefiles for portability to HP-UX and POSIX correctness.
73 PP/01 Permit LOOKUP_foo enabling on the make command-line.
74 Also via indented variable definition in the Makefile.
75 (Debugging by Oliver Heesakkers).
77 PP/02 Restore caching of spamd results with expanded spamd_address.
78 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
80 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
81 Improves build reliability. Fix from: Frank Elsner
83 NM/02 Fix wide character breakage in the rfc2047 coding
84 Fixes bug 1064. Patch from Andrey N. Oktyabrski
86 NM/03 Allow underscore in dnslist lookups
87 Fixes bug 1026. Patch from Graeme Fowler
89 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
90 Code patches from Adam Ciarcinski of NetBSD.
92 NM/04 Fixed exiqgrep to cope with mailq missing size issue
95 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
96 is logged, to avoid truncation. Patch from John Horne.
98 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
99 Patch from Jakob Hirsch.
101 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
102 SQL string expansion failure details.
103 Patch from Andrey Oktyabrski.
105 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
106 Patch from Simon Arlott.
108 PP/09 New lookups functionality failed to compile on old gcc which rejects
109 extern declarations in function scope.
110 Patch from Oliver Fleischmann
112 PP/10 Use sig_atomic_t for flags set from signal handlers.
113 Check getgroups() return and improve debugging.
114 Fixed developed for diagnosis in bug 927 (which turned out to be
117 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
118 Patch from Mark Zealey.
120 PP/12 Bugzilla 1056: Improved spamd server selection.
121 Patch from Mark Zealey.
123 PP/13 Bugzilla 1086: Deal with maildir quota file races.
124 Based on patch from Heiko Schlittermann.
126 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
127 Patch from Uwe Doering, sign-off by Michael Haardt.
129 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
130 variable declaration deep within a block. Bug and patch from
133 PP/15 lookups-Makefile IRIX compatibilty coercion.
135 PP/16 Make DISABLE_DKIM build knob functional.
137 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
138 Patch by Simon Arlott
140 TF/03 Fix valgrind.h portability to C89 compilers that do not support
141 variable argument macros. Our copy now differs from upstream.
147 TF/01 Failure to get a lock on a hints database can have serious
148 consequences so log it to the panic log.
150 TF/02 Log LMTP confirmation messages in the same way as SMTP,
151 controlled using the smtp_confirmation log selector.
153 TF/03 Include the error message when we fail to unlink a spool file.
155 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
156 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
157 for maintaining out-of-tree patches for some time.
159 PP/01 Bugzilla 139: Documentation and portability issues.
160 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
161 Handle per-OS dynamic-module compilation flags.
163 PP/02 Let /dev/null have normal permissions.
164 The 4.73 fixes were a little too stringent and complained about the
165 permissions on /dev/null. Exempt it from some checks.
166 Reported by Andreas M. Kirchwitz.
168 PP/03 Report version information for many libraries, including
169 Exim version information for dynamically loaded libraries. Created
170 version.h, now support a version extension string for distributors
171 who patch heavily. Dynamic module ABI change.
173 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
174 privilege escalation vulnerability whereby the Exim run-time user
175 can cause root to append content of the attacker's choosing to
178 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
181 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
182 If dropping privileges for untrusted macros, we disabled normal logging
183 on the basis that it would fail; for the Exim run-time user, this is not
184 the case, and it resulted in successful deliveries going unlogged.
185 Fixed. Reported by Andreas Metzler.
191 PP/01 Date: & Message-Id: revert to normally being appended to a message,
192 only prepend for the Resent-* case. Fixes regression introduced in
193 Exim 4.70 by NM/22 for Bugzilla 607.
195 PP/02 Include check_rfc2047_length in configure.default because we're seeing
196 increasing numbers of administrators be bitten by this.
198 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
200 PP/03 Bugzilla 994: added openssl_options main configuration option.
202 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
204 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
206 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
208 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
209 without a peer certificate, leading to a segfault because of an
210 assumption that peers always have certificates. Be a little more
211 paranoid. Problem reported by Martin Tscholak.
213 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
214 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
215 NB: ClamAV planning to remove STREAM in "middle of 2010".
216 CL also introduces -bmalware, various -d+acl logging additions and
217 more caution in buffer sizes.
219 PP/09 Implemented reverse_ip expansion operator.
221 PP/10 Bugzilla 937: provide a "debug" ACL control.
223 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
225 PP/12 Bugzilla 973: Implement --version.
227 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
229 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
231 PP/15 Bugzilla 816: support multiple condition rules on Routers.
233 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
234 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
235 ignore trailing whitespace.
237 JJ/02 prevent non-panic DKIM error from being sent to paniclog
239 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
242 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
243 Notification from Dr Andrew Aitchison.
245 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
246 ExtendedDetectionInfo response format.
247 Notification from John Horne.
249 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
252 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
253 XSL and documented dependency on system catalogs, with examples of how
256 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
259 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
260 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
261 configuration file which is writeable by the Exim user or group.
263 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
264 of configuration files to cover files specified with the -C option if
265 they are going to be used with root privileges, not just the default
268 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
269 option (effectively making it always true).
271 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
272 files to be used while preserving root privileges.
274 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
275 that rogue child processes cannot use them.
277 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
278 run-time user, instead of root.
280 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
281 Exim run-time user without dropping privileges.
283 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
284 result string, instead of calling string_vformat() twice with the same
287 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
288 for other users. Others should always drop root privileges if they use
289 -C on the command line, even for a whitelisted configure file.
291 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
293 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
299 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
300 $data_path, and $header_path variables; fixed documentation bugs and
303 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
304 exipick to access non-standard spools, including the "frozen" queue
307 NM/01 Bugzilla 965: Support mysql stored procedures.
308 Patch from Alain Williams
310 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
312 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
313 Patch from Andreas Metzler
315 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
316 Patch from Kirill Miazine
318 NM/05 Bugzilla 671: Added umask to procmail example.
320 JJ/03 installed exipick 20100323.0, fixing doc bug
322 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
323 directory. Notification and patch from Dan Rosenberg.
325 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
327 TK/02 Improve log output when DKIM signing operation fails.
329 MH/01 Treat the transport option dkim_domain as a colon separated
330 list, not as a single string, and sign the message with each element,
331 omitting multiple occurences of the same signer.
333 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
334 Bugzilla 985, 986. Patch by Simon Arlott
336 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
337 Patch by Simon Arlott
339 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
340 MBX locking. Notification from Dan Rosenberg.
346 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
348 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
350 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
352 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
354 NM/04 Bugzilla 915: Flag broken perl installation during build.
360 TK/01 Added patch by Johannes Berg that expands the main option
361 "spamd_address" if it starts with a dollar sign.
363 TK/02 Write list of recipients to X-Envelope-Sender header when building
364 the mbox-format spool file for content scanning (suggested by Jakob
367 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
368 (http://www.dcc-servers.net/) support via dccifd. Activated by
369 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
371 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
372 by Mark Daniel Reidel <mr@df.eu>.
374 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
375 When building exim an external PCRE library is now needed -
376 PCRE is a system library on the majority of modern systems.
377 See entry on PCRE_LIBS in EDITME file.
379 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
380 conversation. Added nologin parameter to request.
381 Patch contributed by Kirill Miazine.
383 TF/01 Do not log submission mode rewrites if they do not change the address.
385 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
387 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
388 log files in place. Contributed by Roberto Lima.
390 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
392 TF/03 Bugzilla 615: When checking the local_parts router precondition
393 after a local_part_suffix or local_part_prefix option, Exim now
394 does not use the address's named list lookup cache, since this
395 contains cached lookups for the whole local part.
397 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
398 Robert Millan. Documentation is in experimental-spec.txt.
400 TF/04 Bugzilla 668: Fix parallel build (make -j).
402 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
404 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
405 Patch provided by Jan Srzednicki.
407 TF/05 Leading white space used to be stripped from $spam_report which
408 wrecked the formatting. Now it is preserved.
410 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
411 that they are available at delivery time.
413 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
415 TF/08 TLS error reporting now respects the incoming_interface and
416 incoming_port log selectors.
418 TF/09 Produce a more useful error message if an SMTP transport's hosts
419 setting expands to an empty string.
421 NM/06 Bugzilla 744: EXPN did not work under TLS.
422 Patch provided by Phil Pennock.
424 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
425 Patch provided by Richard Godbee.
427 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
428 acl_smtp_notquit, added index entry.
430 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
431 Patch provided by Eugene Bujak.
433 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
434 accept(). Patch provided by Maxim Dounin.
436 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
437 Patch provided by Phil Pennock.
439 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
441 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
442 Patch provided by Brad "anomie" Jorsch.
444 NM/14 Bugzilla 622: Added timeout setting to transport filter.
445 Patch provided by Dean Brooks.
447 TK/05 Add native DKIM support (does not depend on external libraries).
449 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
450 Patch provided by Graeme Fowler.
452 NM/16 Bugzilla 851: Documentation example syntax fix.
454 NM/17 Changed NOTICE file to remove references to embedded PCRE.
456 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
459 NM/19 Bugzilla 745: TLS version reporting.
460 Patch provided by Phil Pennock.
462 NM/20 Bugzilla 167: bool: condition support.
463 Patch provided by Phil Pennock.
465 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
466 clients. Patch provided by Phil Pennock.
468 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
469 Patch provided by Brad "anomie" Jorsch.
471 NM/23 Bugzilla 687: Fix misparses in eximstats.
472 Patch provided by Heiko Schlittermann.
474 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
475 Patch provided by Heiko Schlittermann.
477 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
478 plus update to original patch.
480 NM/26 Bugzilla 799: Documentation correction for ratelimit.
482 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
483 Patch provided by David Brownlee.
485 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
487 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
489 NM/30 Bugzilla 888: TLS documentation bugfixes.
491 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
493 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
494 Unlike the original bugzilla I have changed all shell scripts in src tree.
496 NM/33 Bugzilla 898: Transport filter timeout fix.
497 Patch by Todd Rinaldo.
499 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
500 Patch by Serge Demonchaux.
502 NM/35 Bugzilla 39: Base64 decode bug fixes.
503 Patch by Jakob Hirsch.
505 NM/36 Bugzilla 909: Correct connect() call in dcc code.
507 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
509 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
511 NM/39 Bugzilla 911: Fixed MakeLinks build script.
517 TK/01 Add preliminary DKIM support. Currently requires a forked version of
518 ALT-N's libdkim that I have put here:
519 http://duncanthrax.net/exim-experimental/
521 Note to Michael Haardt: I had to rename some vars in sieve.c. They
522 were called 'true' and it seems that C99 defines that as a reserved
523 keyword to be used with 'bool' variable types. That means you could
524 not include C99-style headers which use bools without triggering
525 build errors in sieve.c.
527 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
528 as mailq or other aliases. Changed the --help handling significantly
529 to do whats expected. exim_usage() emits usage/help information.
531 SC/01 Added the -bylocaldomain option to eximstats.
533 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
535 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
537 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
543 PH/01 Another patch from the Sieve maintainer.
545 PH/02 When an IPv6 address is converted to a string for single-key lookup
546 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
547 dots are used instead of colons so that keys in lsearch files need not
548 contain colons. This was done some time before quoting was made available
549 in lsearch files. However, iplsearch files do require colons in IPv6 keys
550 (notated using the quote facility) so as to distinguish them from IPv4
551 keys. This meant that lookups for IP addresses in host lists did not work
552 for iplsearch lookups.
554 This has been fixed by arranging for IPv6 addresses to be expressed with
555 colons if the lookup type is iplsearch. This is not incompatible, because
556 previously such lookups could never work.
558 The situation is now rather anomolous, since one *can* have colons in
559 ordinary lsearch keys. However, making the change in all cases is
560 incompatible and would probably break a number of configurations.
562 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
565 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
566 conversion specification without a maximum field width, thereby enabling
567 a rogue spamd server to cause a buffer overflow. While nobody in their
568 right mind would setup Exim to query an untrusted spamd server, an
569 attacker that gains access to a server running spamd could potentially
570 exploit this vulnerability to run arbitrary code as the Exim user.
572 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
573 $primary_hostname instead of what libspf2 thinks the hosts name is.
575 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
576 a directory entry by the name of the lookup key. Previously, if a
577 symlink pointed to a non-existing file or a file in a directory that
578 Exim lacked permissions to read, a lookup for a key matching that
579 symlink would fail. Now it is enough that a matching directory entry
580 exists, symlink or not. (Bugzilla 503.)
582 PH/03 The body_linecount and body_zerocount variables are now exported in the
585 PH/04 Added the $dnslist_matched variable.
587 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
588 This means they are set thereafter only if the connection becomes
591 PH/06 Added the client_condition to authenticators so that some can be skipped
592 by clients under certain conditions.
594 PH/07 The error message for a badly-placed control=no_multiline_responses left
595 "_responses" off the end of the name.
597 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
599 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
600 (without spaces) instead of just copying the configuration text.
602 PH/10 Added the /noupdate option to the ratelimit ACL condition.
604 PH/11 Added $max_received_linelength.
606 PH/12 Added +ignore_defer and +include_defer to host lists.
608 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
609 way in which PCRE > 7.0 is built.
611 PH/14 Implemented queue_only_load_latch.
613 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
614 MAIL command. The effect was to mangle the value on 64-bit systems.
616 PH/16 Another patch from the Sieve maintainer.
618 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
620 PH/18 If a system quota error occurred while trying to create the file for
621 a maildir delivery, the message "Mailbox is full" was not appended to the
622 bounce if the delivery eventually timed out. Change 4.67/27 below applied
623 only to a quota excession during the actual writing of the file.
625 PH/19 It seems that peer DN values may contain newlines (and other non-printing
626 characters?) which causes problems in log lines. The DN values are now
627 passed through string_printing() before being added to log lines.
629 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
630 and InterBase are left for another time.)
632 PH/21 Added message_body_newlines option.
634 PH/22 Guard against possible overflow in moan_check_errorcopy().
636 PH/23 POSIX allows open() to be a macro; guard against that.
638 PH/24 If the recipient of an error message contained an @ in the local part
639 (suitably quoted, of course), incorrect values were put in $domain and
640 $local_part during the evaluation of errors_copy.
646 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
647 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
650 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
651 issue a MAIL command.
653 PH/02 In an ACL statement such as
655 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
657 if a client was not listed at all, or was listed with a value other than
658 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
659 the condition was not true (as it should be), so access was not denied.
660 The bug was that the ! inversion was incorrectly passed on to the second
661 item. This has been fixed.
663 PH/03 Added additional dnslists conditions == and =& which are different from
664 = and & when the dns lookup returns more than one IP address.
666 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
667 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
669 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
670 FSYNC, which compiles an option called disable_fsync that allows for
671 bypassing fsync(). The documentation is heavily laced with warnings.
673 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
675 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
676 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
677 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
678 including adding "make clean"; (3) Added -fPIC when compiling the test
679 dynamically loaded module, to get rid of a warning.
681 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
682 message fails, move_frozen_messages = true and ignore_bounce_errors_after
683 = 0s. The bug is otherwise harmless.
685 PH/07 There was a bug in the dovecot authenticator such that the value of
686 $auth1 could be overwritten, and so not correctly preserved, after a
687 successful authentication. This usually meant that the value preserved by
688 the server_setid option was incorrect.
690 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
692 PH/09 Installed PCRE release 7.0.
694 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
695 run for batched SMTP input. It is now run at the start of every message
696 in the batch. While fixing this I discovered that the process information
697 (output by running exiwhat) was not always getting set for -bs and -bS
698 input. This is fixed, and it now also says "batched" for BSMTP.
700 PH/11 Added control=no_pipelining.
702 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
703 patch, slightly modified), and move the expansion of helo_data till after
704 the connection is made in the smtp transport (so it can use these
707 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
709 PH/14 Added log_selector = +pid.
711 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
713 PH/16 Add ${if forany and ${if forall.
715 PH/17 Added dsn_from option to vary the From: line in DSNs.
717 PH/18 Flush SMTP output before performing a callout, unless control =
718 no_callout_flush is set.
720 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
721 was true (the default) a successful delivery failed to delete the retry
722 item, thus causing premature timeout of the address. The bug is now
725 PH/20 Added hosts_avoid_pipelining to the smtp transport.
727 PH/21 Long custom messages for fakedefer and fakereject are now split up
728 into multiline reponses in the same way that messages for "deny" and
729 other ACL rejections are.
731 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
732 with slight modification.
734 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
735 draft, changing the syntax and factoring some duplicate code".
737 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
738 for deliveries of the second and subsequent messages over the same SMTP
741 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
742 ${reduce, with only minor "tidies".
744 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
746 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
747 expansion side effects.
749 PH/27 When a message times out after an over-quota error from an Exim-imposed
750 quota, the bounce message says "mailbox is full". This message was not
751 being given when it was a system quota that was exceeded. It now should
754 MH/03 Made $recipients available in local_scan(). local_scan() already has
755 better access to the recipient list through recipients_list[], but
756 $recipients can be useful in postmaster-provided expansion strings.
758 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
759 in the case of a MAIL command with additional options following the
760 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
761 were accidentally chopped off.
763 PH/29 SMTP synchronization checks are implemented when a command is read -
764 there is a check that no more input is waiting when there shouldn't be
765 any. However, for some commands, a delay in an ACL can mean that it is
766 some time before the response is written. In this time, more input might
767 arrive, invalidly. So now there are extra checks after an ACL has run for
768 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
769 pipelining has not been advertised.
771 PH/30 MH's patch to allow iscntrl() characters to be list separators.
773 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
774 returned in the SMTP response when smtp_return_error_details was false.
777 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
778 instead of the C I/O that was originally supplied, because problems were
781 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
782 Exim which did not show up earlier: it was assuming that a call to
783 SSL_CTX_set_info_callback() might give an error value. In fact, there is
784 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
785 was a macro that became an assignment, so it seemed to work. This has
786 changed to a proper function call with a void return, hence the compile
787 error. Exim's code has been fixed.
789 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
792 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
794 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
796 PH/37 If a message is not accepted after it has had an id assigned (e.g.
797 because it turns out to be too big or there is a timeout) there is no
798 "Completed" line in the log. When some messages of this type were
799 selected by exigrep, they were listed as "not completed". Others were
800 picked up by some special patterns. I have improved the selection
801 criteria to be more general.
803 PH/38 The host_find_failed option in the manualroute router can now be set
804 to "ignore", to completely ignore a host whose IP address cannot be
805 found. If all hosts are ignored, the behaviour is controlled by the new
806 host_all_ignored option.
808 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
809 homing or because of multiple MX records with /mx) generated more than
810 one IP address, and the following item turned out to be the local host,
811 all the secondary addresses of the first item were incorrectly removed
812 from the list, along with the local host and any following hosts (which
813 is what is supposed to happen).
815 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
816 whoever called Exim into the -H file. In the case of the daemon it was
817 behaving confusingly. When first started, it used values for whoever
818 started the daemon, but after a SIGHUP it used the Exim user (because it
819 calls itself on a restart). I have changed the code so that it now always
822 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
823 message are rejected with the same error (e.g. no authentication or bad
824 sender address), and a DATA command is nevertheless sent (as can happen
825 with PIPELINING or a stupid MUA), the error message that was given to the
826 RCPT commands is included in the rejection of the DATA command. This is
827 intended to be helpful for MUAs that show only the final error to their
830 PH/42 Another patch from the Sieve maintainer.
832 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
833 Eximstats - Fixed some broken HTML links and added missing column headers
835 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
836 columns for Rejects, Temp Rejects, Ham, and Spam rows.
838 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
840 PH/43 Yet another patch from the Sieve maintainer.
842 PH/44 I found a way to check for a TCP/IP connection going away before sending
843 the response to the final '.' that terminates a message, but only in the
844 case where the client has not sent further data following the '.'
845 (unfortunately, this is allowed). However, in many cases there won't be
846 any further data because there won't be any more messages to send. A call
847 to select() can be used: if it shows that the input is "ready", there is
848 either input waiting, or the socket has been closed. An attempt to read
849 the next input character can distinguish the two cases. Previously, Exim
850 would have sent an OK response which the client would never have see.
851 This could lead to message repetition. This fix should cure that, at
852 least in a lot of common cases.
854 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
855 advertised in response to EHLO.
861 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
862 fixed by 4.65/MH/01 (is this a record?) are fixed:
864 (i) An empty string was always treated as zero by the numeric comparison
865 operators. This behaviour has been restored.
867 (ii) It is documented that the numeric comparison operators always treat
868 their arguments as decimal numbers. This was broken in that numbers
869 starting with 0 were being interpreted as octal.
871 While fixing these problems I realized that there was another issue that
872 hadn't been noticed. Values of message_size_limit (both the global option
873 and the transport option) were treated as octal if they started with 0.
874 The documentation was vague. These values are now always treated as
875 decimal, and I will make that clear in the documentation.
881 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
882 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
885 MH/01 Don't check that the operands of numeric comparison operators are
886 integers when their expansion is in "skipping" mode (fixes bug
887 introduced by 4.64-PH/07).
