1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 NM/01 Workround for PCRE version dependency in version reporting
12 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
13 This fixes portability to compilers other than gcc, notably
14 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
16 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
17 makefiles for portability to HP-UX and POSIX correctness.
19 PP/01 Permit LOOKUP_foo enabling on the make command-line.
20 Also via indented variable definition in the Makefile.
21 (Debugging by Oliver Heesakkers).
23 PP/02 Restore caching of spamd results with expanded spamd_address.
24 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
26 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
27 Improves build reliability. Fix from: Frank Elsner
29 NM/02 Fix wide character breakage in the rfc2047 coding
30 Fixes bug 1064. Patch from Andrey N. Oktyabrski
32 NM/03 Allow underscore in dnslist lookups
33 Fixes bug 1026. Patch from Graeme Fowler
35 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
36 Code patches from Adam Ciarcinski of NetBSD.
38 NM/04 Fixed exiqgrep to cope with mailq missing size issue
41 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
42 is logged, to avoid truncation. Patch from John Horne.
44 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
45 Patch from Jakob Hirsch.
47 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
48 SQL string expansion failure details.
49 Patch from Andrey Oktyabrski.
51 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
52 Patch from Simon Arlott.
54 PP/09 New lookups functionality failed to compile on old gcc which rejects
55 extern declarations in function scope.
56 Patch from Oliver Fleischmann
58 PP/10 Use sig_atomic_t for flags set from signal handlers.
59 Check getgroups() return and improve debugging.
60 Fixed developed for diagnosis in bug 927 (which turned out to be
63 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
64 Patch from Mark Zealey.
66 PP/12 Bugzilla 1056: Improved spamd server selection.
67 Patch from Mark Zealey.
69 PP/13 Bugzilla 1086: Deal with maildir quota file races.
70 Based on patch from Heiko Schlittermann.
72 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
73 Patch from Uwe Doering, sign-off by Michael Haardt.
75 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
76 variable declaration deep within a block. Bug and patch from
79 PP/15 lookups-Makefile IRIX compatibilty coercion.
81 PP/16 Make DISABLE_DKIM build knob functional.
83 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
86 TF/03 Fix valgrind.h portability to C89 compilers that do not support
87 variable argument macros. Our copy now differs from upstream.
93 TF/01 Failure to get a lock on a hints database can have serious
94 consequences so log it to the panic log.
96 TF/02 Log LMTP confirmation messages in the same way as SMTP,
97 controlled using the smtp_confirmation log selector.
99 TF/03 Include the error message when we fail to unlink a spool file.
101 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
102 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
103 for maintaining out-of-tree patches for some time.
105 PP/01 Bugzilla 139: Documentation and portability issues.
106 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
107 Handle per-OS dynamic-module compilation flags.
109 PP/02 Let /dev/null have normal permissions.
110 The 4.73 fixes were a little too stringent and complained about the
111 permissions on /dev/null. Exempt it from some checks.
112 Reported by Andreas M. Kirchwitz.
114 PP/03 Report version information for many libraries, including
115 Exim version information for dynamically loaded libraries. Created
116 version.h, now support a version extension string for distributors
117 who patch heavily. Dynamic module ABI change.
119 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
120 privilege escalation vulnerability whereby the Exim run-time user
121 can cause root to append content of the attacker's choosing to
124 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
127 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
128 If dropping privileges for untrusted macros, we disabled normal logging
129 on the basis that it would fail; for the Exim run-time user, this is not
130 the case, and it resulted in successful deliveries going unlogged.
131 Fixed. Reported by Andreas Metzler.
137 PP/01 Date: & Message-Id: revert to normally being appended to a message,
138 only prepend for the Resent-* case. Fixes regression introduced in
139 Exim 4.70 by NM/22 for Bugzilla 607.
141 PP/02 Include check_rfc2047_length in configure.default because we're seeing
142 increasing numbers of administrators be bitten by this.
144 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
146 PP/03 Bugzilla 994: added openssl_options main configuration option.
148 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
150 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
152 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
154 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
155 without a peer certificate, leading to a segfault because of an
156 assumption that peers always have certificates. Be a little more
157 paranoid. Problem reported by Martin Tscholak.