889 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
890 child addresses, Exim now panics and dies. Previously, because the count
891 is held in a short int, deliveries were likely to be lost. As such a
892 large number of recipients for a single message is ridiculous
893 (performance will be very, very poor), I have chosen to impose a limit
894 rather than extend the field.
900 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
901 leftover -K file (the existence of which was triggered by #402).
902 While we were at it, introduced process PID as part of the -K
903 filename. This should rule out race conditions when creating
906 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
907 processing considerably. Previous code took too long for large mails,
908 triggering a timeout which in turn triggers #401.
910 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
911 in the DK code in transports.c. sendfile() is not really portable,
912 hence the _LINUX specificness.
914 TF/01 In the add_headers option to the mail command in an Exim filter,
915 there was a bug that Exim would claim a syntax error in any
916 header after the first one which had an odd number of characters
919 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
920 callout verification, Exim cached a "reject" for the entire domain. This
921 is correct for most verifications, but it is not correct for a recipient
922 verification with use_sender or use_postmaster set, because in that case
923 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
924 case of MAIL FROM:<> rejection from other early rejections (e.g.
925 rejection of HELO). When verifying a recipient using a non-null MAIL
926 address, the cache is ignored if it shows MAIL FROM:<> rejection.
927 Whatever the result of the callout, the value of the domain cache is
928 left unchanged (for any other kind of callout, getting as far as trying
929 RCPT means that the domain itself is ok).
931 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
934 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
935 manifest itself as EPIPE rather than ECONNECT. When tidying away a
936 session, the daemon ignores ECONNECT errors and logs others; it now
937 ignores EPIPE as well.
939 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
940 (quoted-printable decoding).
942 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
943 later the small subsequent patch to fix an introduced bug.
945 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
947 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
949 PH/08 An error is now given if message_size_limit is specified negative.
951 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
952 to be given (somewhat) arbitrary names.
954 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
957 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
958 miscellaneous code fixes
960 PH/10 Added the log_reject_target ACL modifier to specify where to log
963 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
964 hostname. This is wrong, because it relates to the incoming message (and
965 probably the interface on which it is arriving) and not to the outgoing
966 callout (which could be using a different interface). This has been
967 changed to use the value of the helo_data option from the smtp transport
968 instead - this is what is used when a message is actually being sent. If
969 there is no remote transport (possible with a router that sets up host
970 addresses), $smtp_active_hostname is used.
972 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
973 tweaks were necessary in order to get it to work (see also 21 below):
974 (a) The code assumed that strncpy() returns a negative number on buffer
975 overflow, which isn't the case. Replaced with Exim's string_format()
977 (b) There were several signed/unsigned issues. I just did the minimum
978 hacking in of casts. There is scope for a larger refactoring.
979 (c) The code used strcasecmp() which is not a standard C function.
980 Replaced with Exim's strcmpic() function.
981 (d) The code set only $1; it now sets $auth1 as well.
982 (e) A simple test gave the error "authentication client didn't specify
983 service in request". It would seem that Dovecot has changed its
984 interface. Fortunately there's a specification; I followed it and
985 changed what the client sends and it appears to be working now.
987 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
990 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
991 address A is aliased to B and C, where B exists and C does not. Without
992 -v the output is "A verified" because verification stops after a
993 successful redirection if more than one address is generated. However,
994 with -v the child addresses are also verified. Exim was outputting "A
995 failed to verify" and then showing the successful verification for C,
996 with its parentage. It now outputs "B failed to verify", showing B's
997 parentage before showing the successful verification of C.
999 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1000 look up a TXT record in a specific list after matching in a combined
1003 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1004 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1005 they consult the DNS. I had assumed they would set it the way they
1006 wanted; and indeed my experiments on Linux seem to show that in some
1007 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1008 To be on the safe side, however, I have now made the interface to
1009 host_find_byname() similar to host_find_bydns(), with an argument
1010 containing the DNS resolver options. The host_find_byname() function now
1011 sets these options at its start, just as host_find_bydns() does. The smtp
1012 transport options dns_qualify_single and dns_search_parents are passed to
1013 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1014 of host_find_byname() use the default settings of RES_DEFNAMES
1015 (qualify_single) but not RES_DNSRCH (search_parents).
1017 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1018 spool_read_header() do less string testing, by means of a preliminary
1019 switch on the second character of optional "-foo" lines. (This is
1020 overdue, caused by the large number of possibilities that now exist.
1021 Originally there were few.) While I was there, I also converted the
1022 str(n)cmp tests so they don't re-test the leading "-" and the first
1023 character, in the hope this might squeeze out yet more improvement.
1025 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1026 flag allowing group syntax was set by the header_syntax check but not
1027 turned off, possible causing trouble later; (2) The flag was not being
1028 set at all for the header_verify test, causing "group"-style headers to
1029 be rejected. I have now set it in this case, and also caused header_
1030 verify to ignore an empty address taken from a group. While doing this, I
1031 came across some other cases where the code for allowing group syntax
1032 while scanning a header line wasn't quite right (mostly, not resetting
1033 the flag correctly in the right place). These bugs could have caused
1034 trouble for malformed header lines. I hope it is now all correct.
1036 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1037 with the "reply" argument non-NULL. The code, however (which originally
1038 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1039 but it didn't always do it. This confused somebody who was copying the
1040 code for some other use. I have removed all the tests.
1042 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1043 feature that was used to support insecure browsers during the U.S. crypto
1044 embargo. It requires special client support, and Exim is probably the
1045 only MTA that supported it -- and would never use it because real RSA is
1046 always available. This code has been removed, because it had the bad
1047 effect of slowing Exim down by computing (never used) parameters for the
1048 RSA_EXPORT functionality.
1050 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1051 authenticator to fail if there's a tab character in the incoming data
1052 (there should never be unless someone is messing about, as it's supposed
1053 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1054 if the connection is using TLS or if the remote IP is the same as the
1055 local IP, and the "valid-client-cert option" if a client certificate has
1058 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1059 authenticators. This can be used for authorization after authentication
1060 succeeds. (In the case of plaintext, it servers for both authentication
1063 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1064 if any retry times were supplied.
1066 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1067 connection, where there is no client IP address to check. In this
1068 situation, the verify now always succeeds.
1070 PH/25 Applied John Jetmore's -Mset patch.
1072 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1074 PH/27 In a string expansion for a processed (not raw) header when multiple
1075 headers of the same name were present, leading whitespace was being
1076 removed from all of them, but trailing whitespace was being removed only
1077 from the last one. Now trailing whitespace is removed from each header
1078 before concatenation. Completely empty headers in a concatenation (as
1079 before) are ignored.
1081 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1082 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1084 PH/29 [Removed. This was a change that I later backed out, and forgot to
1085 correct the ChangeLog entry (that I had efficiently created) before
1086 committing the later change.]
1088 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1089 address errors (4xx response to RCPT) over the same connection as other
1090 messages routed to the same hosts. Such deliveries are always "forced",
1091 so retry times are not inspected. This resulted in far too many retries
1092 for the affected addresses. The effect occurred only when there were more
1093 hosts than the hosts_max_try setting in the smtp transport when it had
1094 the 4xx errors. Those hosts that it had tried were not added to the list
1095 of hosts for which the message was waiting, so if all were tried, there
1096 was no problem. Two fixes have been applied:
1098 (i) If there are any address or message errors in an SMTP delivery, none
1099 of the hosts (tried or untried) are now added to the list of hosts
1100 for which the message is waiting, so the message should not be a
1101 candidate for sending over the same connection that was used for a
1102 successful delivery of some other message. This seems entirely
1103 reasonable: after all the message is NOT "waiting for some host".
1104 This is so "obvious" that I'm not sure why it wasn't done
1105 previously. Hope I haven't missed anything, but it can't do any
1106 harm, as the worst effect is to miss an optimization.
1108 (ii) If, despite (i), such a delivery is accidentally attempted, the
1109 routing retry time is respected, so at least it doesn't keep
1110 hammering the server.
1112 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1113 in ${readsocket because some servers need this prod.
1115 PH/32 Added some extra debug output when updating a wait-xxx database.
1117 PH/33 The hint "could be header name not terminated by colon", which has been
1118 given for certain expansion errors for a long time, was not being given
1119 for the ${if def:h_colon_omitted{... case.
1121 PH/34 The spec says: "With one important exception, whenever a domain list is
1122 being scanned, $domain contains the subject domain." There was at least
1123 one case where this was not true.
1125 PH/35 The error "getsockname() failed: connection reset by peer" was being
1126 written to the panic log as well as the main log, but it isn't really
1127 panic-worthy as it just means the connection died rather early on. I have
1128 removed the panic log writing for the ECONNRESET error when getsockname()
1131 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1132 runs only) independently of the message's sender address. This meant
1133 that, if the 4xx error was in fact related to the sender, a different
1134 message to the same recipient with a different sender could confuse
1135 things. In particualar, this can happen when sending to a greylisting
1136 server, but other circumstances could also provoke similar problems.
1137 I have changed the default so that the retry time for these errors is now
1138 based a combination of the sender and recipient addresses. This change
1139 can be overridden by setting address_retry_include_sender=false in the
1142 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1143 remote server are returned as part of bounce messages. This was not
1144 happening for LMTP over a pipe (the lmtp transport), but now it is the
1145 same for both kinds of LMTP.
1147 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1148 in header lines that were in fact CNAMEs. This is no longer the case.
1150 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1151 and queue runs started by the daemon processed all messages. This has
1152 been fixed so that -R and -S can now usefully be given with -q<time>.
1154 PH/40 Import PCRE release 6.7 (fixes some bugs).
1156 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1158 PH/42 Give an error if -q is specified more than once.
1160 PH/43 Renamed the variables $interface_address and $interface_port as
1161 $received_ip_address and $received_port, to make it clear that these
1162 values apply to message reception, and not to the outgoing interface when
1163 a message is delivered. (The old names remain recognized, of course.)
1165 PH/44 There was no timeout on the connect() call when using a Unix domain
1166 socket in the ${readsocket expansion. There now is.
1168 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1169 be meaningful with "accept".
1171 SC/01 Eximstats V1.43
1172 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1174 SC/02 Eximstats V1.44
1175 Use a glob alias rather than an array ref in the generated
1176 parser. This improves both readability and performance.
1178 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1179 Collect SpamAssassin and rejection statistics.
1180 Don't display local sender or destination tables unless
1181 there is data to show.
1182 Added average volumes into the top table text output.
1184 SC/04 Eximstats V1.46
1185 Collect data on the number of addresses (recipients)
1186 as well as the number of messages.
1188 SC/05 Eximstats V1.47
1189 Added 'Message too big' to the list of mail rejection
1190 reasons (thanks to Marco Gaiarin).
1192 SC/06 Eximstats V1.48
1193 Mainlog lines which have GMT offsets and are too short to
1194 have a flag are now skipped.
1196 SC/07 Eximstats V1.49 (Alain Williams)
1197 Added the -emptyok flag.
1199 SC/08 Eximstats V1.50
1200 Fixes for obtaining the IP address from reject messages.
1202 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1203 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1204 whitesspace changes from 4.64-PH/27
1206 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1209 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1210 are found, allow negative numbers in numeric criteria)
1212 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1214 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1215 to match changes made in 4.64-PH/43
1217 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1219 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1220 "domain" is actually an IP address, and force a failure. This locks out
1221 those revolvers/nameservers that support "A-for-A" lookups, in
1222 contravention of the specifications.
1224 PH/48 When a host name was looked up from an IP address, and the subsequent
1225 forward lookup of the name timed out, the host name was left in
1226 $sender_host_name, contrary to the specification.
1228 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1229 restricted to single-key lookups, Exim was not diagnosing an error if
1230 * or *@ was used with a query-style lookup.
1232 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1234 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1235 long ago, but noone interested enough thought of it. Let's just say that
1236 the "1.1" means that there are some new functions that weren't there at
1237 some point in the past.
1239 PH/51 Error processing for expansion failure of helo_data from an smtp
1240 transport during callout processing was broken.
1242 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1243 tested/used via the -bh/-bhc/-bs options.
1245 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1246 bug, fixed in subsequent PCRE releases).
1248 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1249 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1251 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1257 SC/01 Use a glob alias rather than an array ref in eximstats generated
1258 parser. This improves both readability and performance.
1260 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1261 Don't display local sender or destination tables in eximstats unless
1262 there is data to show.
1263 Added average volumes into the eximstats top table text output.
1265 SC/03 Collect data on the number of addresses (recipients) as well
1266 as the number of messages in eximstats.
1268 TF/01 Correct an error in the documentation for the redirect router. Exim
1269 does (usually) call initgroups() when daemonizing.
1271 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1272 with consistent privilege compared to when running as a daemon.
1274 TF/03 Note in the spec that $authenticated_id is not set for local
1275 submissions from trusted users.
1277 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1278 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1280 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1281 by adding some example configuration directives to the default
1282 configuration file. A little bit of work is required to uncomment the
1283 directives and define how usernames and passwords are checked, but
1284 there is now a framework to start from.
1286 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1287 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1288 without this. I don't know how relevant this is to other LDAP libraries.
1290 PH/02 Add the verb name to the "unknown ACL verb" error.
1292 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1294 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1296 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1297 directory not expanded when it should be if an expanded home directory
1298 was set for the address (which is overridden by the transport).
1300 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1303 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1304 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1305 because it is too late at that time, and has no effect.
1307 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1308 security issue with \' (bugzilla #107). I could not use the
1309 PQescapeStringConn() function, because it needs a PGconn value as one of
1312 PH/08 When testing addresses using -bt, indicate those final addresses that
1313 are duplicates that would not cause an additional delivery. At least one
1314 person was confused, thinking that -bt output corresponded to deliveries.
1315 (Suppressing duplicates isn't a good idea as you lose the information
1316 about possibly different redirections that led to the duplicates.)
1318 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1319 systems where poll() doesn't work, in particular OS X.
1321 PH/10 Added more information to debugging output for retry time not reached.
1323 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1324 operations in malware.c.
1326 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1329 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1330 syslog_duplication set false, log lines that would normally be written
1331 both the the main log and to the reject log were not written to syslog at
1334 PH/14 In the default configuration, change the use of "message" in ACL warn
1335 statements to "add_header".
1337 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1338 not followed by a command (e.g. "seen endif").
1340 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1341 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1344 PH/17 Added extra conditions to the default value of delay_warning_condition
1348 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1349 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1350 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1353 The Auto-Submitted: and various List- headers are standardised, whereas I
1354 don't think Precedence: ever was.
1356 PH/18 Refactored debugging code in route_finduser() to show more information,
1357 in particular, the error code if getpwnam() issues one.
1359 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1360 This is apparently needed in addition to the PH/07 change above to avoid
1361 any possible encoding problems.
1363 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1364 but not after initializing Perl.
1366 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1367 output them only if debugging. By default they are written stderr,
1368 apparently, which is not desirable.
1370 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1373 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1376 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1378 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1379 authenticated or an ident call has been made. Suppress the default
1380 values for $authenticated_id and $authenticated_sender (but permit -oMai
1381 and -oMas) when testing with -bh.
1383 PH/24 Re-jigged the order of the tests in the default configuration so that the
1384 tests for valid domains and recipients precede the DNS black list and CSA
1385 tests, on the grounds that those ones are more expensive.
1387 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1388 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1389 This bug exists in every version of Exim that I still have, right back to
1392 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1393 However, an attempt to turn on case-sensitivity in a regex key by
1394 including (?-i) didn't work because the subject string was already
1395 lowercased, and the effects were non-intuitive. It turns out that a
1396 one-line patch can be used to allow (?-i) to work as expected.
1402 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1403 other effects) broke the use of negated acl sub-conditions.
1405 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1408 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1409 "Deny" causes Exim to reject the incoming connection with a 554 error.
1410 Unfortunately, if there is a major crisis, such as a disk failure,
1411 tcp-wrappers gives "deny", whereas what one would like would be some
1412 kind of temporary error. A kludge has been added to help with this.
1413 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1414 554 error is used if errno is still zero or contains ENOENT (which occurs
1415 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1418 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1420 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1421 errors. Otherwise a message that provokes a temporary error (when other
1422 messages do not) can cause a whole host to time out.
1424 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1425 addresses were routed directly to files or pipes from a redirect router.
1426 File deliveries just didn't batch; pipe deliveries might have suffered
1429 PH/06 A failure to get a lock for a hints database would erroneously always say
1430 "Failed to get write lock", even when it was really a read lock.
1432 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1433 of 0600. This has been changed to use the value of the lockfile_mode
1434 option (which defaults to 0600).
1436 PH/08 Applied small patch from the Sieve maintainer.
1438 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1439 folder from quota calculations, a direct delivery into this folder messed
1440 up the contents of the maildirsize file. This was because the regex was
1441 used only to exclude .Trash (or whatever) when the size of the mailbox
1442 was calculated. There was no check that a delivery was happening into an
1443 excluded directory. This bug has been fixed by ignoring all quota
1444 processing for deliveries into excluded directories.
1446 PH/10 Added the maildirfolder_create_regex option to appendfile.
1452 PH/01 The code for finding all the local interface addresses on a FreeBSD
1453 system running IPv6 was broken. This may well have applied to all BSD
1454 systems, as well as to others that have similar system calls. The broken
1455 code found IPv4 interfaces correctly, but gave incorrect values for the
1456 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1457 that it would not match correctly against @[] and not recognize the IPv6
1460 PH/02 The ipliteral router was not recognizing addresses of the form user@
1461 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1463 PH/03 Added disable_ipv6.
1465 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1466 decoded form, because it is most often used to construct To: headers
1467 lines in autoreplies, and the decoded form may well be syntactically
1468 invalid. However, $reply_address has leading white space removed, and all
1469 newlines turned into spaces so that the autoreply transport does not
1472 PH/05 If group was specified without a user on a router, and no group or user
1473 was specified on a transport, the group from the router was ignored.
1475 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1476 for visible compile-time settings that can be used to change these
1477 numbers, for those that want even more. Backwards compatibility with old
1478 spool files has been maintained. However, going back to a previous Exim
1479 release will lost any variables that are in spool files.
1481 PH/07 Two small changes when running in the test harness: increase delay when
1482 passing a TCP/IP connection to a new process, in case the original
1483 process has to generate a bounce, and remove special handling of
1484 127.0.0.2 (sic), which is no longer necessary.
1486 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1487 be the same on different OS.
1489 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1492 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1493 whether --show-vars was specified or not
1495 JJ/02 exipick: Added support for new ACL variable spool format introduced
1498 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1499 syntactically invalid From: or Reply-to: line, and a filter used this to
1500 generate an autoreply, and therefore failed to obtain an address for the
1501 autoreply, Exim could try to deliver to a non-existent relative file
1502 name, causing unrelated and misleading errors. What now happens is that
1503 it logs this as a hard delivery error, but does not attempt to create a
1506 PH/11 The exinext utility has a -C option for testing purposes, but although
1507 the given file was scanned by exinext itself; it wasn't being passed on
1508 when Exim was called.
1510 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1511 an end-of-file indication when reading a command response.
1513 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1514 compiled. In many other places in Exim, IPv6 addresses are always
1515 recognized, so I have changed this. It also means that IPv4 domain
1516 literals of the form [IPV4:n.n.n.n] are now always recognized.
1518 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1519 used if the router is not running as root, for example, when verifying at
1520 ACL time, or when using -bh. The debugging output from this situation was
1521 non-existent - all you got was a failure to exec. I have made two
1524 (a) Failures to set uid/gid, the current directory, or a process leader
1525 in a subprocess such as that created by queryprogram now generate
1526 suitable debugging ouput when -d is set.
1528 (b) The queryprogram router detects when it is not running as root,
1529 outputs suitable debugging information if -d is set, and then runs
1530 the subprocess without attempting to change uid/gid.
1532 PH/15 Minor change to Makefile for building test_host (undocumented testing
1535 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1536 additional section of a DNS packet that returns MX or SRV records.
1537 Instead, it always explicitly searches for A/AAAA records. This avoids
1538 major problems that occur when a DNS server includes only records of one
1539 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1540 fixed another bug: if SRV records were looked up and the corresponding
1541 address records were *not* found in the additional section, the port
1542 values from the SRV records were lost.
1544 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1545 using the correct key (the original address) when searching the retry
1546 rules in order to find which one to use for generating the retry hint.
1548 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1549 adding the default one. Similarly, if it contains a Reply-To: header, the
1550 errors_reply_to option, if set, is not used.
1552 PH/19 When calculating a retry time, Exim used to measure the "time since
1553 failure" by looking at the "first failed" field in the retry record. Now
1554 it does not use this if it is later than than the arrival time of the
1555 message. Instead it uses the arrival time. This makes for better
1556 behaviour in cases where some deliveries succeed, thus re-setting the
1557 "first failed" field. An example is a quota failure for a huge message
1558 when small messages continue to be delivered. Without this change, the
1559 "time since failure" will always be short, possible causing more frequent
1560 delivery attempts for the huge message than are intended.
1561 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1563 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1564 $1, $2, $3) because the numerical variables can be reset during some
1565 expansion items (e.g. "match"), thereby losing the authentication data.
1567 PH/21 Make -bV show the size of off_t variables so that the test suite can
1568 decide whether to run tests for quotas > 2G.
1570 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1571 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1572 filecount value is greater than 2G or if a quota value is greater than 2G
1573 on a system where the size of off_t is not greater than 4, a panic error
1576 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1577 never match. The debug and -bh output now contains an explicit error
1578 message indicating a malformed IPv4 address or mask.
1580 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1581 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1582 PH/23 above applies.
1584 PH/25 Do not write to syslog when running in the test harness. The only
1585 occasion when this arises is a failure to open the main or panic logs
1586 (for which there is an explicit test).
1588 PH/26 Added the /no_tell option to "control=freeze".
1590 PH/27 If a host name lookup failed very early in a connection, for example, if
1591 the IP address matched host_lookup and the reverse lookup yielded a name
1592 that did not have a forward lookup, an error message of the form "no IP
1593 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1594 could be logged. Now it outputs the IP address instead of "NULL".
1596 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1597 allows the sender and the authenticated sender to be set when
1598 submitting a message from within Exim. Since child_open_exim() is
1599 documented for local_scan(), the new function should be too.
1601 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1602 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1603 results in an empty string is now treated as unset.
1605 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1607 PH/31 Added sender_verify_fail logging option.
1609 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1610 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1611 tidied the source and removed it altogether.
1613 PH/33 When a queue run was abandoned because the load average was too high, a
1614 log line was always written; now it is written only if the queue_run log
1615 selector is set. In addition, the log line for abandonment now contains
1616 information about the queue run such as the pid. This is always present
1617 in "start" and "stop" lines but was omitted from the "abandon" line.
1619 PH/34 Omit spaces between a header name and the colon in the error message that
1620 is given when verify = headers_syntax fails (if there are lots of them,
1621 the message gets confusing).
1623 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1624 names, as there are now some PTR records that contain slashes. This check
1625 is only to protect against broken name servers that fall over on strange
1626 characters, so the fact that it applies to all lookups doesn't matter.
1628 PH/36 Now that the new test suite is complete, we can remove some of the
1629 special code in Exim that was needed for the old test suite. For example,
1630 sorting DNS records because real resolvers return them in an arbitrary
1631 order. The new test suite's fake resolver always returns records in the
1634 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1635 bounces) except when queue_only is set, to avoid logging races between
1636 the different processes.
1638 PH/38 Panic-die if .include specifies a non-absolute path.
1640 PH/39 A tweak to the "H" retry rule from its user.
1642 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1643 a label. They prevented compilation on older perls.
1645 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1646 a warning to be raised on newish perls.
1648 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1649 on queue. Changes to match documented behaviour of showing count of
1650 messages matching specified criteria.
1652 PH/40 Changed the default ident timeout from 30s to 5s.
1654 PH/41 Added support for the use of login_cap features, on those BSD systems
1655 that have them, for controlling the resources used by pipe deliveries.
1657 PH/42 The content-scanning code uses fopen() to create files in which to put
1658 message data. Previously it was not paying any attention to the mode of
1659 the files. Exim runs with umask(0) because the rest of the code creates
1660 files with open(), and sets the required mode explicitly. Thus, these
1661 files were ending up world-writeable. This was not a big issue, because,
1662 being within the spool directory, they were not world-accessible. I have
1663 created a function called modefopen, which takes an additional mode
1664 argument. It sets umask(777), creates the file, chmods it to the required
1665 mode, then resets the umask. All the relevant calls to fopen() in the
1666 content scanning code have been changed to use this function.
1668 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1669 to 24 hours. This avoids potential overflow problems when processing G
1670 and H retry rules. I suspect nobody ever tinkers with this value.
1672 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1674 PH/45 When the plaintext authenticator is running as a client, the server's
1675 challenges are checked to ensure they are valid base64 strings. By
1676 default, the authentication attempt is cancelled if an invalid string is
1677 received. Setting client_ignore_invalid_base64 true ignores these errors.
1678 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1679 they are received. Thus, the responses can be made to depend on the
1680 challenges. If an invalid string is ignored, an empty string is placed in
1683 PH/46 Messages that are created by the autoreply transport now contains a
1684 References: header, in accordance with RFCs 2822 and 3834.
1686 PH/47 Added authenticated_sender_force to the smtp transport.
1688 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1690 PH/49 Installed latest patch from the Sieve maintainer.
1692 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1693 was also set, the appendfile transport was unnecessarily scanning a
1694 directory of message files (e.g. for maildir delivery) to find the count
1695 of files (along with the size), even though it did not need this
1696 information. It now does the scan only if it needs to find either the
1697 size of the count of files.
1699 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1701 PH/52 Two bugs concerned with error handling when the smtp transport is
1704 (i) Exim was not creating retry information for temporary errors given
1705 for individual recipients after the DATA command when the smtp transport
1706 was used in LMTP mode. This meant that they could be retried too
1707 frequently, and not timed out correctly.
1709 (ii) Exim was setting the flag that allows error details to be returned
1710 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1711 recipients that were returned after the DATA command.
1713 PH/53 This is related to PH/52, but is more general: for any failing address,
1714 when detailed error information was permitted to be returned to the
1715 sender, but the error was temporary, then after the final timeout, only
1716 "retry timeout exceeded" was returned. Now it returns the full error as
1717 well as "retry timeout exceeded".
1719 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1720 do this, and (what is worse) MTAs that accept it.
1722 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1723 will now be deprecated.
1725 PH/56 New os.c-cygwin from the Cygwin maintainer.
1727 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1728 formats (previously only available in exim formats via -bpr, -bpru,
1729 and -bpra. Now also available in native and exiqgrep formats)
1731 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1732 with very large, slow to parse queues
1734 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1736 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1738 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1739 responses to authentication challenges, though it was showing the
1740 challenges; (ii) I've removed the CR characters from the debug output for
1743 PH/58 Allow for the insertion of a newline as well as a space when a string
1744 is turned into more than one encoded-word during RFC 2047 encoding. The
1745 Sieve code now uses this.
1747 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1748 data_4xx, lost_connection, tls_required.
1750 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1751 message was being sent as an SMTP response.
1753 PH/61 Add -l and -k options to exicyclog.
1755 PH/62 When verifying, if an address was redirected to one new address, so that
1756 verification continued, and the new address failed or deferred after
1757 having set something in $address_data, the value of $address_data was not
1758 passed back to the ACL. This was different to the case when no
1759 redirection occurred. The value is now passed back in both cases.
1761 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1762 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1763 use login_cap.h, so on its own it isn't the distinguishing feature. The
1764 new name refers directly to the setclassresources() function.
1766 PH/65 Added configuration files for NetBSD3.
1768 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1770 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1771 is preferred over IPv4.
1773 PH/68 The bounce_return_message and bounce_return_body options were not being
1774 honoured for bounces generated during the reception of non-SMTP messages.
1775 In particular, this applied to messages rejected by the ACL. This bug has
1776 been fixed. However, if bounce_return_message is true and bounce_return_
1777 body is false, the headers that are returned for a non-SMTP message
1778 include only those that have been read before the error was detected.
1779 (In the case of an ACL rejection, they have all been read.)
1781 PH/69 The HTML version of the specification is now built in a directory called
1782 spec_html instead of spec.html, because the latter looks like a path with
1783 a MIME-type, and this confuses some software.
1785 PH/70 Catch two compiler warnings in sieve.c.
1787 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1788 function verify_get_ident() calls ip_connect() to connect a socket, but
1789 if the "connect()" function timed out, ip_connect() used to close the
1790 socket. However, verify_get_ident() also closes the socket later, and in
1791 between Exim writes to the log, which may get opened at this point. When
1792 the socket was closed in ip_connect(), the log could get the same file
1793 descriptor number as the socket. This naturally causes chaos. The fix is
1794 not to close the socket in ip_connect(); the socket should be closed by
1795 the function that creates it. There was only one place in the code where
1796 this was missing, in the iplookup router, which I don't think anybody now
1797 uses, but I've fixed it anyway.
1799 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1800 well as to direct DNS lookups. Otherwise the handling of names in host
1801 lists is inconsistent and therefore confusing.
1807 PH/01 Two changes to the default runtime configuration:
1809 (1) Move the checks for relay_from_hosts and authenticated clients from
1810 after to before the (commented out) DNS black list checks.
1812 (2) Add control=submission to the relay_from_hosts and authenticated
1813 clients checks, on the grounds that messages accepted by these
1814 statements are most likely to be submissions.
1816 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1818 (1) Generate an error if the third argument for the ${prvs expansion is
1821 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1824 (3) Reset the variables that are obtained from the first argument of
1825 ${prvscheck and used in the second argument before leaving the code,
1826 because their memory is reclaimed, so using them afterwards may do
1829 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1830 one (it's much easier than Tom thought :-).
1832 (5) Because of (4), we can now allow for the use of $prvscheck_result
1833 inside the third argument.
1835 PH/03 For some reason, the default setting of PATH when running a command from
1836 a pipe transport was just "/usr/bin". I have changed it to
1839 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1840 anything to be listed in the output from -bV.
1842 PH/05 When a filter generated an autoreply, the entire To: header line was
1843 quoted in the delivery log line, like this:
1845 => >A.N.Other <ano@some.domain> <original@ddress> ...
1847 This has been changed so that it extracts the operative address. There
1848 may be more than one such address. If so, they are comma-separated, like
1851 => >ano@some.domain,ona@other.domain <original@ddress> ...
1853 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1854 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1855 not being looked up in the rDNS to get a host name, Exim was showing the
1856 IP address twice in Received: lines, even though the IP addresses were
1857 identical. For example:
1859 Received: from [1.2.3.4] (helo=[1.2.3.4])
1861 However, if the real host name was known, it was omitting the HELO data
1862 if it matched the actual IP address. This has been tidied up so that it
1863 doesn't show the same IP address twice.
1865 PH/07 When both +timestamp and +memory debugging was on, the value given by
1866 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1867 called by the debug printing, thereby overwriting the timestamp buffer.
1868 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1870 PH/08 When the original message was included in an autoreply transport, it
1871 always said "this is a copy of the message, including all the headers",
1872 even if body_only or headers_only was set. It now gives an appropriate
1875 PH/09 Applied a patch from the Sieve maintainer which:
1877 o fixes some comments
1878 o adds the (disabled) notify extension core
1879 o adds some debug output for the result of if/elsif tests
1880 o points to the current vacation draft in the documentation
1881 and documents the missing references header update
1885 o fixes a bug in processing the envelope test (when testing
1886 multiple envelope elements, the last element determinted the
1889 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1890 Electronic Mail") by including:
1892 Auto-submitted: auto-generated
1894 in the messages that it generates (bounce messages and others, such as
1895 warnings). In the case of bounce messages for non-SMTP mesages, there was
1896 also a typo: it was using "Auto_submitted" (underscore instead of
1897 hyphen). Since every message generated by Exim is necessarily in response
1898 to another message, thes have all been changed to:
1900 Auto-Submitted: auto-replied
1902 in accordance with these statements in the RFC:
1904 The auto-replied keyword:
1906 - SHOULD be used on messages sent in direct response to another
1907 message by an automatic process,
1909 - MUST NOT be used on manually-generated messages,
1911 - MAY be used on Delivery Status Notifications (DSNs) and Message
1912 Disposition Notifications (MDNs),
1914 - MUST NOT be used on messages generated by automatic or periodic
1915 processes, except for messages which are automatic responses to
1918 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1919 to the default Received: header definition.
1921 PH/12 Added log selector acl_warn_skipped (default on).
1923 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1924 variables because (a) they are in the wrong storage pool and (b) even if
1925 they were copied, it wouldn't work properly because of the caching.
1927 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1928 checking when decoding. Apparently there are clients that generate
1929 overlong encoded strings. Why am I not surprised?
1931 PH/15 If the first argument of "${if match_address" was not empty, but did not
1932 contain an "@" character, Exim crashed. Now it writes a panic log message
1933 and treats the condition as false.
1935 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1937 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1938 extension "envelope-auth". The code is finished and in agreement with
1939 other implementations, but there is no documentation so far and in fact,
1940 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1941 not changing the active code.
1943 Print executed "if" and "elsif" statements when debugging is used. This
1944 helps a great deal to understand what a filter does.
1946 Document more things not specified clearly in RFC3028. I had all this
1947 sorted out, when out of a sudden new issues came to my mind. Oops."
1949 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1952 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1953 canonical form (as documented). However, after a host name lookup from
1954 the IP address, check_host() was doing a simple string comparison with
1955 addresses acquired from the DNS when checking that the found name did
1956 have the original IP as one of its addresses. Since any found IPv6
1957 addresses are likely to be in abbreviated form, the comparison could
1958 fail. Luckily, there already exists a function for doing the comparison
1959 by converting both addresses to binary, so now that is used instead of
1960 the text comparison.
1962 PH/20 There was another similar case to PH/19, when a complete host name was
1963 given in a host list; looking up its IP address could give an abbreviated
1964 form, whereas the current host's name might or might not be abbreviated.
1965 The same fix has been applied.
1971 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1972 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1975 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1976 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1978 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1980 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1981 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1982 Also, for "Auto-Submitted", treat anything other than "no" as
1983 non-personal, in accordance with RFC 3834. (Previously it treated
1984 anything starting "auto-" as non-personal.)
1986 TF/01 The control=submission/name=... option had a problem with syntax
1987 errors if the name included a slash character. The /name= option
1988 now slurps the rest of the string, so it can include any characters
1989 but it must come last in the list of options (after /sender_retain
1992 PH/05 Some modifications to the interface to the fake nameserver for the new
2000 TK/01 Added the "success_on_redirect" address verification option. See
2001 NewStuff for rationale and an example.
2003 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2005 PH/02 Patch to exigrep to allow it to work on syslog lines.
2007 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2008 fread() to skip over the body file's header line, because in Cygwin the
2009 header line is locked and is inaccessible.
2011 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2012 co-exist for some time) to make it clear that it is the Exim ID that is
2013 referenced, not the Message-ID: header line.
2015 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2016 string_format() function, because snprintf() does not exist on all
2019 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2020 ${stat: expansion item.
2022 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2023 protocol synchronization error", to keep the pedants happy.
2025 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2026 well as for IRIX systems, when gcc is being used. See the host.c source
2029 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2031 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2034 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2035 transport and to the smtp transport in LMTP mode.
2037 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2039 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2040 run a filter in a subprocess. This could lead to confusion in subsequent
2041 lookups in the parent process. There should also be a search_tidyup() at
2042 the end of the subprocess.
2044 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2045 only if the host matched helo_try_verify_hosts, which caused the
2046 verification to occur when the EHLO/HELO command was issued. The ACL just
2047 tested the remembered result. Now, if a previous verification attempt has
2048 not happened, "verify = helo" does it there and then.
2050 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2052 TK/03 Fix log output including CR from clamd.
2054 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2055 exist provoked a memory error which could cause a segfault.
2057 PH/15 Installed PCRE 6.2
2059 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2061 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2062 of the problem. Specifically, suggested +O2 rather than +O1 for the
2065 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2067 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2068 record, the port was not correctly logged when the outgoing_port log
2069 selector was set (it logged the transort's default port).
2071 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2072 fallback_hosts, and "hosts" in the smtp transport.
2074 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2075 host errors such as "Connection refused".
2077 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2078 authentication with radiusclient 0.4.9:
2080 - Error returned from rc_read_config was caught wrongly
2081 - Username/password not passed on to radius server due to wrong length.
2083 The presumption is that some radiusclient API changes for 4.51/PH/17
2084 were not taken care of correctly. The code is still untested by me (my
2085 Linux distribution still has 0.3.2 of radiusclient), but it was
2086 contributed by a Radius user.
2088 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2089 expanding the "port" option of the smtp transport.
2091 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2092 while reading a MIME header. Thanks to Tom Hughes for a patch.
2094 PH/24 Include config.h inside local_scan.h so that configuration settings are
2097 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2098 that in an ACL for RCPT (for example), you can examine exactly what was
2101 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2102 commands, but it was not correctly comparing the address with the actual
2103 client host address. Thus, it would show the EHLO address in Received:
2104 header lines when this was not necessary.
2106 PH/27 Added the % operator to ${eval:}.
2108 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2109 it should be ignoring failures (because with -C, for example, it has lost
2110 privilege). It wasn't ignoring creation failures other than "already
2113 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2116 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2117 because an input line was too long, either on its own, or by virtue of
2118 too many continuations, the temporary file was not being removed, and the
2119 return code was incorrect.
2121 PH/31 Missing "BOOL" in function definition in filtertest.c.
2123 PH/32 Applied Sieve patches from the maintainer.
2125 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2127 PH/33 Added "verify = not_blind".
2129 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2130 Local/Makefile (with some defaults set). These are used in built scripts
2131 such as exicyclog, but they have never been used in the exim_install
2132 script (though there are many overriding facilities there). I have
2133 arranged that the exim_install script now takes note of these two
2136 PH/35 Installed configuration files for Dragonfly.
2138 PH/36 When a locally submitted message by a trusted user did not contain a
2139 From: header, and the sender address was obtained from -f or from an SMTP
2140 MAIL command, and the trusted user did not use -F to supply a sender
2141 name, $originator_name was incorrectly used when constructing a From:
2142 header. Furthermore, $originator_name was used for submission mode
2143 messages from external hosts without From: headers in a similar way,
2144 which is clearly wrong.
2146 PH/37 Added control=suppress_local_fixups.
2148 PH/38 When log_selector = +received_sender was set, and the addition of the
2149 sender made the log line's construction buffer exactly full, or one byte
2150 less than full, an overflow happened when the terminating "\n" was
2153 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2154 when the result of a list match is failure because a DNS lookup failed.
2156 PH/40 RM_COMMAND is now used in the building process.
2158 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2159 the "build-* directories that it finds.
2161 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2162 address in a domain literal was a prefix of an interface address.
2164 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2165 when verifying a sender address, unless rewrite_headers is false.
2167 PH/44 Wrote a long comment about why errors_to addresses are verified as
2168 recipients, not senders.
2170 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2171 the ratelimit ACL was added.
2173 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2175 PH/46 Added extra information about PostgreSQL errors to the error string.
2177 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2178 avoiding the need to install special zones in a real server. This is
2179 backwards compatible; if it can't find the fake resolver, it drops back.
2180 Thus, both old and new test suites can be run.
2182 TF/02 Added util/ratelimit.pl
2184 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2185 clock is set back in time.
2187 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2188 Candler <B.Candler@pobox.com>.
2190 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2191 OK for addresses that are the result of redirections.
2193 PH/48 A number of further additions for the benefit of the new test suite,
2194 including a fake gethostbyname() that interfaces to the fake DNS resolver
2197 TF/06 The fix for widen_domains has also been applied to qualify_single and
2198 search_parents which are the other dnslookup options that can cause
2201 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2204 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2206 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2207 service" if it encountered an MX record with an empty target hostname.
2208 The message is now "an MX or SRV record indicated no SMTP service".
2210 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2211 if the DNS of the sending site is misconfigured. This is quite a
2212 common situation. This change restores the behaviour of treating a
2213 helo verification defer as a failure.
2215 PH/51 If self=fail was set on a router, the bounce message did not include the
2216 actual error message.
2222 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2224 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2225 command itself ended in error, the underlying message about the transport
2226 filter timeout was being overwritten with the pipe command error. Now the
2227 underlying error message should be appended to the second error message.
2229 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2231 PH/02 Reduce the amount of output that "make" produces by default. Full output
2232 can still be requested.
2234 PH/03 The warning log line about a condition test deferring for a "warn" verb
2235 was being output only once per connection, rather than after each
2236 occurrence (because it was using the same function as for successful
2237 "warn" verbs). This seems wrong, so I have changed it.
2239 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2240 it should not have, which might have caused a crash in the right
2241 circumstances, but probably never did.
2243 PH/04 Installed a modified version of Tony Finch's patch to make submission
2244 mode fix the return path as well as the Sender: header line, and to
2245 add a /name= option so that you can make the user's friendly name appear
2248 TF/03 Added the control = fakedefer ACL modifier.
2250 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2251 Mark Lowes for thorough testing.
2253 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2255 TK/03 Merged latest SRS patch from Miles Wilton.
2257 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2258 with the definition in sysexits.h (which is #included earlier).
2259 Fortunately, Exim does not actually use EX_OK. The code used to try to
2260 preserve the sysexits.h value, by assumimg that macro definitions were
2261 scanned for macro replacements. I have been disabused of this notion,
2262 so now the code just undefines EX_OK before #including unistd.h.
2264 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2265 in the smtp transport. When a block could not be written in a single
2266 write() function, the timeout was being re-applied to each part-write.
2267 This seems wrong - if the receiver was accepting one byte at a time it
2268 would take for ever. The timeout is now adjusted when this happens. It
2269 doesn't have to be particularly precise.
2271 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2272 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2274 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2275 without a domain if the check to <postmaster@domain> fails.
2277 SC/01 Eximstats: added -xls and the ability to specify output files
2278 (patch written by Frank Heydlauf).