159 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
160 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
161 NB: ClamAV planning to remove STREAM in "middle of 2010".
162 CL also introduces -bmalware, various -d+acl logging additions and
163 more caution in buffer sizes.
165 PP/09 Implemented reverse_ip expansion operator.
167 PP/10 Bugzilla 937: provide a "debug" ACL control.
169 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
171 PP/12 Bugzilla 973: Implement --version.
173 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
175 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
177 PP/15 Bugzilla 816: support multiple condition rules on Routers.
179 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
180 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
181 ignore trailing whitespace.
183 JJ/02 prevent non-panic DKIM error from being sent to paniclog
185 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
188 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
189 Notification from Dr Andrew Aitchison.
191 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
192 ExtendedDetectionInfo response format.
193 Notification from John Horne.
195 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
198 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
199 XSL and documented dependency on system catalogs, with examples of how
202 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
205 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
206 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
207 configuration file which is writeable by the Exim user or group.
209 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
210 of configuration files to cover files specified with the -C option if
211 they are going to be used with root privileges, not just the default
214 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
215 option (effectively making it always true).
217 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
218 files to be used while preserving root privileges.
220 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
221 that rogue child processes cannot use them.
223 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
224 run-time user, instead of root.
226 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
227 Exim run-time user without dropping privileges.
229 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
230 result string, instead of calling string_vformat() twice with the same
233 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
234 for other users. Others should always drop root privileges if they use
235 -C on the command line, even for a whitelisted configure file.
237 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
239 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
245 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
246 $data_path, and $header_path variables; fixed documentation bugs and
249 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
250 exipick to access non-standard spools, including the "frozen" queue
253 NM/01 Bugzilla 965: Support mysql stored procedures.
254 Patch from Alain Williams
256 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
258 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
259 Patch from Andreas Metzler
261 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
262 Patch from Kirill Miazine
264 NM/05 Bugzilla 671: Added umask to procmail example.
266 JJ/03 installed exipick 20100323.0, fixing doc bug
268 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
269 directory. Notification and patch from Dan Rosenberg.
271 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
273 TK/02 Improve log output when DKIM signing operation fails.
275 MH/01 Treat the transport option dkim_domain as a colon separated
276 list, not as a single string, and sign the message with each element,
277 omitting multiple occurences of the same signer.
279 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
280 Bugzilla 985, 986. Patch by Simon Arlott
282 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
283 Patch by Simon Arlott
285 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
286 MBX locking. Notification from Dan Rosenberg.
292 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
294 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
296 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
298 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
300 NM/04 Bugzilla 915: Flag broken perl installation during build.
306 TK/01 Added patch by Johannes Berg that expands the main option
307 "spamd_address" if it starts with a dollar sign.
309 TK/02 Write list of recipients to X-Envelope-Sender header when building
310 the mbox-format spool file for content scanning (suggested by Jakob
313 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
314 (http://www.dcc-servers.net/) support via dccifd. Activated by
315 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
317 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
318 by Mark Daniel Reidel <mr@df.eu>.
320 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
321 When building exim an external PCRE library is now needed -
322 PCRE is a system library on the majority of modern systems.
323 See entry on PCRE_LIBS in EDITME file.
325 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
326 conversation. Added nologin parameter to request.
327 Patch contributed by Kirill Miazine.
329 TF/01 Do not log submission mode rewrites if they do not change the address.
331 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
333 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
334 log files in place. Contributed by Roberto Lima.
336 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
338 TF/03 Bugzilla 615: When checking the local_parts router precondition
339 after a local_part_suffix or local_part_prefix option, Exim now
340 does not use the address's named list lookup cache, since this
341 contains cached lookups for the whole local part.
343 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
344 Robert Millan. Documentation is in experimental-spec.txt.
346 TF/04 Bugzilla 668: Fix parallel build (make -j).
348 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
350 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
351 Patch provided by Jan Srzednicki.
353 TF/05 Leading white space used to be stripped from $spam_report which
354 wrecked the formatting. Now it is preserved.
356 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
357 that they are available at delivery time.
359 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
361 TF/08 TLS error reporting now respects the incoming_interface and
362 incoming_port log selectors.