2280 SC/02 Eximstats: use FileHandles for outputing results.
2282 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2284 SC/04 Eximstats: fixed display of large numbers with -nvr option
2286 SC/05 Eximstats: fixed merging of reports with empty tables.
2288 SC/06 Eximstats: added the -include_original_destination flag
2290 SC/07 Eximstats: removed tabs and trailing whitespace.
2292 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2294 TK/06 MBOX spool code: Add real "From " MBOX separator line
2295 so the .eml file is really in mbox format (even though
2296 most programs do not really care). Patch from Alex Miller.
2298 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2299 The latter is generated from $received_to and is only set if the
2300 message has one envelope recipient. SA can use these headers,
2301 obviously out-of-the-box. Patch from Alex Miller.
2303 PH/08 The ${def test on a variable was returning false if the variable's
2304 value was "0", contrary to what the specification has always said!
2305 The result should be true unless the variable is empty.
2307 PH/09 The syntax error of a character other than { following "${if
2308 def:variable_name" (after optional whitespace) was not being diagnosed.
2309 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2310 accidental colon was present, for example, could give incorrect results.
2312 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2313 result is used (not including appendfile, where other changes are about
2316 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2317 This involved changing a lot of size variables from int to off_t. It
2318 should work with maildirs and everything.
2320 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2321 spamd dying while we are connected to it.
2323 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2326 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2327 function for BDB 4.3.
2329 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2331 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2332 the sources for PCRE have been re-arranged and more files are now
2335 PH/15 The code I had for printing potentially long long variables in PH/11
2336 above was not the best (it lost precision). The length of off_t variables
2337 is now inspected at build time, and an appropriate printing format (%ld
2338 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2339 to be "long long int" or "long int". This is needed for the internal
2340 formatting function string_vformat().
2342 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2343 the configuration file to be ":syslog", then the script "guesses" where
2344 the logs files are, rather than using the compiled in default. In our
2345 case the guess is not the same as the compiled default, so the script
2346 suddenly stopped working when I started to use syslog. The patch checks
2347 to see if log_file_path is "". If so, it attempts to read it from exim
2348 with no configuration file to get the compiled in version, before it
2349 falls back to the previous guessing code."
2351 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2352 implementing BATV in an Exim configuration. See NewStuff for the gory
2355 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2356 Makefile that are specific to HP-UX.
2358 PH/18 If the "use_postmaster" option was set for a recipient callout together
2359 with the "random" option, the postmaster address was used as the MAIL
2360 FROM address for the random test, but not for the subsequent recipient
2361 test. It is now used for both.
2363 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2364 patch removes a few documentation additions to RFC 3028, because the
2365 latest draft now contains them. It adds the new en;ascii-case comparator
2366 and a new error check for 8bit text in MIME parts. Comparator and
2367 require names are now matched exactly. I enabled the subaddress
2368 extension, but it is not well tested yet (read: it works for me)."
2370 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2371 rework some of the code of TK/09 above to avoid the hardwired use of
2372 "%lld" and "long long". Replaced the call to snprintf() with a call to
2375 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2376 records point to non-existent hosts", "retry timeout exceeded", and
2377 "retry time not reached for any host after a long failure period".
2379 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2380 experimental DomainKeys support:
2382 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2383 (2) On an error such as an illegally used "control", the wrong name for
2384 the control was given.
2386 These problems did NOT occur unless DomainKeys support was compiled.
2388 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2390 PH/24 Added ${if match_ip condition.
2392 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2393 databases so that it will be absolutely obvious if a crash occurs in the
2394 DB library. This is a regular occurrence (often caused by mis-matched
2397 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2398 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2399 picked up on a user's system that detects such things. There doesn't seem
2400 to be a gcc warning option for this - only an attribute that has to be
2401 put on the function's prototype. It seems that in Fedora Core 4 they have
2402 set this on a number of new functions. No doubt there will be more in due
2405 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2406 specify a transport. However, if an address that was verified by such a
2407 router was the subject of a callout, Exim crashed because it tried to
2408 read the rcpt_include_affixes from the non-existent transport. Now it
2409 just assumes that the setting of that option is false. This bug was
2410 introduced by 4.51/PH/31.
2412 PH/28 Changed -d+all to exclude +memory, because that information is very
2413 rarely of interest, but it makes the output a lot bigger. People tend to
2414 do -d+all out of habit.
2416 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2417 code in os-type was giving problems when libc.so lives in lib64, like on
2420 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2421 aren't the modern standard, and it seems that some systems' include files
2422 don't always have them. Exim was already checking for some of the newer
2423 ones like T_AAAA, and defining it itself. I've added checks for all the
2424 record types that Exim uses.
2426 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2427 not automatically generating a new one, as it is supposed to. This
2428 prevented TLS from working. If the file did exist, but contained invalid
2429 data, a new version was generated, as expected. It was only the case of a
2430 non-existent file that was broken.
2432 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2433 with a change in libdomainkeys > 0.64.
2435 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2436 from DNS. If the selector record carries the flag, it now has
2437 precedence over the domain-wide flag.
2439 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2441 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2442 the use of a port name that isn't defined in /etc/services) occurred, the
2443 message was deferred as in a normal delivery, and thus remained on the
2444 spool, instead of being failed because of the mua_wrapper setting. This
2445 is now fixed, and I tidied up some of the mua_wrapper messages at the
2448 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2449 the messages in a hash of arrays rather than using individual hashes.
2450 This is a bit cleaner and results in dramatic memory savings, albeit
2451 at a slight CPU cost.
2453 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2454 as requested by Marc Sherman.
2456 SC/10 Eximstats: added histograms for user specified patterns as requested
2459 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2461 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2462 fopen() in the content-scanning modules that did not already have it.
2468 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2469 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2471 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2473 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2475 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2476 to test Sieve filters that use "vacation".
2478 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2479 that changes the way the GnuTLS parameters are stored in the cache file.
2480 The new format can be generated externally. For backward compatibility,
2481 if the data in the cache doesn't make sense, Exim assumes it has read an
2482 old-format file, and it generates new data and writes a new file. This
2483 means that you can't go back to an older release without removing the
2486 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2487 work if there are any delivery delays because "one_time" forces the
2488 parent to be marked "delivered", so its unseen clone is never tried
2489 again. For this reason, Exim now forbids the simultaneous setting of
2492 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2493 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2494 entry if you want to know the details. The fix, however, neglected to
2495 consider the case when local delivery batching is involved. The test for
2496 "previously delivered" was not happening when checking to see if an
2497 address could be batched with a previous (undelivered) one; under
2498 certain circumstances this could lead to multiple deliveries to the same
2501 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2502 in its include files, and this causes problems building Exim.
2504 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2505 header_syntax) but Exim was just ignoring anything given after a slash.
2506 In particular, this caused confusion with an attempt to use "verify =
2507 reverse_host_lookup/defer_ok". An error is now given when options are
2508 supplied for verify items that do not have them. (Maybe reverse_host_
2509 lookup should have a defer_ok option, but that's a different point.)
2511 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2512 defined by RFC 821) to 2048, because there were problems with some AUTH
2513 commands, and RFC 1869 says the size should be increased for extended
2514 SMTP commands that take arguments.
2516 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2519 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2520 "unknown" error; now it says that the functionality isn't in the binary.
2522 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2523 an address' error message when a string expansion fails (syntax or
2524 whatever). Otherwise the password may appear in the log. Following change
2525 PH/42 below, there is no longer a chance of it appearing in a bounce
2528 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2530 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2531 of its addresses was ever tried. (Bugzilla bug #2.)
2533 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2534 the result incorrectly in the debug output. (It correctly added a newline
2535 to what was transported.)
2537 TF/01 Added $received_time.
2539 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2540 commented out examples of how to interface to a virus scanner and to
2541 SpamAssassin. Also added commented examples of av_scanner and
2542 spamd_address settings.
2544 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2545 and controls are allowed in which ACLs. There were a couple of minor
2546 errors. Some of the entries in the conditions table (which is a table of
2547 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2548 negation of where the condition IS allowed.
2550 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2552 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2553 header file does not have a version number, so I've had to invent a new
2554 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2555 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2556 radiusclient), but it was contributed by a Radius user.
2558 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2559 files or directories, for OpenSSL.
2561 PH/19 When an Exim process that is running as root has to create an Exim log
2562 file, it does so in a subprocess that runs as exim:exim so as to get the
2563 ownership right at creation (otherwise, other Exim processes might see
2564 the file with the wrong ownership). There was no test for failure of this
2565 fork() call, which would lead to the process getting stuck as it waited
2566 for a non-existent subprocess. Forks do occasionally fail when resources
2567 run out. I reviewed all the other calls to fork(); they all seem to check
2570 PH/20 When checking for unexpected SMTP input at connect time (before writing
2571 the banner), Exim was not dealing correctly with a non-positive return
2572 from the read() function. If the client had disconnected by this time,
2573 the result was a log entry for a synchronization error with an empty
2574 string after "input=" when read() returned zero. If read() returned -1
2575 (an event I could not check), uninitialized data bytes were printed.
2576 There were reports of junk text (parts of files, etc) appearing after
2579 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2581 PH/22 Added support for macro redefinition, and (re)definition in between
2582 driver and ACL definitions.
2584 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2585 forgetting to use the resulting value; it was using the unexpanded value.
2587 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2588 hadn't been configured. The fix is from Juergen Kreileder, who
2589 understands it better than I do:
2591 "Here's what I see happening with three configured cyrus_sasl
2592 authenticators configured (plain, login, cram-md5):
2594 On startup auth_cyrus_sasl_init() gets called for each of these.
2595 This means three calls to sasl_listmech() without a specified mech_list.
2596 => SASL tests which mechs of all available mechs actually work
2597 => three warnings about OTP not working
2598 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2600 With the patch, sasl_listmech() also gets called three times. But now
2601 SASL's mech_list option is set to the server_mech specified in the the
2602 authenticator. Or in other words, the answer from sasl_listmech()
2603 gets limited to just the mech you're testing for (which is different
2605 => the return list contains just 'plain' or 'login', 'cram-md5' or
2606 nothing depending on the value of ob->server_mech.
2608 I've just tested the patch: Authentication still works fine,
2609 unavailable mechs specified in the exim configuration are still
2610 caught, and the auth.log warnings about OTP are gone."
2612 PH/25 When debugging is enabled, the contents of the command line are added
2613 to the debugging output, even when log_selector=+arguments is not
2616 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2617 answer is "GNU", and only if the return is "GNU/something" is the answer
2620 PH/27 $acl_verify_message is now set immediately after the failure of a
2621 verification in an ACL, and so is available in subsequent modifiers. In
2622 particular, the message can be preserved by coding like this:
2624 warn !verify = sender
2625 set acl_m0 = $acl_verify_message
2627 Previously, $acl_verify_message was set only while expanding "message"
2628 and "log_message" when a very denied access.
2630 PH/28 Modified OS/os.c-Linux with
2632 -#ifndef OS_LOAD_AVERAGE
2633 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2635 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2636 nomenclature these days.)
2638 PH/29 Installed patch from the Sieve maintainer that adds the options
2639 sieve_useraddress and sieve_subaddress to the redirect router.
2641 PH/30 In these circumstances:
2642 . Two addresses routed to the same list of hosts;
2643 . First host does not offer TLS;
2644 . First host accepts first address;
2645 . First host gives temporary error to second address;
2646 . Second host offers TLS and a TLS session is established;
2647 . Second host accepts second address.
2648 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2649 and peerdn, if requested) that were in fact used only for the second
2652 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2653 attention to any local part prefix or suffix that was matched by the
2654 router that accepted the address. It now behaves in the same way as it
2655 does for delivery: the affixes are removed from the local part unless
2656 rcpt_include_affixes is set on the transport.
2658 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2659 timeout during the DATA phase of an incoming message.
2661 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2662 applied a patch sanctioned by the Sieve maintainer.
2664 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2665 the uid or gid is negative. A case of a negative gid caused this to be
2666 noticed. The fix allows for either to be negative.
2668 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2669 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2672 PH/36 The change PH/12 above was broken. Fixed it.
2674 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2675 the grounds that routing the same address twice would always produce the
2676 same answer. This might have been true once, but it is certainly no
2677 longer true now. Routing a child address may depend on the previous
2678 routing that produced that child. Some complicated redirection strategies
2679 went wrong when messages had multiple recipients, and made Exim's
2680 behaviour dependent on the order in which the addresses were given.
2682 I have moved the duplicate checking until after the routing is complete.
2683 Exim scans the addresses that are assigned to local and remote
2684 transports, and removes any duplicates. This means that more work will be
2685 done, as duplicates will always all be routed, but duplicates are
2686 presumably rare, so I don't expect this is of any significance.
2688 For deliveries to pipes, files, and autoreplies, the duplicate checking
2689 still happens during the routing process, since they are not going to be
2692 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2693 It corrects a timeout issue with spamd. This is Ian's comment: "The
2694 background is that sometimes spamd either never reads data from a
2695 connection it has accepted, or it never writes response data. The exiscan
2696 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2697 blindly assumes that writes won't block so it may never time out."
2699 PH/39 Allow G after quota size as well as K and M.
2701 PH/40 The value set for $authenticated_id in an authenticator may not contain
2702 binary zeroes or newlines because the value is written to log lines and
2703 to spool files. There was no check on this. Now the value is run through
2704 the string_printing() function so that such characters are converted to
2705 printable escape sequences.
2707 PH/41 $message_linecount is a new variable that contains the total number of
2708 lines in the message. Compare $body_linecount, which is the count for the
2711 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2712 bounce and delay warning messages, except in certain special cases, which
2715 (a) An SMTP error message from a remote host;
2716 (b) A message specified in a :fail: redirection;
2717 (c) A message specified in a "fail" command in a system filter;
2718 (d) A message specified in a FAIL return from the queryprogram router;
2719 (e) A message specified by the cannot_route_message router option.
2721 In these cases only, Exim does include the error details in bounce and
2722 warning messages. There are also a few cases where bland messages such
2723 as "unrouteable address" or "local delivery error" are given.
2725 PH/43 $value is now also set for the "else" part of a ${run expansion.
2727 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2728 being worked on, but at least Exim now implements the latest version to
2731 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2732 process to complete was treated as a delivery failure, a timeout while
2733 writing the message to the pipe was logged, but erroneously treated as a
2734 successful delivery. Such timeouts include transport filter timeouts. For
2735 consistency with the overall process timeout, these timeouts are now
2736 treated as errors, giving rise to delivery failures by default. However,
2737 there is now a new Boolean option for the pipe transport called
2738 timeout_defer, which, if set TRUE, converts the failures into defers for
2739 both kinds of timeout. A transport filter timeout is now identified in
2742 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2743 systems where "make" and "gmake" are different, calling "gmake" at top
2744 level broke things. I've arranged for the value of $(MAKE) to be passed
2745 from the Makefile to this script so that it can call the same version of
2749 A note about Exim versions 4.44 and 4.50
2750 ----------------------------------------
2752 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2753 changes of various kinds. As a consequence, a big documentation update was
2754 needed. This delayed the release for rather longer than seemed good, especially
2755 in the light of a couple of (minor) security issues. Therefore, the changes
2756 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2757 release. So 4.44 and 4.50 are in effect two different branches that both start
2760 I have left the 4.50 change log unchanged; it contains all the changes since
2761 4.43. The change log for 4.44 is below; many of its items are identical to
2762 those for 4.50. This seems to be the most sensible way to preserve the
2763 historical information.
2769 1. Minor wording change to the doc/README.SIEVE file.
2771 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2772 computation of the current number of files was incorrect.
2774 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2775 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2778 4. Give more explanation in the error message when the command for a transport
2779 filter fails to execute.
2781 5. There are several places where Exim runs a non-Exim command in a
2782 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2783 was being done only for the command run by the queryprogram router. It is
2784 now done for all such subprocesses. The other cases are: ${run, transport
2785 filters, and the commands run by the lmtp and pipe transports.
2787 6. Added CONFIGURE_GROUP build-time option.
2789 7. Some older OS have a limit of 256 on the maximum number of file
2790 descriptors. Exim was using setrlimit() to set 1000 as a large value
2791 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2792 systems. I've change it so that if it can't get 1000, it tries for 256.
2794 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2795 was an oversight, and furthermore, ever since the addition of extra
2796 controls (e.g. 4.43/32), the checks on when to allow different forms of
2797 "control" were broken. There should now be diagnostics for all cases when a
2798 control that does not make sense is encountered.
2800 9. Added the /retain_sender option to "control=submission".
2802 10. $recipients is now available in the predata ACL (oversight).
2804 11. Tidy the search cache before the fork to do a delivery from a message
2805 received from the command line. Otherwise the child will trigger a lookup
2806 failure and thereby defer the delivery if it tries to use (for example) a
2807 cached ldap connection that the parent has called unbind on.
2809 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2810 of $address_data from the recipient verification was clobbered by the
2811 sender verification.
2813 13. The value of address_data from a sender verification is now available in
2814 $sender_address_data in subsequent conditions in the ACL statement.
2816 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2818 15. Added a new option "connect=<time>" to callout options, to set a different
2821 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2822 was its contents. (It was OK if the option was not defined at all.)
2824 17. A "Completed" log line is now written for messages that are removed from
2825 the spool by the -Mrm option.
2827 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2828 information about exactly what failed.
2830 19. Added -dd to debug only the daemon process.
2832 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2833 handles timeouts, both on the server side and network timeouts. Renamed the
2834 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2836 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2837 It is now set to "smtps".
2839 22. $host_address is now set to the target address during the checking of
2840 ignore_target_hosts.
2842 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2843 being passed; this would have caused $sender_host_name to have been used if
2844 matching the list had actually called for a host name (not very likely,
2845 since this list is usually IP addresses). A host name is now passed as
2848 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2849 code that specifies a non-restarting handler (typically sigaction() in
2850 modern systems) in an attempt to fix a rare and obscure crash bug.
2852 25. Narrowed the window for a race in the daemon that could cause it to ignore
2853 SIGCHLD signals. This is not a major problem, because they are used only to
2854 wake it up if nothing else does.
2856 26. A malformed maildirsize file could cause Exim to calculate negative values
2857 for the mailbox size or file count. Odd effects could occur as a result.
2858 The maildirsize information is now recalculated if the size or filecount
2861 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2862 support for a long time. Removed HAVE_SYS_VFS_H.
2864 28. Installed the latest version of exipick from John Jetmore.
2866 29. In an address list, if the pattern was not a regular expression, an empty
2867 subject address (from a bounce message) matched only if the pattern was an
2868 empty string. Non-empty patterns were not even tested. This was the wrong
2869 because it is perfectly reasonable to use an empty address as part of a
2870 database query. An empty address is now tested by patterns that are
2871 lookups. However, all the other forms of pattern expect the subject to
2872 contain a local part and a domain, and therefore, for them, an empty
2873 address still always fails if the pattern is not itself empty.
2875 30. Exim went into a mad DNS loop when attempting to do a callout where the
2876 host was specified on an smtp transport, and looking it up yielded more
2877 than one IP address.
2879 31. Re-factored the code for checking spool and log partition space into a
2880 function that finds that data and another that does the check. The former
2881 is then used to implement four new variables: $spool_space, $log_space,
2882 $spool_inodes, and $log_inodes.
2884 32. The RFC2047 encoding function was originally intended for short strings
2885 such as real names; it was not keeping to the 75-character limit for
2886 encoded words that the RFC imposes. It now respects the limit, and
2887 generates multiple encoded words if necessary. To be on the safe side, I
2888 have increased the buffer size for the ${rfc2047: expansion operator from
2891 33. It is now permitted to omit both strings after an "if" condition; if the
2892 condition is true, the result is "true". As before, when the second string
2893 is omitted, a false condition yields an empty string. This makes it less
2894 cumbersome to write custom ACL and router conditions.
2896 34. Failure to deliver a bounce message always caused it to be frozen, even if
2897 there was an errors_to setting on the router. The errors_to setting is now
2900 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2901 canonical form (fully expanded) before being placed in
2902 $sender_host_address.
2904 36. The table in the code that translates DNS record types into text (T_A to
2905 "A" for instance) was missing entries for NS and CNAME. It is just possible
2906 that this could have caused confusion if both these types were looked up
2907 for the same domain, because the text type is used as part of Exim's
2908 per-process caching. But the chance of anyone hitting this buglet seems
2911 37. The dnsdb lookup has been extended in a number of ways.
2913 (1) There is a new type, "zns", which walks up the domain tree until it
2914 finds some nameserver records. It should be used with care.
2916 (2) There is a new type, "mxh", which is like "mx" except that it returns
2917 just the host names, not the priorities.
2919 (3) It is now possible to give a list of domains (or IP addresses) to be
2920 looked up. The behaviour when one of the lookups defers can be
2921 controlled by a keyword.
2923 (4) It is now possible to specify the separator character for use when
2924 multiple records are returned.
2926 38. The dnslists ACL condition has been extended: it is now possible to supply
2927 a list of IP addresses and/or domains to be looked up in a particular DNS
2930 39. Added log_selector=+queue_time_overall.
2932 40. When running the queue in the test harness, wait just a tad after forking a
2933 delivery process, to get repeatability of debugging output.
2935 41. Include certificate and key file names in error message when GnuTLS fails
2936 to set them up, because the GnuTLS error message doesn't include the name
2937 of the failing file when there is a problem reading it.
2939 42. Allow both -bf and -bF in the same test run.
2941 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2943 44. The "Exiscan patch" is now merged into the mainline Exim source.
2945 45. Sometimes the final signoff response after QUIT could fail to get
2946 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2947 < 0 before doing a fflush(). This bug looks as though it goes back to the
2948 introduction of TLS in release 3.20, but "sometimes" must have been rare
2949 because the tests only now provoked it.
2951 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2952 (this can affect the format of dates).
2954 47. exim_tidydb, when checking for the continued existence of a message for
2955 which it has found a message-specific retry record, was not finding
2956 messages that were in split spool directories. Consequently, it was
2957 deleting retry records that should have stayed in existence.
2959 48. Steve fixed some bugs in eximstats.
2961 49. The SPA authentication driver was not abandoning authentication and moving
2962 on to the next authenticator when an expansion was forced to fail,
2963 contradicting the general specification for all authenticators. Instead it
2964 was generating a temporary error. It now behaves as specified.
2966 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2967 (the order specifies the preference for clients). The order is now AES256,
2968 AES128, 3DES, ARCFOUR128.
2970 51. Small patch to Sieve code - explicitly set From: when generating an
2973 52. Exim crashed if a remote delivery caused a very long error message to be
2974 recorded - for instance if somebody sent an entire SpamAssassin report back
2975 as a large number of 550 error lines. This bug was coincidentally fixed by
2976 increasing the size of one of Exim's internal buffers (big_buffer) that
2977 happened as part of the Exiscan merge. However, to be on the safe side, I
2978 have made the code more robust (and fixed the comments that describe what
2981 53. Now that there can be additional text after "Completed" in log lines (if
2982 the queue_time_overall log selector is set), a one-byte patch to exigrep
2983 was needed to allow it to recognize "Completed" as not the last thing in
2986 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2987 patch that reportedly fixes this has been added. I am not expert enough to
2988 create a test for it. This is what the patch creator wrote:
2990 "I found a little strange behaviour of ldap code when working with
2991 Windows 2003 AD Domain, where users was placed in more than one
2992 Organization Units. When I tried to give exim partial DN, the exit code
2993 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2994 But simultaneously result of request was absolutely normal ldap result,
2995 so I produce this patch..."
2997 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2998 REFERENCE, so I have modified the code to exclude the patch when that macro
3001 55. Some experimental protocols are using DNS PTR records for new purposes. The
3002 keys for these records are domain names, not reversed IP addresses. The
3003 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3004 leaves it alone. Component reversal etc. now happens only for IP addresses.
3007 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3009 57. Double the size of the debug message buffer (to 2048) so that more of very
3010 long debug lines gets shown.
3012 58. The exicyclog utility now does better if the number of log files to keep
3013 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3015 59. Two changes related to the smtp_active_hostname option:
3017 (1) $smtp_active_hostname is now available as a variable.
3018 (2) The default for smtp_banner uses $smtp_active_hostname instead
3019 of $primary_hostname.
3021 60. The host_aton() function is supposed to be passed a string that is known
3022 to be a valid IP address. However, in the case of IPv6 addresses, it was
3023 not checking this. This is a hostage to fortune. Exim now panics and dies
3024 if the condition is not met. A case was found where this could be provoked
3025 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3026 components; fortuitously, this particular loophole had already been fixed
3027 by change 4.50/55 above.
3029 If there are any other similar loopholes, the new check in host_aton()
3030 itself should stop them being exploited. The report I received stated that
3031 data on the command line could provoke the exploit when Exim was running as
3032 exim, but did not say which command line option was involved. All I could
3033 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3034 running as the user.
3037 61. There was a buffer overflow vulnerability in the SPA authentication code
3038 (which came originally from the Samba project). I have added a test to the
3039 spa_base64_to_bits() function which I hope fixes it.
3042 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3043 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3045 63. The daemon start-up calls getloadavg() while still root for those OS that
3046 need the first call to be done as root, but it missed one case: when
3047 deliver_queue_load_max is set with deliver_drop_privilege. This is
3048 necessary for the benefit of the queue runner, because there is no re-exec
3049 when deliver_drop_privilege is set.
3051 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3052 This has been fixed.
3054 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3055 was in use, was not putting the data itself into the right store pool;
3056 consequently, it could be overwritten for a subsequent message in the same
3057 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3060 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3062 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3063 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3064 the function were treating the return as a boolean value, which happened to
3065 work because 0=false and not-0=true, but is not correct code.
3067 68. The host_aton() function was not handling scoped IPv6 addresses (those
3068 with, for example, "%eth0" on the end) correctly.
3070 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3071 negated items (that is, ~something) in unsigned ints. Some compilers
3072 apparently mutter when there is no cast.
3074 70. If an address verification called from an ACL failed, and did not produce a
3075 user-specific message (i.e. there was only a "system" message), nothing was
3076 put in $acl_verify_message. In this situation, it now puts the system
3079 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3080 session; change 4.31/43 added the unwanted input to the log line - except
3081 that it did not do this in the start of session case. It now does.
3083 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3084 This is wrong and can cause the other end to generate a synchronization
3085 error if it is another Exim or anything else that does the synchronization
3086 check. A QUIT command is no longer sent after a timeout.
3088 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3089 during host lookups.
3091 74. The defer_ok option of callout verification was not working if it was used
3092 when verifying addresses in header lines, that is, for this case:
3094 verify = header_sender/callout=defer_ok
3096 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3097 those file descriptors could be used for SMTP connections. If anything
3098 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3099 could be sent to the SMTP client, causing chaos. The daemon now opens
3100 stdin, stdout, and stderr to /dev/null when it puts itself into the
3103 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3104 log by default. The user can override this with suitable Perl magic.
3106 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3107 the log message when discard triggers, was not working for the DATA ACL or
3108 for the non-SMTP ACL.
3110 78. Error message wording change in sieve.c.
3112 79. If smtp_accept_max_per_host was set, the number of connections could be
3113 restricted to fewer than expected, because the daemon was trying to set up
3114 a new connection before checking whether the processes handling previous
3115 connections had finished. The check for completed processes is now done
3116 earlier. On busy systems, this bug wouldn't be noticed because something
3117 else would have woken the daemon, and it would have reaped the completed
3120 80. If a message was submitted locally by a user whose login name contained one
3121 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3122 It caused a spool format error. I have fixed the spool reading code. A
3123 related problem was that the "from" clause in the Received: line became
3124 illegal because of the space(s). It is now covered by ${quote_local_part.
3126 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3129 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3131 83. Patch from Sieve maintainer to fix unterminated string problem in
3132 "vacation" handling.
3134 84. Some minor changes to the Linux configuration files to help with other
3135 OS variants using glibc.
3137 85. One more patch for Sieve to update vacation handling to latest spec.
3140 ----------------------------------------------------
3141 See the note above about the 4.44 and 4.50 releases.
3142 ----------------------------------------------------
3148 1. Change 4.43/35 introduced a bug that caused file counts to be
3149 incorrectly computed when quota_filecount was set in an appendfile
3152 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3153 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3156 3. Give more explanation in the error message when the command for a transport
3157 filter fails to execute.
3159 4. There are several places where Exim runs a non-Exim command in a
3160 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3161 was being done only for the command run by the queryprogram router. It is
3162 now done for all such subprocesses. The other cases are: ${run, transport
3163 filters, and the commands run by the lmtp and pipe transports.
3165 5. Some older OS have a limit of 256 on the maximum number of file
3166 descriptors. Exim was using setrlimit() to set 1000 as a large value
3167 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3168 systems. I've change it so that if it can't get 1000, it tries for 256.
3170 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3171 was an oversight, and furthermore, ever since the addition of extra
3172 controls (e.g. 4.43/32), the checks on when to allow different forms of
3173 "control" were broken. There should now be diagnostics for all cases when a
3174 control that does not make sense is encountered.
3176 7. $recipients is now available in the predata ACL (oversight).
3178 8. Tidy the search cache before the fork to do a delivery from a message
3179 received from the command line. Otherwise the child will trigger a lookup
3180 failure and thereby defer the delivery if it tries to use (for example) a
3181 cached ldap connection that the parent has called unbind on.
3183 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3184 of $address_data from the recipient verification was clobbered by the
3185 sender verification.
3187 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3188 was its contents. (It was OK if the option was not defined at all.)
3190 11. A "Completed" log line is now written for messages that are removed from
3191 the spool by the -Mrm option.
3193 12. $host_address is now set to the target address during the checking of
3194 ignore_target_hosts.
3196 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3197 being passed; this would have caused $sender_host_name to have been used if
3198 matching the list had actually called for a host name (not very likely,
3199 since this list is usually IP addresses). A host name is now passed as
3202 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3203 code that specifies a non-restarting handler (typically sigaction() in
3204 modern systems) in an attempt to fix a rare and obscure crash bug.
3206 15. Narrowed the window for a race in the daemon that could cause it to ignore
3207 SIGCHLD signals. This is not a major problem, because they are used only to
3208 wake it up if nothing else does.
3210 16. A malformed maildirsize file could cause Exim to calculate negative values
3211 for the mailbox size or file count. Odd effects could occur as a result.
3212 The maildirsize information is now recalculated if the size or filecount
3215 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3216 support for a long time. Removed HAVE_SYS_VFS_H.
3218 18. Updated exipick to current release from John Jetmore.
3220 19. Allow an empty sender to be matched against a lookup in an address list.
3221 Previously the only cases considered were a regular expression, or an
3224 20. Exim went into a mad DNS lookup loop when doing a callout where the
3225 host was specified on the transport, if the DNS lookup yielded more than
3228 21. The RFC2047 encoding function was originally intended for short strings
3229 such as real names; it was not keeping to the 75-character limit for
3230 encoded words that the RFC imposes. It now respects the limit, and
3231 generates multiple encoded words if necessary. To be on the safe side, I
3232 have increased the buffer size for the ${rfc2047: expansion operator from
3235 22. Failure to deliver a bounce message always caused it to be frozen, even if
3236 there was an errors_to setting on the router. The errors_to setting is now
3239 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3240 canonical form (fully expanded) before being placed in
3241 $sender_host_address.
3243 24. Updated eximstats to version 1.33
3245 25. Include certificate and key file names in error message when GnuTLS fails
3246 to set them up, because the GnuTLS error message doesn't include the name
3247 of the failing file when there is a problem reading it.
3249 26. Expand error message when OpenSSL has problems setting up cert/key files.
3252 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3253 (this can affect the format of dates).
3255 28. exim_tidydb, when checking for the continued existence of a message for
3256 which it has found a message-specific retry record, was not finding
3257 messages that were in split spool directories. Consequently, it was
3258 deleting retry records that should have stayed in existence.
3260 29. eximstats updated to version 1.35
3261 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3262 1.35 - bugfix such that pie charts by volume are generated correctly
3264 30. The SPA authentication driver was not abandoning authentication and moving
3265 on to the next authenticator when an expansion was forced to fail,
3266 contradicting the general specification for all authenticators. Instead it
3267 was generating a temporary error. It now behaves as specified.
3269 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3270 (the order specifies the preference for clients). The order is now AES256,
3271 AES128, 3DES, ARCFOUR128.
3273 31. Small patch to Sieve code - explicitly set From: when generating an
3276 32. Exim crashed if a remote delivery caused a very long error message to be
3277 recorded - for instance if somebody sent an entire SpamAssassin report back
3278 as a large number of 550 error lines. This bug was coincidentally fixed by
3279 increasing the size of one of Exim's internal buffers (big_buffer) that
3280 happened as part of the Exiscan merge. However, to be on the safe side, I
3281 have made the code more robust (and fixed the comments that describe what
3284 33. Some experimental protocols are using DNS PTR records for new purposes. The
3285 keys for these records are domain names, not reversed IP addresses. The
3286 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3287 leaves it alone. Component reversal etc. now happens only for IP addresses.
3290 34. The host_aton() function is supposed to be passed a string that is known
3291 to be a valid IP address. However, in the case of IPv6 addresses, it was
3292 not checking this. This is a hostage to fortune. Exim now panics and dies
3293 if the condition is not met. A case was found where this could be provoked
3294 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3295 components; fortuitously, this particular loophole had already been fixed
3296 by change 4.50/55 or 4.44/33 above.
3298 If there are any other similar loopholes, the new check in host_aton()
3299 itself should stop them being exploited. The report I received stated that
3300 data on the command line could provoke the exploit when Exim was running as
3301 exim, but did not say which command line option was involved. All I could
3302 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3303 running as the user.
3306 35. There was a buffer overflow vulnerability in the SPA authentication code
3307 (which came originally from the Samba project). I have added a test to the
3308 spa_base64_to_bits() function which I hope fixes it.
3311 36. The daemon start-up calls getloadavg() while still root for those OS that
3312 need the first call to be done as root, but it missed one case: when
3313 deliver_queue_load_max is set with deliver_drop_privilege. This is
3314 necessary for the benefit of the queue runner, because there is no re-exec
3315 when deliver_drop_privilege is set.
3317 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3318 was in use, was not putting the data itself into the right store pool;
3319 consequently, it could be overwritten for a subsequent message in the same
3320 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3323 38. Sometimes the final signoff response after QUIT could fail to get
3324 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3325 < 0 before doing a fflush(). This bug looks as though it goes back to the
3326 introduction of TLS in release 3.20, but "sometimes" must have been rare
3327 because the tests only now provoked it.
3333 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3334 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3335 Now it returns an integer. A number of calls were still expecting a T/F
3336 return. Fortuitously, in all cases, the tests worked in OK situations,
3337 which is the norm. However, things would have gone wrong on any write
3338 failures on the smtp file descriptor. This function is used when sending
3339 messages over SMTP and also when doing verify callouts.
3341 2. When Exim is called to do synchronous delivery of a locally submitted
3342 message (the -odf or -odi options), it no longer closes stderr before doing
3345 3. Implemented the mua_wrapper option.
3347 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3349 5. Implemented the functions header_remove(), header_testname(),
3350 header_add_at_position(), and receive_remove_recipient(), and exported them
3353 6. If an ACL "warn" statement specified the addition of headers, Exim already
3354 inserted X-ACL-Warn: at the start if there was no header name. However, it
3355 was not making this test for the second and subsequent header lines if
3356 there were newlines in the string. This meant that an invalid header could
3357 be inserted if Exim was badly configured.
3359 7. Allow an ACL "warn" statement to add header lines at the start or after all
3360 the Received: headers, as well as at the end.
3362 8. Added the rcpt_4xx retry error code.
3364 9. Added postmaster_mailfrom=xxx to callout verification option.
3366 10. Added mailfrom=xxxx to the callout verification option, for verify=
3369 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3370 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3372 12. Inserted some casts to stop certain compilers warning when using pointer
3373 differences as field lengths or precisions in printf-type calls (mostly
3374 affecting debugging statements).
3376 13. Added optional readline() support for -be (dynamically loaded).
3378 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3379 same clock tick as a message's arrival, so that its received time was the
3380 same as the "first fail" time on the retry record, and that message
3381 remained on the queue past the ultimate address timeout, every queue runner
3382 would try a delivery (because it was past the ultimate address timeout) but
3383 after another failure, the ultimate address timeout, which should have then
3384 bounced the address, did not kick in. This was a "< instead of <=" error;
3385 in most cases the first failure would have been in the next clock tick
3386 after the received time, and all would be well.
3388 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3389 being recognized when the domain list was tested by the match_domain
3390 condition in an expansion string.
3392 16. Added the ${str2b64: operator.
3394 17. Exim was always calling setrlimit() to set a large limit for the number of
3395 processes, without checking whether the existing limit was already
3396 adequate. (It did check for the limit on file descriptors.) Furthermore,
3397 errors from getrlimit() and setrlimit() were being ignored. Now they are
3398 logged to the main and panic logs, but Exim does carry on, to try to do its
3399 job under whatever limits there are.
3401 18. Imported PCRE 5.0.
3403 19. Trivial typo in log message " temporarily refused connection" (the leading
3406 20. If the log selector return_path_on_delivery was set and an address was
3407 redirected to /dev/null, the delivery process crashed because it assumed
3408 that a return path would always be set for a "successful" delivery. In this
3409 case, the whole delivery is bypassed as an optimization, and therefore no
3412 21. Internal re-arrangement: the function for sending a challenge and reading
3413 a response while authentication was assuming a zero-terminated challenge
3414 string. It's now changed to take a pointer and a length, to allow for
3415 binary data in such strings.
3417 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3419 23. Exim was not respecting finduser_retries when seeking the login of the
3420 uid under which it was called; it was always trying 10 times. (The default
3421 setting of finduser_retries is zero.) Also, it was sleeping after the final
3422 failure, which is pointless.
3424 24. Implemented tls_on_connect_ports.
3426 25. Implemented acl_smtp_predata.
3428 26. If the domain in control=submission is set empty, Exim assumes that the
3429 authenticated id is a complete email address when it generates From: or
3430 Sender: header lines.
3432 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3433 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3434 chown and chgrp in /bin and hostname in /usr/bin.
3436 28. Exim was keeping the "process log" file open after each use, just as it
3437 does for the main log. This opens the possibility of it remaining open for
3438 long periods when the USR1 signal hits a daemon. Occasional processlog
3439 errors were reported, that could have been caused by this. Anyway, it seems
3440 much more sensible not to leave this file open at all, so that is what now
3443 29. The long-running daemon process does not normally write to the log once it
3444 has entered its main loop, and it closes the log before doing so. This is
3445 so that log files can straightforwardly be renamed and moved. However,
3446 there are a couple of unusual error situations where the daemon does write
3447 log entries, and I had neglected to close the log afterwards.
3449 30. The text of an SMTP error response that was received during a remote
3450 delivery was being truncated at 512 bytes. This is too short for some of
3451 the long messages that one sometimes sees. I've increased the limit to
3454 31. It is now possible to make retry rules that apply only when a message has a
3455 specific sender, in particular, an empty sender.
3457 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3458 it possible to be selective about when SMTP synchronization is enforced.
3460 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3462 32. Implemented hosts_connection_nolog.
3464 33. Added an ACL for QUIT.
3466 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3469 35. Added mailbox_size and mailbox_filecount to appendfile.
3471 36. Added control = no_multiline_responses to ACLs.
3473 37. There was a bug in the logic of the code that waits for the clock to tick
3474 in the case where the clock went backwards by a substantial amount such
3475 that the microsecond fraction of "now" was more than the microsecond
3476 fraction of "then" (but the whole seconds number was less).
3478 38. Added support for the libradius Radius client library this is found on
3479 FreeBSD (previously only the radiusclient library was supported).
3485 1. When certain lookups returned multiple values in the form name=value, the
3486 quoting of the values was not always being done properly. Specifically:
3487 (a) If the value started with a double quote, but contained no whitespace,
3489 (b) If the value contained whitespace other than a space character (i.e.
3490 tabs or newlines or carriage returns) it was not quoted.
3491 This fix has been applied to the mysql and pgsql lookups by writing a
3492 separate quoting function and calling it from the lookup code. The fix
3493 should probably also be applied to nisplus, ibase and oracle lookups, but
3494 since I cannot test any of those, I have not disturbed their existing code.
3496 2. A hit in the callout cache for a specific address caused a log line with no
3497 reason for rejecting RCPT. Now it says "Previous (cached) callout
3498 verification failure".
3500 3. There was an off-by-one bug in the queryprogram router. An over-long
3501 return line was truncated at 256 instead of 255 characters, thereby
3502 overflowing its buffer with the terminating zero. As well as fixing this, I
3503 have increased the buffer size to 1024 (and made a note to document this).
3505 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3506 when Exim is waiting for an SMTP response from a remote server, Exim
3507 restarts its select() call on the socket, thereby resetting its timeout.
3508 This is not a problem when such interrupts are rare. Somebody set up a cron
3509 job to run exiwhat every 2 minutes, which is less than the normal select()
3510 timeout (5 or 10 minutes). This meant that the select() timeout never
3511 kicked in because it was always reset. I have fixed this by comparing the
3512 time when an interrupt arrives with the time at the start of the first call
3513 to select(). If more time than the timeout has elapsed, the interrupt is
3514 treated as a timeout.
3516 5. Some internal re-factoring in preparation for the addition of Sieve
3517 extensions (by MH). In particular, the "personal" test is moved to a
3518 separate function, and given an option for scanning Cc: and Bcc: (which is
3519 not set for Exim filters).
3521 6. When Exim created an email address using the login of the caller as the
3522 local part (e.g. when creating a From: or Sender: header line), it was not
3523 quoting the local part when it contained special characters such as @.
3525 7. Installed new OpenBSD configuration files.
3527 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3528 try to make them clearer.
3530 9. Callout options, other than the timeout value, were being ignored when
3531 verifying sender addresses in header lines. For example, when using
3533 verify = header_sender/callout=no_cache
3535 the cache was (incorrectly) being used.