364 TF/09 Produce a more useful error message if an SMTP transport's hosts
365 setting expands to an empty string.
367 NM/06 Bugzilla 744: EXPN did not work under TLS.
368 Patch provided by Phil Pennock.
370 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
371 Patch provided by Richard Godbee.
373 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
374 acl_smtp_notquit, added index entry.
376 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
377 Patch provided by Eugene Bujak.
379 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
380 accept(). Patch provided by Maxim Dounin.
382 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
383 Patch provided by Phil Pennock.
385 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
387 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
388 Patch provided by Brad "anomie" Jorsch.
390 NM/14 Bugzilla 622: Added timeout setting to transport filter.
391 Patch provided by Dean Brooks.
393 TK/05 Add native DKIM support (does not depend on external libraries).
395 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
396 Patch provided by Graeme Fowler.
398 NM/16 Bugzilla 851: Documentation example syntax fix.
400 NM/17 Changed NOTICE file to remove references to embedded PCRE.
402 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
405 NM/19 Bugzilla 745: TLS version reporting.
406 Patch provided by Phil Pennock.
408 NM/20 Bugzilla 167: bool: condition support.
409 Patch provided by Phil Pennock.
411 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
412 clients. Patch provided by Phil Pennock.
414 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
415 Patch provided by Brad "anomie" Jorsch.
417 NM/23 Bugzilla 687: Fix misparses in eximstats.
418 Patch provided by Heiko Schlittermann.
420 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
421 Patch provided by Heiko Schlittermann.
423 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
424 plus update to original patch.
426 NM/26 Bugzilla 799: Documentation correction for ratelimit.
428 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
429 Patch provided by David Brownlee.
431 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
433 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
435 NM/30 Bugzilla 888: TLS documentation bugfixes.
437 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
439 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
440 Unlike the original bugzilla I have changed all shell scripts in src tree.
442 NM/33 Bugzilla 898: Transport filter timeout fix.
443 Patch by Todd Rinaldo.
445 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
446 Patch by Serge Demonchaux.
448 NM/35 Bugzilla 39: Base64 decode bug fixes.
449 Patch by Jakob Hirsch.
451 NM/36 Bugzilla 909: Correct connect() call in dcc code.
453 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
455 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
457 NM/39 Bugzilla 911: Fixed MakeLinks build script.
463 TK/01 Add preliminary DKIM support. Currently requires a forked version of
464 ALT-N's libdkim that I have put here:
465 http://duncanthrax.net/exim-experimental/
467 Note to Michael Haardt: I had to rename some vars in sieve.c. They
468 were called 'true' and it seems that C99 defines that as a reserved
469 keyword to be used with 'bool' variable types. That means you could
470 not include C99-style headers which use bools without triggering
471 build errors in sieve.c.
473 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
474 as mailq or other aliases. Changed the --help handling significantly
475 to do whats expected. exim_usage() emits usage/help information.
477 SC/01 Added the -bylocaldomain option to eximstats.
479 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
481 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
483 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
489 PH/01 Another patch from the Sieve maintainer.
491 PH/02 When an IPv6 address is converted to a string for single-key lookup
492 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
493 dots are used instead of colons so that keys in lsearch files need not
494 contain colons. This was done some time before quoting was made available
495 in lsearch files. However, iplsearch files do require colons in IPv6 keys
496 (notated using the quote facility) so as to distinguish them from IPv4
497 keys. This meant that lookups for IP addresses in host lists did not work
498 for iplsearch lookups.
500 This has been fixed by arranging for IPv6 addresses to be expressed with
501 colons if the lookup type is iplsearch. This is not incompatible, because
502 previously such lookups could never work.
504 The situation is now rather anomolous, since one *can* have colons in
505 ordinary lsearch keys. However, making the change in all cases is
506 incompatible and would probably break a number of configurations.
508 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
511 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
512 conversion specification without a maximum field width, thereby enabling
513 a rogue spamd server to cause a buffer overflow. While nobody in their
514 right mind would setup Exim to query an untrusted spamd server, an
515 attacker that gains access to a server running spamd could potentially
516 exploit this vulnerability to run arbitrary code as the Exim user.