3537 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3538 only the Cygwin environment.
3540 11. When return_path_on_delivery was set as a log selector, if different remote
3541 addresses in the same message used different return paths and parallel
3542 remote delivery occurred, the wrong values would sometimes be logged.
3543 (Whenever a remote delivery process finished, the return path value from
3544 the most recently started remote delivery process was logged.)
3546 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3547 lines when AUTH and/or TLS are in use. This is the "received protocol"
3548 field. Exim used to use "asmtp" for authenticated SMTP, without any
3549 indication (in the protocol name) for TLS use. Now it follows the RFC and
3550 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3551 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3552 names appear in log lines as well as in Received: header lines.
3554 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3555 extensions, and comparison tests, and to fix some bugs.
3557 14. Changes to the "personal" filter test:
3559 (1) The test was buggy in that it was just doing the equivalent of
3560 "contains" tests on header lines. For example, if a user's address was
3561 anne@some.where, the "personal" test would incorrectly be true for
3563 To: susanne@some.where
3565 This test is now done by extracting each address from the header in turn,
3566 and checking the entire address. Other tests that are part of "personal"
3567 are now done using regular expressions (for example, to check local parts
3568 of addresses in From: header lines).
3570 (2) The list of non-personal local parts in From: addresses has been
3571 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3572 taken from the Sieve specification recommendations.
3574 (3) If the message contains any header line starting with "List-" it is
3575 treated as non-personal.
3577 (4) The test for "circular" in the Subject: header line has been removed
3578 because it now seems ill-conceived.
3580 15. Minor typos in src/EDITME comments corrected.
3582 16. Installed latest exipick from John Jetmore.
3584 17. If headers_add on a router specified a text string that was too long for
3585 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3586 of string_sprintf() is now avoided.
3588 18. $message_body_size was not set (it was always zero) when running the DATA
3589 ACL and the local_scan() function.
3591 19. For the "mail" command in an Exim filter, no default was being set for
3592 the once_repeat time, causing a random time value to be used if "once" was
3593 specified. (If the value happened to be <= 0, no repeat happened.) The
3594 default is now 0s, meaning "never repeat". The "vacation" command was OK
3595 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3596 (I found it when inspecting the code).
3598 20. There is now an overall timeout for performing a callout verification. It
3599 defaults to 4 times the callout timeout, which applies to individual SMTP
3600 commands during the callout. The overall timeout applies when there is more
3601 than one host that can be tried. The timeout is checked before trying the
3602 next host. This prevents very long delays if there are a large number of
3603 hosts and all are timing out (e.g. when the network connections are timing
3604 out). The value of the overall timeout can be changed by specifying an
3605 additional sub-option for "callout", called "maxwait". For example:
3607 verify = sender/callout=5s,maxwait=20s
3609 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3610 to the end before writing, but this should make it even safer).
3612 22. Exim was forgetting that it had advertised PIPELINING for the second and
3613 subsequent messages on an SMTP connection. It was also not resetting its
3614 memory on STARTTLS and an internal HELO.
3616 23. When Exim logs an SMTP synchronization error within a session, it now
3617 records whether PIPELINING has been advertised or not.
3619 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3620 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3621 rather than long int.
3623 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3625 26. Added the never_mail option to autoreply.
3631 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3632 crash if the getsockname() call failed; this can happen if a connection is
3633 closed very soon after it is established. The problem was simply in the
3634 order in which certain operations were done, causing Exim to try to write
3635 to the SMTP stream before it had set up the file descriptor. The bug has
3636 been fixed by making things happen in the correct order.
3642 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3643 before the connection was closed, thus losing the rejection response.
3645 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3646 some early Solaris releases, but causes trouble in current releases where
3647 socklen_t is defined.
3649 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3652 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3655 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3656 that releasing the top of it at the end releases what was used for sub-
3657 expansions (unless the block got too big). However, discard this block if
3658 the first thing is a variable or header, so that we can use its block when
3659 it is dynamic (useful for very large $message_headers, for example).
3661 6. Lookups now cache *every* query, not just the most recent. A new, separate
3662 store pool is used for this. It can be recovered when all lookup caches are
3663 flushed. Lookups now release memory at the end of their result strings.
3664 This has involved some general refactoring of the lookup sources.
3666 7. Some code has been added to the store_xxx() functions to reduce the amount
3667 of flapping under certain conditions.
3669 8. log_incoming_interface used to affect only the <= reception log lines. Now
3670 it causes the local interface and port to be added to several more SMTP log
3671 lines, for example "SMTP connection from", and rejection lines.
3673 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3675 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3677 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3678 could be overwritten at the end of the current message (or the start of a
3679 new message if it was set in a HELO ACL). The value is now preserved for
3680 the duration of the SMTP connection.
3682 12. If a transport had a headers_rewrite setting, and a matching header line
3683 contained an unqualified address, that address was qualified, even if it
3684 did not match any rewriting rules. The underlying bug was that the values
3685 of the flags that permit the existence of unqualified sender and recipient
3686 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3687 non-local messages, and by -bnq for local messages) were not being
3688 preserved with the message after it was received.
3690 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3691 "next input=" as part of the text comprising the host identity instead of
3692 the correct text. The code was using the same buffer for two different
3693 strings. However, depending on which order the printing function evaluated
3694 its arguments, the bug did not always show up. Under Linux, for example, my
3695 test suite worked just fine.
3697 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3698 doesn't work with some older versions of Perl. It has been changed to "my",
3699 which in any case is probably the better facility to use.
3701 15. A really picky compiler found some instances of statements for creating
3702 error messages that either had too many or two few arguments for the format
3705 16. The size of the buffer for calls to the DNS resolver has been increased
3706 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3707 for addresses that have a lot of PTR records. This alleviates a problem; it
3708 does not fully solve it.
3710 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3711 buffer now truncates the list and logs the incident, which is the same
3712 action as happens when Exim is looking up a host name and its aliases.
3713 Previously in this situation something unpredictable would happen;
3714 sometimes it was "internal error: store_reset failed".
3716 18. If a server dropped the connection unexpectedly when an Exim client was
3717 using GnuTLS and trying to read a response, the client delivery process
3718 crashed while trying to generate an error log message.
3720 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3721 string, for example:
3723 warn message = H1: something\nH2: something
3725 the text was added as a single header line from Exim's point of view
3726 though it ended up OK in the delivered message. However, searching for the
3727 second and subsequent header lines using $h_h2: did not work. This has been
3728 fixed. Similarly, if a system filter added multiple headers in this way,
3729 the routers could not see them.
3731 20. Expanded the error message when iplsearch is called with an invalid key to
3732 suggest using net-iplsearch in a host list.
3734 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3735 ACLs are no longer actually imposed (and a message to that effect is
3738 22. If a "gecos" field in a passwd entry contained escaped characters, in
3739 particular, if it contained a \" sequence, Exim got it wrong when building
3740 a From: or a Sender: header from that name. A second bug also caused
3741 incorrect handling when an unquoted " was present following a character
3742 that needed quoting.
3744 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3745 was not being matched caselessly.
3747 24. Arranged for all hyphens in the exim.8 source to be escaped with
3750 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3751 properly. Recipient callout cache records were still being keyed to include
3752 the sender, even when use_sender was set false. This led to far more
3753 callouts that were necessary. The sender is no longer included in the key
3754 when use_sender is false.
3756 26. Added "control = submission" modifier to ACLs.
3758 27. Added the ${base62d: operator to decode base 62 numbers.
3760 28. dnsdb lookups can now access SRV records.
3762 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3763 the configuration file.
3765 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3766 (-v) mode. This makes the output for a verbose queue run more intelligible.
3768 31. Added a use_postmaster feature to recipient callouts.
3770 32. Added the $body_zerocount variable, containing the number of binary zero
3771 bytes in the message body.
3773 33. The time of last modification of the "new" subdirectory is now used as the
3774 "mailbox time last read" when there is a quota error for a maildir
3777 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3779 35. Added +ignore_unknown as a special item in host lists.
3781 36. Code for decoding IPv6 addresses in host lists is now included, even if
3782 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3783 address was recognized as an IP address, but was then not correctly decoded
3784 into binary, causing unexpected and incorrect effects when compared with
3791 1. Very minor rewording of debugging text in manualroute to say "list of
3792 hosts" instead of "hostlist".
3794 2. If verify=header_syntax was set, and a header line with an unqualified
3795 address (no domain) and a large number of spaces between the end of the
3796 name and the colon was received, the reception process suffered a buffer
3797 overflow, and (when I tested it) crashed. This was caused by some obsolete
3798 code that should have been removed. The fix is to remove it!
3800 3. When running in the test harness, delay a bit after writing a bounce
3801 message to get a bit more predictability in the log output.
3803 4. Added a call to search_tidyup() just before forking a reception process. In
3804 theory, someone could use a lookup in the expansion of smtp_accept_max_
3805 per_host which, without the tidyup, could leave open a database connection.
3807 5. Added the variables $recipient_data and $sender_data which get set from a
3808 lookup success in an ACL "recipients" or "senders" condition, or a router
3809 "senders" option, similar to $domain_data and $local_part_data.
3811 6. Moved the writing of debug_print from before to after the "senders" test
3814 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3815 problems for message scanning, either using a data ACL, or using
3816 local_scan() because the Received: header was not generated till after they
3817 were called (in order to set the time as the time of reception completion).
3818 I have revised the way this works. The header is now generated after the
3819 body is received, but before the ACL or local_scan() are called. After they
3820 are run, the timestamp in the header is updated.
3826 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3827 before starting a queue runner without re-exec. This happened only when
3828 deliver_drop_privilege was set or when the Exim user was set to root. The
3829 effect of the bug was that timeouts during subsequent deliveries caused
3830 crashes instead of being properly handled. The handler is now left at its
3831 default (and expected) setting.
3833 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3834 message, again when deliver_drop_privilege is set or Exim is run as root.
3835 The bug described in (1) was not present in this case, but the tidying up
3836 of the other signals was missing. I have made the two cases consistent.
3838 3. The ignore_target_hosts setting on a manualroute router was being ignored
3839 for hosts that were looked up using the /MX notation.
3841 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3844 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3845 operated on the sender address. After changing the $sender_address to <>
3846 for the sender address verify, Exim was re-instated it as the original
3847 (before rewriting) address, but remembering that it had rewritten it, so it
3848 wasn't rewriting it again. This bug also had the effect of breaking the
3849 sender address verification caching when the sender address was rewritten.
3851 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3852 This has been changed so that if the ip literal address matches
3853 ignore_target_hosts, the router declines.
3855 7. Added expansion conditions match_domain, match_address, and match_local_
3856 part (NOT match_host).
3858 8. The placeholder for the Received: header didn't have a length field set.
3860 9. Added code to Exim itself and to exim_lock to test for a specific race
3861 condition that could lead to file corruption when using MBX delivery. The
3862 issue is with the lockfile that is created in /tmp. If this file is removed
3863 after a process has opened it but before that process has acquired a lock,
3864 there is the potential for a second process to recreate the file and also
3865 acquire a lock. This could lead to two Exim processes writing to the file
3866 at the same time. The added code performs the same test as UW imapd; it
3867 checks after acquiring the lock that its file descriptor still refers to
3868 the same named file.
3870 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3871 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3874 11. Added the smtp_active_hostname option. If used, this will typically be made
3875 to depend on the incoming interface address. Because $interface_address is
3876 not set up until the daemon has forked a reception process, error responses
3877 that can happen earlier (such as "too many connections") no longer contain
3880 12. If an expansion in a condition on a "warn" statement fails because a lookup
3881 defers, the "warn" statement is abandoned, and the next ACL statement is
3882 processed. Previously this caused the whole ACL to be aborted.
3884 13. Added the iplsearch lookup type.
3886 14. Added ident_timeout as a log selector.
3888 15. Added tls_certificate_verified as a log selector.
3890 16. Added a global option tls_require_ciphers (compare the smtp transport
3891 option of the same name). This controls incoming TLS connections.
3893 17. I finally figured out how to make tls_require_ciphers do a similar thing
3894 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3895 before starting the TLS session.
3897 18. Tabs are now shown as \t in -bP output.
3899 19. If the log selector return_path_on_delivery was set, Exim crashed when
3900 bouncing a message because it had too many Received: header lines.
3902 20. If two routers both had headers_remove settings, and the first one included
3903 a superfluous trailing colon, the final name in the first list and the
3904 first name in the second list were incorrectly joined into one item (with a
3905 colon in the middle).
3911 1. Added -C and -D options to the exinext utility, mainly to make it easier
3912 to include in the automated testing, but these could be helpful when
3913 multiple configurations are in use.
3915 2. The exinext utility was not formatting the output nicely when there was
3916 an alternate port involved in the retry record key, nor when there was a
3917 message id as well (for retries that were specific to a specific message
3918 and a specific host). It was also confused by IPv6 addresses, because of
3919 the additional colons they contain. I have fixed the IPv4 problem, and
3920 patched it up to do a reasonable job for IPv6.
3922 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3923 delivery, the log line now contains "pipelined" if PIPELINING was used.
3925 4. An SMTP transport process used to panic and die if the bind() call to set
3926 an explicit outgoing interface failed. This has been changed; it is now
3927 treated in the same way as a connect() failure.
3929 5. A reference to $sender_host_name in the part of a conditional expansion
3930 that was being skipped was still causing a DNS lookup. This no longer
3933 6. The def: expansion condition was not recognizing references to header lines
3934 that used bh_ and bheader_.
3936 7. Added the _cache feature to named lists.
3938 8. The code for checking quota_filecount in the appendfile transport was
3939 allowing one more file than it should have been.
3941 9. For compatibility with Sendmail, the command line option
3949 and sets the incoming protocol and host name (for trusted callers). The
3950 host name and its colon can be omitted when only the protocol is to be set.
3951 Note the Exim already has two private options, -pd and -ps, that refer to
3952 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3953 "s", but I don't think that's a major issue.
3955 10. A number of refactoring changes to the code, none of which should affect
3958 (a) The number of logging options was getting close to filling up the
3959 32-bit word that was used as a bit map. I have split them into two classes:
3960 those that are passed in the argument to log_write(), and those that are
3961 only ever tested independently outside of that function. These are now in
3962 separate 32-bit words, so there is plenty of room for expansion again.
3963 There is no change in the user interface or the logging behaviour.
3965 (b) When building, for example, log lines, the code previously used a
3966 macro that called string_cat() twice, in order to add two strings. This is
3967 not really sufficiently general. Furthermore, there was one instance where
3968 it was actually wrong because one of the argument was used twice, and in
3969 one call a function was used. (As it happened, calling the function twice
3970 did not affect the overall behaviour.) The macro has been replaced by a
3971 function that can join an arbitrary number of extra strings onto a growing
3974 (c) The code for expansion conditions now uses a table and a binary chop
3975 instead of a serial search (which was left over from when there were very
3976 few conditions). Also, it now recognizes conditions like "pam" even when
3977 the relevant support is not compiled in: a suitably worded error message is
3978 given if an attempt is made to use such a condition.
3980 11. Added ${time_interval:xxxxx}.
3982 12. A bug was causing one of the ddress fields not to be passed back correctly
3983 from remote delivery subprocesses. The field in question was not being
3984 subsequently used, so this caused to problems in practice.
3986 13. Added new log selectors queue_time and deliver_time.
3988 14. Might have fixed a bug in maildirsizefile handling that threw up
3989 "unexpected character" debug warnings, and recalculated the data
3990 unnecessarily. In any case, I expanded the warning message to give more
3993 15. Added the message "Restricted characters in address" to the statements in
3994 the default ACL that block characters like @ and % in local parts.
3996 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3997 Three changes have been made:
3999 (a) There was a serious bug; a negative response to MAIL caused the whole
4000 recipient domain to be cached as invalid, thereby blocking all messages
4001 to all local parts at the same domain, from all senders. This bug has
4002 been fixed. The domain is no longer cached after a negative response to
4003 MAIL if the sender used is not empty.
4005 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4008 (c) A new callout option, "use_sender" has been added for people who want
4009 the modified behaviour.
4015 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4018 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4019 indeed breaks things for older releases.
4021 3. Added additional logging to the case where there is a problem reading data
4022 from a filter that is running in a subprocess using a pipe, in order to
4023 try to track down a specific problem.
4025 4. Testing facility fudge: when running in the test harness and attempting
4026 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4027 getting "No route to host". Convert this to a timeout.
4029 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4032 6. Some OS don't have socklen_t but use size_t instead. This affects the
4033 fifth argument of getsockopt() amongst other things. This is now
4034 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4035 can be set for individual OS. I have set it for SunOS5, OSF1, and
4036 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4037 some earlier ones do not.
4039 7. Change 4.30/15 was not doing the test caselessly.
4041 8. The standard form for an IPv6 address literal was being rejected by address
4042 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4043 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4044 this, as well as the form without the "IPv6" on the front (but only when
4045 address literals are enabled, of course).
4047 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4049 10. Exim crashed if a message with an empty sender address specified by -f
4050 encountered a router with an errors_to setting. This could be provoked only
4051 by a command such as
4055 where an empty string was supplied; "<>" did not hit this bug.
4057 11. Installed PCRE release 4.5.
4059 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4060 remained set. It is now erased.
4062 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4063 times from message ids (which are base 36 rather than the normal 62).
4065 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4066 were being counted as actual protocol errors, and logged if the log
4067 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4068 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4069 and DATA following a set of rejected RCPTs do not count as protocol errors.
4070 In other words, Exim assumes they were pipelined, though this may not
4071 actually be the case. Of course, in all cases the client gets an
4072 appropriate error code.
4074 15. If a lookup fails in an ACL condition, a message about the failure may
4075 be available; it is used if testing the ACL cannot continue, because most
4076 such messages specify what the cause of the deferral is. However, some
4077 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4078 that caused an old message to be retained and used if a later statement
4079 caused a defer, replacing the real cause of the deferral.
4081 16. If an IP address had so many PTR records that the DNS lookup buffer
4082 was not large enough to hold them, Exim could crash while trying to process
4083 the truncated data. It now detects and logs this case.
4085 17. Further to 4.21/58, another change has been made: if (and only if) the
4086 first line of a message (the first header line) ends with CRLF, a bare LF
4087 in a subsequent header line has a space inserted after it, so as not to
4088 terminate the header.
4090 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4091 unnecessarily, used atoi() instead of strtol(), and didn't check the
4092 termination when getting file sizes from file names by regex.
4094 19. Completely re-implemented the support for maildirsize files, in the light
4095 of a number of problems with the previous contributed implementation
4096 (4.30/29). In particular:
4098 . If the quota is zero, the maildirsize file is maintained, but no quota is
4101 . If the maildir directory does not exist, it is created before any attempt
4102 to write a maildirsize file.
4104 . The quota value in the file is just a cache; if the quota is changed in
4105 the transport, the new value overrides.
4107 . A regular expression is available for excluding directories from the
4110 20. The autoreply transport checks the characters in options that define the
4111 message's headers; it allows continued headers, but it was checking with
4112 isspace() after an embedded newline instead of explicitly looking for a
4115 21. If all the "regular" hosts to which an address was routed had passed their
4116 expiry times, and had not reached their retry times, the address was
4117 bounced, even if fallback hosts were defined. Now Exim should go on to try
4120 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4121 equivalent code in the SMTP transport. Some hosts send humungous responses
4122 to HELO/EHLO, more than 1024 it seems.
4124 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4125 gives compiler warnings in some environments. I've now done it "properly",
4128 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4129 (because of problems with the built-in one) was declared to return uschar *
4130 instead of char *, causing compiler failure.
4132 25. Fixed a file descriptor leak when processing alias/forward files.
4134 26. Fixed a minor format string issue in dbfn.c.
4136 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4138 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4139 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4142 29. When a sender address is verified, it is cached, to save repeating the test
4143 when there is more than one recipient in a message. However, when the
4144 verification involves a callout, it is possible for different callout
4145 options to be set for different recipients. It is too complicated to keep
4146 track of this in the cache, so now Exim always runs a verification when a
4147 callout is required, relying on the callout cache for the optimization.
4148 The overhead is duplication of the address routing, but this should not be
4151 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4152 to be verified with callout=postmaster, and the main callout worked but the
4153 postmaster check failed, the verification correctly failed. However, if a
4154 subsequent RCPT command asked for sender verification *without* the
4155 postmaster check, incorrect caching caused this verification also to fail,
4158 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4159 it was not caching the DNS options (qualify_single, search_parents) that
4160 were used when the lookup failed. A subsequent lookup with different
4161 options therefore always gave the same answer, though there were cases
4162 where it should not have. (Example: a "domains = !$mx_any" option on a
4163 dnslookup router: the "domains" option is always processed without any
4164 widening, but the router might have qualify_single set.) Now Exim uses the
4165 cached value only when the same options are set.
4167 32. Added John Jetmore's "exipick" utility to the distribution.
4169 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4170 than a timeout (e.g. a certificate is required, and is not provided), an
4171 Exim server now closes the connection immediately. Previously it waited for
4172 the client to close - but if the client is SSL, it seems that they each
4173 wait for each other, leading to a delay before one of them times out.
4175 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4176 maintained 0.8.x compatibility because I don't think many are using it, and
4177 it is clearly obsolete.
4179 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4182 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4183 client certificate was expired. A simple patch fixes this, though I don't
4184 understand the full logic of why the verify callback is called multiple
4187 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4188 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4189 which causes problems with some clients (such as the Certicom SSL Plus
4190 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4191 disables the coutermeasure allowing Eudora to connect."
4193 38. Exim was not checking that a write() to a log file succeeded. This could
4194 lead to Bad Things if a log got too big, in particular if it hit a file
4195 size limit. Exim now panics and dies if it cannot write to a log file, just
4196 as it does if it cannot open a log file.
4198 39. Modified OS/Makefile-Linux so that it now contains
4200 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4202 The two -D definitions ensure that Exim is compiled with large file
4203 support, which makes it possible to handle log files that are bigger than
4206 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4207 instance) a domain was checked against a named list that involved a lookup,
4208 causing $domain_data to be set, then another domain was checked against the
4209 same list, then the first domain was re-checked, the value of $domain_data
4210 after the final check could be wrong. In particular, if the second check
4211 failed, it could be set empty. This bug probably also applied to
4214 41. The strip_trailing_dot option was not being applied to the address given
4215 with the -f command-line option.
4217 42. The code for reading a message's header from the spool was incrementing
4218 $received_count, but never initializing it. This meant that the value was
4219 incorrect (doubled) while delivering a message in the same process in which
4220 it was received. In the most common configuration of Exim, this never
4221 happens - a fresh exec is done - but it can happen when
4222 deliver_drop_privilege is set.
4224 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4225 it now includes up to 150 characters of the unexpected data in the log
4228 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4229 and building data strings. The size of both of these buffers was 10 000
4230 bytes - far larger than anybody would *ever* want, thought I. Needless to
4231 say, somebody hit the limit. I have increased the maximum line length to
4232 20 000 and the maximum data length of concatenated lines to 100 000. I have
4233 also fixed two bugs, because there was no checking on these buffers. Tsk,
4234 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4235 buffer is too small.
4237 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4238 lsearch lookups. Now it does.
4240 46. When parsing a route_list item in a manualroute router, a fixed-length
4241 buffer was used for the list of hosts. I made this 1024 bytes long,
4242 thinking that nobody would ever have a list of hosts that long. Wrong.
4243 Somebody had a whole pile of complicated expansion conditions, and the
4244 string was silently truncated, leading to an expansion error. It turns out
4245 that it is easier to change to an unlimited length (owing to other changes
4246 that have happened since this code was originally written) than to build
4247 structure for giving a limitation error. The length of the item that
4248 expands into the list of hosts is now unlimited.
4250 47. The lsearch lookup could not handle data where the length of text line was
4251 more than 4095 characters. Such lines were truncated, leading to shortened
4252 data being returned. It should now handle lines of any length.
4254 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4255 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4258 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4259 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4260 no longer happens in comment lines. A list of replacements is now placed
4261 at the head of all of the source files, except those whose only change is
4262 to replace PERL_COMMAND in the very first #! line.
4264 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4265 messages on the queue, with a bottom-up merge sort, using code contributed
4266 by Michael Haardt. This should make operations like -bp somewhat faster on
4267 large queues. It won't affect queue runners, except when queue_run_in_order
4270 51. Installed eximstats 1.31 in the distribution.
4272 52. Added support for SRV lookups to the dnslookup router.
4274 53. If an ACL referred to $message_body or $message_body_end, the value was not
4275 reset for any messages that followed in the same SMTP session.
4277 54. The store-handling optimization for building very long strings was not
4278 differentiating between the different store pools. I don't think this
4279 actually made any difference in practice, but I've tidied it.
4281 55. While running the routers to verify a sender address, $sender_address
4282 was still set to the sender address. This is wrong, because when routing to
4283 send a bounce to the sender, it would be empty. Therefore, I have changed
4284 it so that, while verifying a sender address, $sender_address is set to <>.
4285 (There is no change to what happens when verifying a recipient address.)
4287 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4288 target A or AAAA records (if not already returned) without resetting the
4289 qualify_single or search_parents options of the DNS resolver. These are
4290 inappropriate in this case because the targets of MX and SRV records must
4291 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4292 target that, when qualified, matched something in the local domain. These
4293 two options are now turned off when doing these lookups.
4295 57. It seems that at least some releases of Reiserfs (which does not have the
4296 concept of a fixed number of inodes) returns zero and not -1 for the
4297 number of available inodes. This interacted badly with check_spool_inodes,
4298 which assumed that -1 was the "no such thing" setting. What I have done is
4299 to check that the total number of inodes is greater than zero before doing
4300 the test of how many are available.
4302 58. When a "warn" ACL statement has a log_message modifier, the message is
4303 remembered, and not repeated. This is to avoid a lot of repetition when a
4304 message has many recipients that cause the same warning to be written.
4305 Howewer, Exim was preserving the list of already written lines for an
4306 entire SMTP session, which doesn't seem right. The memory is now reset if a
4307 new message is started.
4309 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4310 debugging output unless log_rewrite was also set.
4312 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4313 of (int)(handle) when we know that handle contains (void *)(-1).
4315 61. The Exim daemon panic-logs an error return when it closes the incoming
4316 connection. However "connection reset by peer" seems to be common, and
4317 isn't really an error worthy of noting specially, so that particular error
4320 62. When Exim is trying to find all the local interfaces, it used to panic and
4321 die if the ioctl to get the interface flags failed. However, it seems that
4322 on at least one OS (Solaris 9) it is possible to have an interface that is
4323 included in the list of interfaces, but for which you get a failure error
4324 for this call. This happens when the interface is not "plumbed" into a
4325 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4326 failure of the "get flags" call assumes that the interface is down.
4328 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4329 makes life easier for people who are doing arithmetic on fields extracted
4330 from dates, where you often get leading zeros that should not be
4331 interpreted as octal.
4333 64. Added qualify_domain to the redirect router, to override the global
4336 65. If a pathologically long header line contained very many addresses (the
4337 report of this problem mentioned 10 000) and each of them was rewritten,
4338 Exim could use up a very large amount of memory. (It kept on making new
4339 copies of the header line as it rewrote, and never released the old ones.)
4340 At the expense of a bit more processing, the header rewriting function has
4341 been changed so that it no longer eats memory in this way.
4343 66. The generation of the Received: header has been moved from the time that a
4344 message starts to be received, to the time that it finishes. The timestamp
4345 in the Received: header should now be very close to that of the <= log
4346 line. There are two side-effects of this change:
4348 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4349 logged header lines no longer include the local Received: line, because
4350 it has not yet been created. The same applies to a copy of the message
4351 that is returned to a non-SMTP sender when a message is rejected.
4353 (b) When a filter file is tested using -bf, no additional Received: header
4354 is added to the test message. After some thought, I decided that this
4357 This change does not affect the value of $received_for. It is still set
4358 after address rewriting, but before local_scan() is called.
4360 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4362 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4363 gave an unhelpful panic error message, and a defer error. I have managed to
4364 change this behaviour so that it now rejects any supplied certificate,
4365 which seems right, as the list of acceptable certificates is empty.
4367 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4368 gave an unhelpful defer error. I have not managed to make this reject any
4369 supplied certificates, but the error message it gives is "no certificate
4370 supplied", which is not helpful.
4372 70. exigrep's output now also includes lines that are not associated with any
4373 message, but which match the given pattern. Implemented by a patch from
4374 Martin Sluka, which also tidied up the Perl a bit.
4376 71. Recipient callout verification, like sender verification, was using <> in
4377 the MAIL FROM command. This isn't really the right thing, since the actual
4378 sender may affect whether the remote host accepts the recipient or not. I
4379 have changed it to use the actual sender in the callout; this means that
4380 the cache record is now keyed on a recipient/sender pair, not just the
4381 recipient address. There doesn't seem to be a real danger of callout loops,
4382 since a callout by the remote host to check the sender would use <>.
4383 [SEE ABOVE: changed after hitting problems.]
4385 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4386 temporary errors. However, in the case of such a code being given after
4387 the end of a data transmission (i.e. after ".") Exim was failing to write
4388 a retry record for the message. (Yes, there was some broken host that was
4389 actually sending 8xx at this point.)
4391 73. An unknown lookup type in a host list could cause Exim to panic-die when
4392 the list was checked. (An example that provoked this was putting <; in the
4393 middle of a list instead of at the start.) If this happened during a DATA
4394 ACL check, a -D file could be left lying around. This kind of configuration
4395 error no longer causes Exim to die; instead it causes a defer errror. The
4396 incident is still logged to the main and panic logs.
4398 74. Buglet left over from Exim 3 conversion. The message "too many messages
4399 in one connection" was written to the rejectlog but not the mainlog, except
4400 when address rewriting (yes!) was being logged.
4402 75. Added write_rejectlog option.
4404 76. When a system filter was run not as root (that is, when system_filter_user
4405 was set), the values of the $n variables were not being returned to the
4406 main process; thus, they were not subsequently available in the $sn
4409 77. Added +return_path_on_delivery log selector.
4411 78. A connection timeout was being treated differently from recipients deferred
4412 when testing hosts_max_try with a message that was older than the host's
4413 retry timeout. (The host should not be counted, thus allowing all hosts to
4414 be tried at least once before bouncing.) This may have been the cause of an
4415 occasionally reported bug whereby a message would remain on the queue
4416 longer than the retry timeout, but would be bounced if a delivery was
4417 forced. I say "may" because I never totally pinned down the problem;
4418 setting up timeout/retry tests is difficult. See also the next item.
4420 79. The ultimate address timeout was not being applied to errors that involved
4421 a combination of host plus message (for example, a timeout on a MAIL
4422 command). When an address resolved to a number of possible hosts, and they
4423 were not all tried for each delivery (e.g. because of hosts_max_try), a
4424 message could remain on the queue longer than the retry timeout.
4426 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4429 81. Fixed an obscure SMTP outgoing bug which required at least the following
4430 conditions: (a) there was another message waiting for the same server;
4431 (b) the server returned 5xx to all RCPT commands in the first message so
4432 that the message was not completed; (c) the server dropped the connection
4433 or gave a negative response to the RSET that Exim sends to abort the
4434 transaction. The observed case was a dropped connection after DATA that had
4435 been sent in pipelining mode. That is, the server had advertised PIPELINING
4436 but was not implementing it correctly. The effect of the bug was incorrect
4437 behaviour, such as trying another host, and this could lead to a crash.
4443 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4444 and daemon.c were passed as pointers to ints; they should have been
4445 pointers to socklen_t variables (which are typically unsigned ints).
4447 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4450 3. Fixed a really odd bug that affected only the testing scheme; patching a
4451 certain fixed string in the binary changed the value of another string that
4452 happened to be identical to the end of the original first string.
4454 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4455 name", it returns that address as the IP address. On some operating
4456 systems (e.g. Solaris), it also passes back the IP address string as the
4457 "host name". However, on others (e.g. Linux), it passes back an empty
4458 string. Exim wasn't checking for this, and was changing the host name to an
4459 empty string, assuming it had been canonicized.
4461 5. Although rare, it is permitted to have more than one PTR record for a given
4462 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4463 all the names associated with an address, because they do in Solaris.
4464 However, it seems that they do not in Linux for data that comes from the
4465 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4466 I found this out when I moved to a new Linux workstation and tried to run
4467 the Exim test suite.
4469 To get round this problem I have changed the code so that it now does its
4470 own call to the DNS to look up PTR records when searching for a host name.
4471 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4472 addresses that are only in /etc/hosts are still found.
4474 This behaviour is, however, controlled by an option called host_lookup_
4475 order, which defaults to "bydns:byaddr". If people want to use the other
4476 order, or indeed, just use one or the other means of lookup, they can
4477 specify it in this variable.
4479 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4480 some operating systems, this comes back from gethostbyaddr() as an empty
4481 string, and this is what Exim used to test for. However, it seems that in
4482 other systems, "." is yielded. Exim now tests for this case too.
4484 7. The values of check_spool_space and check_log_space are now held internally
4485 as a number of kilobytes instead of an absolute number of bytes. If a
4486 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4487 kilobyte. This means that much larger values can be stored.
4489 8. Exim monitor: an attempt to get the action menu when not actually pointing
4490 at a message produces an empty menu entitled "No message selected". This
4491 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4492 no entries in it ("Shell widget menu has zero width and/or height"). So I
4493 have added a single, blank menu entry in this case.
4495 9. Added ${quote_local_part.
4497 10. MIME decoding is now applied to the contents of Subject: header lines when
4500 11. Now that a reference to $sender_host_address automatically causes a reverse
4501 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4502 host lookup before query-style lookups in lists that might use this
4503 variable. This has therefore been abolished, and the "net-" prefix is no
4504 longer necessary for query-style lookups.
4506 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4507 have been a typo for LFLAGS, so it has been changed.
4509 13. The install script calls Exim with "-C /dev/null" in order to find the
4510 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4511 to be output. Howeve, since Exim outputs its version number before the
4512 error, it didn't break the script. It just looked ugly. I fixed this by
4513 always allowing "-C /dev/null" if the caller is root.
4515 14. Ignore overlarge ACL variable number when reading spool file - insurance
4516 against a later release with more variables having written the file.
4518 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4519 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4520 this, as well as the form without the "IPv6" on the front.
4522 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4523 OS/Makefile-Darwin file.
4525 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4526 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4528 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4529 variable. However, some OS use ints for inodes. I've added cast to long int
4530 to get rid of the compiler warning.
4532 19. I had forgotten to lock out "/../" in configuration file names when
4533 ALT_CONFIG_PREFIX was set.
4535 20. Routers used for verification do not need to specify transports. However,
4536 if such a router generated a host list, and callout was configured, Exim
4537 crashed, because it could not find a port number from the (non-existent)
4538 transport. It now assumes port 25 in this circumstance.
4540 21. Added the -t option to exigrep.
4542 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4543 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4544 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4545 them is set, LOOKUP_LSEARCH is forced.
4547 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4549 24. Added sender and host information to the "rejected by local_scan()" log
4550 line; previously there was no indication of these.
4552 25. Added .include_if_exists.
4554 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4555 Linux. It turns out that not all file systems support this. Apparently some
4556 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4557 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4558 not supported on the file descriptor, is now ignored when Exim is trying to
4559 sync a directory. This applies only to Linux.
4561 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4563 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4564 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4567 29. Added support for maildirsize files from supplied patch (modified a bit).
4569 30. The use of :fail: followed by an empty string could lead Exim to respond to
4570 sender verification failures with (e.g.):
4572 550 Verification failed for <xxx>
4573 550 Sender verify failed
4575 where the first response line was missing the '-' that indicates it is not
4576 the final line of the response.
4578 31. The loop for finding the name of the user that called Exim had a hardwired
4579 limit of 10; it now uses the value of finduser_retries, which is used for
4580 all other user lookups.
4582 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4585 33. Exim was neglecting to zero errno before one call of strtol() when
4586 expanding a string and expecting an integer value. On some systems this
4587 resulted in spurious "integer overflow" errors. Also, it was casting the
4588 result into an int without checking.
4590 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4591 did not work. The code looks as if it has *never* worked, though it appears
4592 to have been documented since at least releast 1.62. I have made it work.
4594 35. The "timeout_DNS" error in retry rules, also documented since at least
4595 1.62, also never worked. As it isn't clear exactly what this means, and
4596 clearly it isn't a major issue, I have abolished the feature by treating it
4597 as "timeout", and writing a warning to the main and panic logs.
4599 36. The display of retry rules for -brt wasn't always showing the error code
4602 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4603 timeout_connect_A, timeout_connect_MX.
4605 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4606 to the empty sender.
4608 39. The daemon was not analysing the content of -oX till after it had closed
4609 stderr and disconnected from the controlling terminal. This meant that any
4610 syntax errors were only noted on the panic log, and the return code from
4611 the command was 0. By re-arranging the code a little, I've made the
4612 decoding happen first, so such errors now appear on stderr, and the return
4613 code is 1. However, the actual setting up of the sockets still happens in
4614 the disconnected process, so errors there are still only recorded on the
4617 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4618 connections (as happens on some IP stacks) was logged at start up time as
4619 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4620 it from "IPv6 and IPv4", which means that two separate sockets are being
4623 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4624 says whether AF_INET or AF_INET6 was passed as an argument.
4626 42. Exiwhat output was messed up when time zones were included in log
4629 43. Exiwhat now gives more information about the daemon's listening ports,
4630 and whether -tls-on-connect was used.
4632 44. The "port" option of the smtp transport is now expanded.
4634 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4635 silently ignored. Now an error message is written to the main and panic
4638 46. There's a new ACL modifier called "logwrite" which writes to a log file
4639 as soon as it is encountered.
4641 47. Added $local_user_uid and $local_user_gid at routing time.
4643 48. Exim crashed when trying to verify a sender address that was being
4646 49. Exim was recognizing only a space character after ".include". It now also
4647 recognizes a tab character.
4649 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4650 extracting the relevant information from the specification. The man page no
4651 longer contains scrambled data for the -d option, and I've added a section
4652 at the front about calling Exim under different names.
4654 51. Added "extra_headers" argument to the "mail" command in filter files.
4656 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4659 53. Installed eximstats 1.29.
4661 54. Added transport_filter_timeout as a generic transport option.
4663 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4664 Cc: header lines. This was required by RFC 822, but it not required by RFC
4667 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4668 incoming messages that did not have them. Now it does so only if the
4669 message originates locally, that is, if there is no associated remote host
4670 address. When Resent- header lines are present, this applies to the Resent-
4671 lines rather than the non-Resent- lines.
4673 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4674 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4676 58. Messages for configuration errors now include the name of the main
4677 configuration file - useful now that there may be more than one file in a
4678 list (.included file names were always shown).
4680 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4681 for those rare installations that do not start the daemon as root or run it
4682 setuid root. I've cut out the call to initgroups() if the daemon is not
4685 60. The Exim user and group can now be bound into the binary as text strings
4686 that are looked up at the start of Exim's processing.
4688 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4690 62. Added $mailstore_basename variable.
4692 63. Installed patch to sieve.c from Michael Haardt.
4694 64. When Exim failed to open the panic log after failing to open the main log,
4695 the original message it was trying to log was written to stderr and debug
4696 output, but if they were not available (the usual case in production), it
4697 was lost. Now it is written to syslog before the two lines that record the
4698 failures to open the logs.
4700 65. Users' Exim filters run in subprocesses under the user's uid. It is
4701 possible for a "deliver" command or an alias in a "personal" command to
4702 provoke an address rewrite. If logging of address rewriting is configured,
4703 this fails because the process is not running as root or exim. There may be
4704 a better way of dealing with this, but for the moment (because 4.30 needs
4705 to be released), I have disabled address rewrite logging when running a
4706 filter in a non-root, non-exim process.
4712 1. The buildconfig auxiliary program wasn't quoting the value set for
4713 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4714 not defined. This bug was masked in 4.22 by the effect that was fixed in
4717 2. Some messages that were rejected after a message id was allocated were
4718 shown as "incomplete" by exigrep. It no longer does this for messages that
4719 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4721 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4722 have allow_domain_literals set, the ID did not get logged in the <= line.
4723 Domain literals are now always recognized in Message-ID: header lines.
4725 4. The first argument for a ${extract expansion item is the key name or field
4726 number. Leading and trailing spaces in this item were not being ignored,
4727 causing some misleading effects.
4729 5. When deliver_drop_privilege was set, single queue runner processes started
4730 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4731 same command in the process it spins off) were not dropping privilege.
4733 6. When the daemon running as "exim" started a queue runner, it always
4734 re-executed Exim in the spun-off process. This is a waste of effort when
4735 deliver_drop_privilege is set. The new process now just calls the
4736 queue-runner function directly.
4742 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4745 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4746 ignored. Though the use of -oP was forcing the writing of a pid file, it
4747 was always written to the default place.
4749 3. If the message "no IP address found for host xxxx" is generated during
4750 incoming verification, it is now followed by identification of the incoming
4751 connection (so you can more easily find what provoked it).
4753 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4755 5. Added some features to "harden" Exim a bit more against certain attacks:
4757 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4758 be put in Local/Makefile. This is like the never_users runtime option,
4759 but it cannot be overridden. The default setting is "root".
4761 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4762 prefix string with which any file named in a -C command line option
4765 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4766 is retained for -C and -D only if the caller of Exim is root. Without
4767 it, the exim user may also use -C and -D and retain privilege.
4769 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4770 command line option is disabled.
4772 6. Macro names set by the -D option must start with an upper case letter, just
4773 like macro names defined in the configuration file.
4775 7. Added "dereference=" facility to LDAP.
4777 8. Two instances of the typo "uknown" in the source files are fixed.
4779 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4780 the Configure-Makefile script screwed up while processing it.
4782 10. Incorporated PCRE 4.4.