518 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
519 $primary_hostname instead of what libspf2 thinks the hosts name is.
521 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
522 a directory entry by the name of the lookup key. Previously, if a
523 symlink pointed to a non-existing file or a file in a directory that
524 Exim lacked permissions to read, a lookup for a key matching that
525 symlink would fail. Now it is enough that a matching directory entry
526 exists, symlink or not. (Bugzilla 503.)
528 PH/03 The body_linecount and body_zerocount variables are now exported in the
531 PH/04 Added the $dnslist_matched variable.
533 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
534 This means they are set thereafter only if the connection becomes
537 PH/06 Added the client_condition to authenticators so that some can be skipped
538 by clients under certain conditions.
540 PH/07 The error message for a badly-placed control=no_multiline_responses left
541 "_responses" off the end of the name.
543 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
545 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
546 (without spaces) instead of just copying the configuration text.
548 PH/10 Added the /noupdate option to the ratelimit ACL condition.
550 PH/11 Added $max_received_linelength.
552 PH/12 Added +ignore_defer and +include_defer to host lists.
554 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
555 way in which PCRE > 7.0 is built.
557 PH/14 Implemented queue_only_load_latch.
559 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
560 MAIL command. The effect was to mangle the value on 64-bit systems.
562 PH/16 Another patch from the Sieve maintainer.
564 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
566 PH/18 If a system quota error occurred while trying to create the file for
567 a maildir delivery, the message "Mailbox is full" was not appended to the
568 bounce if the delivery eventually timed out. Change 4.67/27 below applied
569 only to a quota excession during the actual writing of the file.
571 PH/19 It seems that peer DN values may contain newlines (and other non-printing
572 characters?) which causes problems in log lines. The DN values are now
573 passed through string_printing() before being added to log lines.
575 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
576 and InterBase are left for another time.)
578 PH/21 Added message_body_newlines option.
580 PH/22 Guard against possible overflow in moan_check_errorcopy().
582 PH/23 POSIX allows open() to be a macro; guard against that.
584 PH/24 If the recipient of an error message contained an @ in the local part
585 (suitably quoted, of course), incorrect values were put in $domain and
586 $local_part during the evaluation of errors_copy.
592 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
593 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
596 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
597 issue a MAIL command.
599 PH/02 In an ACL statement such as
601 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
603 if a client was not listed at all, or was listed with a value other than
604 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
605 the condition was not true (as it should be), so access was not denied.
606 The bug was that the ! inversion was incorrectly passed on to the second
607 item. This has been fixed.
609 PH/03 Added additional dnslists conditions == and =& which are different from
610 = and & when the dns lookup returns more than one IP address.
612 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
613 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
615 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
616 FSYNC, which compiles an option called disable_fsync that allows for
617 bypassing fsync(). The documentation is heavily laced with warnings.
619 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
621 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
622 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
623 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
624 including adding "make clean"; (3) Added -fPIC when compiling the test
625 dynamically loaded module, to get rid of a warning.
627 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
628 message fails, move_frozen_messages = true and ignore_bounce_errors_after
629 = 0s. The bug is otherwise harmless.
631 PH/07 There was a bug in the dovecot authenticator such that the value of
632 $auth1 could be overwritten, and so not correctly preserved, after a
633 successful authentication. This usually meant that the value preserved by
634 the server_setid option was incorrect.
636 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
638 PH/09 Installed PCRE release 7.0.
640 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
641 run for batched SMTP input. It is now run at the start of every message
642 in the batch. While fixing this I discovered that the process information
643 (output by running exiwhat) was not always getting set for -bs and -bS
644 input. This is fixed, and it now also says "batched" for BSMTP.
646 PH/11 Added control=no_pipelining.
648 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
649 patch, slightly modified), and move the expansion of helo_data till after
650 the connection is made in the smtp transport (so it can use these
653 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
655 PH/14 Added log_selector = +pid.
657 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
659 PH/16 Add ${if forany and ${if forall.
661 PH/17 Added dsn_from option to vary the From: line in DSNs.
663 PH/18 Flush SMTP output before performing a callout, unless control =
664 no_callout_flush is set.
666 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
667 was true (the default) a successful delivery failed to delete the retry
668 item, thus causing premature timeout of the address. The bug is now
671 PH/20 Added hosts_avoid_pipelining to the smtp transport.
673 PH/21 Long custom messages for fakedefer and fakereject are now split up
674 into multiline reponses in the same way that messages for "deny" and
675 other ACL rejections are.
677 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
678 with slight modification.
680 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
681 draft, changing the syntax and factoring some duplicate code".
683 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
684 for deliveries of the second and subsequent messages over the same SMTP
687 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
688 ${reduce, with only minor "tidies".
690 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
692 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
693 expansion side effects.
695 PH/27 When a message times out after an over-quota error from an Exim-imposed
696 quota, the bounce message says "mailbox is full". This message was not
697 being given when it was a system quota that was exceeded. It now should
700 MH/03 Made $recipients available in local_scan(). local_scan() already has
701 better access to the recipient list through recipients_list[], but
702 $recipients can be useful in postmaster-provided expansion strings.
704 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
705 in the case of a MAIL command with additional options following the
706 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
707 were accidentally chopped off.
709 PH/29 SMTP synchronization checks are implemented when a command is read -
710 there is a check that no more input is waiting when there shouldn't be
711 any. However, for some commands, a delay in an ACL can mean that it is
712 some time before the response is written. In this time, more input might
713 arrive, invalidly. So now there are extra checks after an ACL has run for
714 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
715 pipelining has not been advertised.
717 PH/30 MH's patch to allow iscntrl() characters to be list separators.
719 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
720 returned in the SMTP response when smtp_return_error_details was false.
723 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
724 instead of the C I/O that was originally supplied, because problems were
727 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
728 Exim which did not show up earlier: it was assuming that a call to
729 SSL_CTX_set_info_callback() might give an error value. In fact, there is
730 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
731 was a macro that became an assignment, so it seemed to work. This has
732 changed to a proper function call with a void return, hence the compile
733 error. Exim's code has been fixed.
735 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
738 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
740 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
742 PH/37 If a message is not accepted after it has had an id assigned (e.g.
743 because it turns out to be too big or there is a timeout) there is no
744 "Completed" line in the log. When some messages of this type were
745 selected by exigrep, they were listed as "not completed". Others were
746 picked up by some special patterns. I have improved the selection
747 criteria to be more general.
749 PH/38 The host_find_failed option in the manualroute router can now be set
750 to "ignore", to completely ignore a host whose IP address cannot be
751 found. If all hosts are ignored, the behaviour is controlled by the new
752 host_all_ignored option.
754 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
755 homing or because of multiple MX records with /mx) generated more than
756 one IP address, and the following item turned out to be the local host,
757 all the secondary addresses of the first item were incorrectly removed
758 from the list, along with the local host and any following hosts (which
759 is what is supposed to happen).
761 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
762 whoever called Exim into the -H file. In the case of the daemon it was
763 behaving confusingly. When first started, it used values for whoever
764 started the daemon, but after a SIGHUP it used the Exim user (because it
765 calls itself on a restart). I have changed the code so that it now always
768 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
769 message are rejected with the same error (e.g. no authentication or bad
770 sender address), and a DATA command is nevertheless sent (as can happen
771 with PIPELINING or a stupid MUA), the error message that was given to the
772 RCPT commands is included in the rejection of the DATA command. This is
773 intended to be helpful for MUAs that show only the final error to their
776 PH/42 Another patch from the Sieve maintainer.
778 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
779 Eximstats - Fixed some broken HTML links and added missing column headers
781 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
782 columns for Rejects, Temp Rejects, Ham, and Spam rows.
784 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
786 PH/43 Yet another patch from the Sieve maintainer.
788 PH/44 I found a way to check for a TCP/IP connection going away before sending
789 the response to the final '.' that terminates a message, but only in the
790 case where the client has not sent further data following the '.'
791 (unfortunately, this is allowed). However, in many cases there won't be
792 any further data because there won't be any more messages to send. A call
793 to select() can be used: if it shows that the input is "ready", there is
794 either input waiting, or the socket has been closed. An attempt to read
795 the next input character can distinguish the two cases. Previously, Exim
796 would have sent an OK response which the client would never have see.
797 This could lead to message repetition. This fix should cure that, at
798 least in a lot of common cases.
800 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
801 advertised in response to EHLO.
807 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
808 fixed by 4.65/MH/01 (is this a record?) are fixed:
810 (i) An empty string was always treated as zero by the numeric comparison
811 operators. This behaviour has been restored.
813 (ii) It is documented that the numeric comparison operators always treat
814 their arguments as decimal numbers. This was broken in that numbers
815 starting with 0 were being interpreted as octal.
817 While fixing these problems I realized that there was another issue that
818 hadn't been noticed. Values of message_size_limit (both the global option
819 and the transport option) were treated as octal if they started with 0.
820 The documentation was vague. These values are now always treated as
821 decimal, and I will make that clear in the documentation.
827 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
828 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
831 MH/01 Don't check that the operands of numeric comparison operators are
832 integers when their expansion is in "skipping" mode (fixes bug
833 introduced by 4.64-PH/07).
835 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
836 child addresses, Exim now panics and dies. Previously, because the count
837 is held in a short int, deliveries were likely to be lost. As such a
838 large number of recipients for a single message is ridiculous
839 (performance will be very, very poor), I have chosen to impose a limit
840 rather than extend the field.
846 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
847 leftover -K file (the existence of which was triggered by #402).
848 While we were at it, introduced process PID as part of the -K
849 filename. This should rule out race conditions when creating
852 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
853 processing considerably. Previous code took too long for large mails,
854 triggering a timeout which in turn triggers #401.
856 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
857 in the DK code in transports.c. sendfile() is not really portable,
858 hence the _LINUX specificness.
860 TF/01 In the add_headers option to the mail command in an Exim filter,
861 there was a bug that Exim would claim a syntax error in any
862 header after the first one which had an odd number of characters
865 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
866 callout verification, Exim cached a "reject" for the entire domain. This
867 is correct for most verifications, but it is not correct for a recipient
868 verification with use_sender or use_postmaster set, because in that case
869 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
870 case of MAIL FROM:<> rejection from other early rejections (e.g.
871 rejection of HELO). When verifying a recipient using a non-null MAIL
872 address, the cache is ignored if it shows MAIL FROM:<> rejection.
873 Whatever the result of the callout, the value of the domain cache is
874 left unchanged (for any other kind of callout, getting as far as trying
875 RCPT means that the domain itself is ok).
877 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
880 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
881 manifest itself as EPIPE rather than ECONNECT. When tidying away a
882 session, the daemon ignores ECONNECT errors and logs others; it now
883 ignores EPIPE as well.
885 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
886 (quoted-printable decoding).
888 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
889 later the small subsequent patch to fix an introduced bug.
891 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
893 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
895 PH/08 An error is now given if message_size_limit is specified negative.
897 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
898 to be given (somewhat) arbitrary names.
900 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
903 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
904 miscellaneous code fixes
906 PH/10 Added the log_reject_target ACL modifier to specify where to log
909 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
910 hostname. This is wrong, because it relates to the incoming message (and
911 probably the interface on which it is arriving) and not to the outgoing
912 callout (which could be using a different interface). This has been
913 changed to use the value of the helo_data option from the smtp transport
914 instead - this is what is used when a message is actually being sent. If
915 there is no remote transport (possible with a router that sets up host
916 addresses), $smtp_active_hostname is used.
918 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
919 tweaks were necessary in order to get it to work (see also 21 below):
920 (a) The code assumed that strncpy() returns a negative number on buffer
921 overflow, which isn't the case. Replaced with Exim's string_format()
923 (b) There were several signed/unsigned issues. I just did the minimum
924 hacking in of casts. There is scope for a larger refactoring.
925 (c) The code used strcasecmp() which is not a standard C function.
926 Replaced with Exim's strcmpic() function.
927 (d) The code set only $1; it now sets $auth1 as well.
928 (e) A simple test gave the error "authentication client didn't specify
929 service in request". It would seem that Dovecot has changed its
930 interface. Fortunately there's a specification; I followed it and
931 changed what the client sends and it appears to be working now.
933 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
936 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
937 address A is aliased to B and C, where B exists and C does not. Without
938 -v the output is "A verified" because verification stops after a
939 successful redirection if more than one address is generated. However,
940 with -v the child addresses are also verified. Exim was outputting "A
941 failed to verify" and then showing the successful verification for C,
942 with its parentage. It now outputs "B failed to verify", showing B's
943 parentage before showing the successful verification of C.
945 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
946 look up a TXT record in a specific list after matching in a combined
949 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
950 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
951 they consult the DNS. I had assumed they would set it the way they
952 wanted; and indeed my experiments on Linux seem to show that in some
953 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
954 To be on the safe side, however, I have now made the interface to
955 host_find_byname() similar to host_find_bydns(), with an argument
956 containing the DNS resolver options. The host_find_byname() function now
957 sets these options at its start, just as host_find_bydns() does. The smtp
958 transport options dns_qualify_single and dns_search_parents are passed to
959 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
960 of host_find_byname() use the default settings of RES_DEFNAMES
961 (qualify_single) but not RES_DNSRCH (search_parents).
963 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
964 spool_read_header() do less string testing, by means of a preliminary
965 switch on the second character of optional "-foo" lines. (This is
966 overdue, caused by the large number of possibilities that now exist.
967 Originally there were few.) While I was there, I also converted the
968 str(n)cmp tests so they don't re-test the leading "-" and the first
969 character, in the hope this might squeeze out yet more improvement.
971 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
972 flag allowing group syntax was set by the header_syntax check but not
973 turned off, possible causing trouble later; (2) The flag was not being
974 set at all for the header_verify test, causing "group"-style headers to
975 be rejected. I have now set it in this case, and also caused header_
976 verify to ignore an empty address taken from a group. While doing this, I
977 came across some other cases where the code for allowing group syntax
978 while scanning a header line wasn't quite right (mostly, not resetting
979 the flag correctly in the right place). These bugs could have caused
980 trouble for malformed header lines. I hope it is now all correct.
982 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
983 with the "reply" argument non-NULL. The code, however (which originally
984 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
985 but it didn't always do it. This confused somebody who was copying the
986 code for some other use. I have removed all the tests.
988 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
989 feature that was used to support insecure browsers during the U.S. crypto
990 embargo. It requires special client support, and Exim is probably the
991 only MTA that supported it -- and would never use it because real RSA is
992 always available. This code has been removed, because it had the bad
993 effect of slowing Exim down by computing (never used) parameters for the
994 RSA_EXPORT functionality.
996 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
997 authenticator to fail if there's a tab character in the incoming data
998 (there should never be unless someone is messing about, as it's supposed
999 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1000 if the connection is using TLS or if the remote IP is the same as the
1001 local IP, and the "valid-client-cert option" if a client certificate has
1004 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1005 authenticators. This can be used for authorization after authentication
1006 succeeds. (In the case of plaintext, it servers for both authentication
1009 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1010 if any retry times were supplied.
1012 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1013 connection, where there is no client IP address to check. In this
1014 situation, the verify now always succeeds.
1016 PH/25 Applied John Jetmore's -Mset patch.
1018 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1020 PH/27 In a string expansion for a processed (not raw) header when multiple
1021 headers of the same name were present, leading whitespace was being
1022 removed from all of them, but trailing whitespace was being removed only
1023 from the last one. Now trailing whitespace is removed from each header
1024 before concatenation. Completely empty headers in a concatenation (as
1025 before) are ignored.
1027 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1028 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1030 PH/29 [Removed. This was a change that I later backed out, and forgot to
1031 correct the ChangeLog entry (that I had efficiently created) before
1032 committing the later change.]
1034 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1035 address errors (4xx response to RCPT) over the same connection as other
1036 messages routed to the same hosts. Such deliveries are always "forced",
1037 so retry times are not inspected. This resulted in far too many retries
1038 for the affected addresses. The effect occurred only when there were more
1039 hosts than the hosts_max_try setting in the smtp transport when it had
1040 the 4xx errors. Those hosts that it had tried were not added to the list
1041 of hosts for which the message was waiting, so