4784 11. The SMTP synchronization check was not operating right at the start of an
4785 SMTP session. For example, it could not catch a HELO sent before the client
4786 waited for the greeting. There is now a check for outstanding input at the
4787 point when the greeting is written. Because of the duplex, asynchronous
4788 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4789 way, but not yet received, when the check is performed.
4791 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4792 on TCP/IP sockets, because this apparently causes some broken clients to
4795 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4796 unchanged) from the Cygwin maintainer.
4798 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4799 was supported instead of testing for mbx. Effectively a typo.
4801 15. The spa authenticator server code was not checking that the input it
4802 received was valid base64.
4804 16. The debug output line for the "set" modifier in ACLs was not showing the
4805 name of the variable that was being set.
4807 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4809 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4810 lookup on its own. Something else was needed to trigger the lookup. For
4811 example, a match in host_lookup or the need for a host name in a host list.
4812 Now, if $sender_host_name is referenced and the host name has not yet been
4813 looked up, a lookup is performed. If the lookup fails, the variable remains
4814 empty, and $host_lookup_failed is set to "1".
4816 19. Added "eqi" as a case-independent comparison operator.
4818 20. The saslauthd authentication condition could segfault if neither service
4819 nor realm was specified.
4821 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4822 error message that was logged was misleading, and incoming SMTP
4823 connections were dropped. The message is now more accurate, and temporary
4824 errors are given to SMTP connections.
4826 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4827 (except RSET) with a 421 error, until QUIT is received. However, it was
4828 failing to send a response to QUIT.
4830 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4831 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4832 ACL is now run after the helo_try_verify_hosts code.
4834 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4835 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4836 case-independent, so other case variants are also recognized). Apparently
4837 some systems use these upper case variants.
4839 25. If more than two messages were waiting for the same host, and a transport
4840 filter was specified for the transport, Exim sent two messages over the
4841 same TCP/IP connection, and then failed with "socket operation on non-
4842 socket" when it tried to send the third.
4844 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4846 27. The extern definition of crypt16() in expand.c was not being excluded when
4847 the OS had its own crypt16() function.
4849 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4850 as a preferred synonym for return_size_limit, both as an option and as an
4853 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4855 30. Changed the default configuration ACL to relax the local part checking rule
4856 for addresses that are not in any local domains. For these addresses,
4857 slashes and pipe symbols are allowed within local parts, but the sequence
4858 /../ is explicitly forbidden.
4860 31. SPA server authentication was not clearing the challenge buffer before
4863 32. log_message in a "warn" ACL statement was writing to the reject log as
4864 well as to the main log, which contradicts the documentation and doesn't
4865 seem right (because no rejection is happening). So I have stopped it.
4867 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4868 However, I am unable to do any testing of this.
4870 34. Fixed an infelicity in the appendfile transport. When checking directories
4871 for a mailbox, to see if any needed to be created, it was accidentally
4872 using path names with one or more superfluous leading slashes; tracing
4873 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4875 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4876 contents are added to the log line that is written for every discarded
4877 recipient. (Previously a log_message setting was ignored.)
4879 36. The ${quote: operator now quotes the string if it is empty.
4881 37. The install script runs exim in order to find its version number. If for
4882 some reason other than non-existence or emptiness, which it checks, it
4883 could not run './exim', it was installing it with an empty version number,
4884 i.e. as "exim-". This error state is now caught, and the installation is
4887 38. An argument was missing from the function that creates an error message
4888 when Exim fails to connect to the socket for saslauthd authentication.
4889 This could cause Exim to crash, or give a corrupted message.
4891 39. Added isip, isip4, and isip6 to ${if conditions.
4893 40. The ACL variables $acl_xx are now saved with the message, and can be
4894 accessed later in routers, transports, and filters.
4896 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4897 strings in the file are not string-expanded.
4899 42. If a MAIL command specified a SIZE value that was too large to fit into an
4900 int variable, the check against message_size_limit failed. Such values are
4901 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4902 day this will have to be increased, but I don't think I want to be around
4903 when emails are that large.
4910 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4911 iconv() is not standard in FreeBSD.
4913 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4914 IPv6 enabled. The observed symptom was a segmentation fault on return from
4915 the function os_common_find_running_interfaces() in src/os.c.
4917 3. In the check_special_case() function in daemon.c I had used "errno" as an
4918 argument name, which causes warnings on some systems. This was basically a
4919 typo, since it was named "eno" in the comments!
4921 4. The code that waits for the clock to tick (at a resolution of some fraction
4922 of a second) so as to ensure message-id uniqueness was always waiting for
4923 at least one whole tick, when it could have waited for less. [This is
4924 almost certainly not relevant at current processor speeds, where it is
4925 unlikely to ever wait at all. But we try to future-proof.]
4927 5. The function that sleeps for a time interval that includes fractions of a
4928 second contained a race. It did not block SIGALRM between setting the
4929 timer, and suspending (a couple of lines later). If the interval was short
4930 and the sigsuspend() was delayed until after it had expired, the suspension
4931 never ended. On busy systems this could lead to processes getting stuck for
4934 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4935 process, before it forks any delivery processes. The open lookup caching
4936 mechanism meant that the open file or database connection was passed into
4937 the delivery process. The problem was that delivery processes always tidy
4938 up cached lookup data. This could cause a problem for the next delivery
4939 process started by the queue runner, because the external queue runner
4940 process does not know about the closure. So the next delivery process
4941 still has data in the lookup cache. In the case of a file lookup, there was
4942 no problem because closing a file descriptor in a subprocess doesn't affect
4943 the parent. However, if the lookup was caching a connection to a database,
4944 the connection was closed, and the second delivery process was likely to
4945 see errors such as "PGSQL: query failed: server closed the connection
4946 unexpectedly". The problem has been fixed by closing all cached lookups
4947 in a queue runner before running a delivery process.
4949 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4950 seem to have the "const" qualifier which it has on other OS. I've
4953 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4954 *of the same type* (client or server) with the same public name that an
4955 error should be diagnosed.
4957 9. When Exim looked up a host name for an IP address, but failed to find the
4958 original IP address when looking up the host name (a safety check), it
4959 output the message "<ip address> does not match any IP for NULL", which was
4960 confusing, to say the least. The bug was that the host name should have
4961 appeared instead of "NULL".
4963 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4964 user that is built into the binary, and the -C or -D options is used, root
4965 privilege is dropped before the configuration file is read. In addition,
4966 logging is switched to stderr instead of the normal log files. If the
4967 configuration then re-defines the Exim user, the unprivileged environment
4968 is probably not what is expected, so Exim logs a panic warning message (but
4971 However, if deliver_drop_privilege is set, the unprivileged state may well
4972 be exactly what is intended, so the warning has been cut out in that case,
4973 and Exim is allowed to try to write to its normal log files.
4979 1. smtp_return_error_details was not giving details for temporary sender
4980 or receiver verification errors.
4982 2. Diagnose a configuration error if two authenticators have the same public
4985 3. Exim used not to create the message log file for a message until the first
4986 delivery attempt. This could be confusing when incoming messages were held
4987 for policy or load reasons. The message log file is now created at the time
4988 the message is received, and an initial "Received" line is written to it.
4990 4. The automatically generated man page for command line options had a minor
4991 bug that caused no ill effects; however, a more serious problem was that
4992 the procedure for building the man page automatically didn't always
4993 operate. Consequently, release 4.20 contains an out-of-date version. This
4994 shouldn't happen again.
4996 5. When building Exim with embedded Perl support, the script that builds the
4997 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4998 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5000 6. The freeze_tell option was not being used for messages that were frozen on
5001 arrival, either by an ACL or by local_scan().
5003 7. Added the smtp_incomplete_transaction log selector.
5005 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5006 was accepting AUTH without a new EHLO.
5008 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5009 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5012 10. Logging of TCP/IP connections (when configured) now happens in the main
5013 daemon process instead of the child process, so that the TCP/IP connection
5014 count is more accurate (but it can never be perfect).
5016 11. The use of "drop" in a nested ACL was not being handled correctly in the
5017 outer ACL. Now, if condition failure induced by the nested "drop" causes
5018 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5019 or "require"), the connection is dropped.
5021 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5022 that yield "discard" can now be used with an "accept" or a "discard" verb,
5023 but an error is generated for any others (because I can't see a useful way
5024 to define what should happen).
5026 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5027 are now processed in the same way as lines in the Exim configuration file.
5028 In particular, continuation lines are supported.
5030 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5032 15. Added -ti meaning -t -i.
5034 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5035 domains, and warn by logging if others are found.
5037 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5038 returned from GIFCONF when Exim is trying to find the list of interfaces on
5039 a host. The code in os.c has been modified to copy each ifreq to an aligned
5040 structure in all cases.
5042 Also, in some cases, the returned ifreq's were being copied to a 'struct
5043 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5044 means the last couple of bytes of an IPv6 address could be chopped if the
5045 ifreq contained only a normal sockaddr (14 bytes storage).
5047 18. Named domain lists were not supported in the hosts_treat_as_local option.
5048 An entry such as +xxxx was not recognized, and was treated as a literal
5051 19. Ensure that header lines added by a DATA ACL are included in the reject log
5052 if the ACL subsequently rejects the message.
5054 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5055 MD5 (which is deprecated).
5057 21. When testing a filter file using -bf, Exim was writing a message when it
5058 took the sender from a "From " line in the message, but it was not doing so
5059 when it took $return_path from a Return-Path: header line. It now does.
5061 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5062 with a valid header line field name (a series of printing characters
5063 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5065 23. Changed "disc" in the source to "disk" to conform to the documentation and
5066 the book and for uniformity.
5068 24. Ignore Sendmail's -Ooption=value command line item.
5070 25. When execve() failed while trying to run a command in a pipe transport,
5071 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5072 could be confused with a return value of 69 from the command itself. This
5073 has been changed to 127, the value the shell returns if it is asked to run
5074 a non-existent command. The wording for the related log line suggests a
5075 non-existent command as the problem.
5077 26. If received_header_text expands to an empty string, do not add a Received:
5078 header line to the message. (Well, it adds a token one on the spool, but
5079 marks it "old" so that it doesn't get used or transmitted.)
5081 27. Installed eximstats 1.28 (addition of -nt option).
5083 28. There was no check for failure on the call to getsockname() in the daemon
5084 code. This can fail if there is a shortage of resources on the system, with
5085 ENOMEM, for example. A temporary error is now given on failure.
5087 29. Contrary to the C standard, it seems that in some environments, the
5088 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5089 program. Exim now does this explicitly; it affects the formatting of
5090 timestamps using strftime().
5092 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5093 complaints. I've now initialized all the variables, to avoid this.
5095 32. Header lines added by a system filter were not being "seen" during
5096 transport-time rewrites.
5098 33. The info_callback() function passed to OpenSSL is set up with type void
5099 (*)(SSL *, int, int), as described somewhere. However, when calling the
5100 function (actually a macro) that sets it up, the type void(*)() is
5101 expected. I've put in a cast to prevent warnings from picky compilers.
5103 34. If a DNS black list lookup found a CNAME record, but there were no A
5104 records associated with the domain it pointed at, Exim crashed.
5106 35. If a DNS black list lookup returned more than one A record, Exim ignored
5107 all but the first. It now scans all returned addresses if a particular IP
5108 value is being sought. In this situation, the contents of the
5109 $dnslist_value variable are a list of all the addresses, separated by a
5112 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5113 to accept a host name and all its aliases if the forward lookup for any of
5114 them yielded the IP address of the incoming connection. Now it accepts only
5115 those names whose forward lookup yields the correct IP address. Any other
5116 names are discarded. This closes a loophole whereby a rogue DNS
5117 administrator could create reverse DNS records to break through a
5118 wildcarded host restriction in an ACL.
5120 37. If a user filter or a system filter that ran in a subprocess used any of
5121 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5122 the wrong values were passed to the pipe command ($thisaddress had the
5123 value of $0, $0 had the value of $1, etc). This bug was introduced by
5124 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5126 38. Improved the line breaking for long SMTP error messages from ACLs.
5127 Previously, if there was no break point between 40 and 75 characters, Exim
5128 left the rest of the message alone. Two changes have been made: (a) I've
5129 reduced the minimum length to 35 characters; (b) if it can't find a break
5130 point between 35 and 75 characters, it looks ahead and uses the first one
5131 that it finds. This may give the occasional overlong line, but at least the
5132 remaining text gets split now.
5134 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5135 file descriptors might be low, and that setting 1000 would always raise it.
5136 It turns out that in some environments, the limit is already over 1000 and
5137 that lowering it causes trouble. So now Exim takes care not to decrease it.
5139 40. When delivering a message, the value of $return_path is set to $sender_
5140 address at the start of routing (routers may change the value). By an
5141 oversight, this default was not being set up when an address was tested by
5142 -bt or -bv, which affected the outcome if any router or filter referred to
5145 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5146 log only when "message" or "log_message" are set. However, if one of the
5147 conditions was an address verification, or a call to a nested ACL, the
5148 messages generated by the underlying test were being passed through. This
5149 no longer happens. The underlying message is available in $acl_verify_
5150 message for both "message" and "log_message" expansions, so it can be
5151 passed through if needed.
5153 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5154 new expansion $bh_ to give the encoded byte string without charset
5155 translation. Translation happens only if iconv() is available; HAVE_ICONV
5156 indicates this at build time. HEADERS_CHARSET gives the charset to
5157 translate to; headers_charset can change it in the configuration, and
5158 "headers charset" can change it in an individual filter file.
5160 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5161 that creates RFC 2047 encoded "words" labels them as that charset instead
5162 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5163 expansion operator; (ii) when Exim creates a From: line for a local
5164 message; (iii) when a header line is rewritten to include a "phrase" part.
5166 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5167 buggy, causing it to skip the first lines of messages whose message ID
5168 ended in 'D'. This would not have bitten before Exim release 4.14, because
5169 message IDs were unlikely to end in 'D' before then. The effect was to have
5170 incorrect size information for certain domains.
5172 45. #include "config.h" was missing at the start of the crypt16.c module. This
5173 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5176 46. If there was a timeout during a "random" callout check, Exim treated it as
5177 a failure of the random address, and carried on sending RSET and the real
5178 address. If the delay was just some slowness somewhere, the response to the
5179 original RCPT would be taken as a response to RSET and so on, causing
5180 mayhem of various kinds.
5182 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5183 when I implemented it. It didn't allow for the fact that some option values
5184 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5185 the right test for positive values.
5187 48. Domain names in DNS records are case-independent. Exim always looks them up
5188 in lower case. Some resolvers return domain names in exactly the case they
5189 appear in the zone file, that is, they may contain uppercase letters. Not
5190 all resolvers do this - some return always lower case. Exim was treating a
5191 change of case by a resolver as a change of domain, similar to a widening
5192 of a domain abbreviation. This triggered its re-routing code and so it was
5193 trying to route what was effectively the same domain again. This normally
5194 caused routing to fail (because the router wouldn't handle the domain
5195 twice). Now Exim checks for this case specially, and just changes the
5196 casing of the domain that it ultimately uses when it transmits the message
5199 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5202 50. If a filter generated a file delivery with a non-absolute name (possible if
5203 no home directory exists for the router), the forbid_file option was not
5206 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5207 the existing equality matching.
5209 52. Exim was using ints instead of ino_t variables in some places where it was
5210 dealing with inode numbers.
5212 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5213 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5214 called TMPDIR, and if it finds it is different, it changes its value.
5216 54. The smtp_printf() function is now made available to local_scan() so
5217 additional output lines can be written before returning. There is also an
5218 smtp_fflush() function to enable the detection of a dropped connection.
5219 The variables smtp_input and smtp_batched_input are exported to
5222 55. Changed the default runtime configuration: the message "Unknown user"
5223 has been removed from the ACL, and instead placed on the localuser router,
5224 using the cannot_route_message feature. This means that any verification
5225 failures that generate their own messages won't get overridden. Similarly,
5226 the "Unrouteable address" message that was in the ACL for unverifiable
5227 relay addresses has also been removed.
5229 56. Added hosts_avoid_esmtp to the smtp transport.
5231 57. The exicyclog script was not checking for the esoteric option
5232 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5233 will work only if exicyclog is run under the appropriate euid.
5235 58. Following a discussion on the list, the rules by which Exim recognises line
5236 endings on incoming messages have been changed. The -dropcr and drop_cr
5237 options are now no-ops, retained only for backwards compatibility. The
5238 following line terminators are recognized: LF CRLF CR. However, special
5239 processing applies to CR:
5241 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5242 nor a local message in the state where . is a terminator.
5244 (ii) If a bare CR is encountered in a header line, an extra space is added
5245 after the line terminator so as not to end the header. The reasoning
5246 behind this is that bare CRs in header lines are most likely either
5247 to be mistakes, or people trying to play silly games.
5249 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5250 was being incorrectly given as 18 bytes larger than it should have been.
5251 This is a VOB (very old bug).
5253 60. This may never have affected anything current, but just in case it has:
5254 When the local host is found other than at the start of a list of hosts,
5255 the local host, those with the same MX, and any that follow, are discarded.
5256 When the list in question was part of a longer list of hosts, the following
5257 hosts (not currently being processed) were also being discarded. This no
5258 longer happens. I'm not sure if this situation could ever has previously
5261 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5264 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5265 header. This is something that is recommended in a new Internet Draft, and
5266 is something that is documented as being done by Sendmail. There are two
5267 possible values. For messages generated by the autoreply transport, Exim
5270 Auto-Submitted: auto-replied
5272 whereas for all other generated messages (e.g. bounces) it adds
5274 Auto-Submitted: auto-generated
5276 63. The "personal" condition in filters now includes a test for the
5277 Auto-Submitted: header. If it contains the string "auto-" the message it
5278 not considered personal.
5280 64. Added rcpt_include_affixes as a generic transport option.
5282 65. Added queue_only_override (default true).
5284 66. Added the syslog_duplication option.
5286 67. If what should have been the first header line of a message consisted of
5287 a space followed by a colon, Exim was mis-interpreting it as a header line.
5288 It isn't of course - it is syntactically invalid and should therefore be
5289 treated as the start of the message body. The misbehaviour could have
5290 caused a number of strange effects, including loss of data in subsequent
5291 header lines, and spool format errors.
5293 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5294 client host had authenticated. This control can now be exercised by an ACL
5295 for more flexibility.
5297 69. By default, callouts do not happen when testing with -bh. There is now a
5298 variant, -bhc, which does actually run the callout code, including
5299 consulting and updating the callout cache.
5301 70. Added support for saslauthd authentication, courtesy of Alexander
5304 71. If statvfs() failed on the spool or log directories while checking their
5305 size for availability, Exim confusingly gave the error "space shortage".
5306 Furthermore, in debugging mode it crashed with a floating point exception.
5307 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5308 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5309 serious problem, Exim now writes to the main and panic logs when this
5310 happens, with details of the failure. It then refuses to accept the
5311 incoming message, giving the message "spool directory problem" or "log
5312 directory problem" with a 421 code for SMTP messages.
5314 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5315 0, 1, and 2 exist, because some OS complain for execs without them (see
5316 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5317 descriptors. However, the code omitted to check that the open succeeded,
5318 causing mysterious errors if for some reason the permissions on /dev/null
5319 got screwed. Now Exim writes a message to the main and panic logs, and
5320 bombs out if it can't open /dev/null.
5322 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5323 interact so that it is all more flexible. It is supposed to remain
5324 backwards compatible. Also added extra_local_interfaces.
5326 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5327 to bomb out with an assertion failure - to the client this appears as a
5328 connection drop. This problem occurs in the part of the code that was taken
5329 from the Samba project. Fortunately, the assertion is in a very simple
5330 function, so I have fixed this by reproducing the function inline in the
5331 one place where it is called, and arranging for authentication to fail
5332 instead of killing the process with assert().
5334 75. The SPA client code was not working when the server requested OEM rather
5335 than Unicode encoding.
5337 76. Added code to make require_files with a specific uid setting more usable in
5338 the case where statting the file as root fails - usually a non-root-mounted
5339 NFS file system. When this happens and the failure is EACCES, Exim now
5340 forks a subprocess and does the per-uid checking as the relevant uid.
5342 77. Added process_log_path.
5344 78. If log_file_path was not explicitly set, a setting of check_log_space or
5345 check_log_inodes was ignored.
5347 79. If a space check for the spool or log partitions fails, the incident is now
5348 logged. Of course, in the latter case the data may get lost...
5350 80. Added the %p formatting code to string_format() so that it can be used to
5351 print addresses in debug_print(). Adjusted all the address printing in the
5352 debugging in store.c to use %p rather than %d.
5354 81. There was a concern that a line of code in smtp_in.c could overflow a
5355 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5356 initially expressed, the concern was not well-founded, because trailing
5357 spaces are removed early. However, if the trailing spaces were followed by
5358 a NULL, they did not get removed, so the overflow was possible. Two fixes
5361 (a) I re-wrote the offending code in a cleaner fashion.
5362 (b) If an incoming SMTP command contains a NULL character, it is rejected
5365 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5366 runs initgroups(), so that if the Exim user is in any additional groups,
5367 they will be used during message reception.
5373 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob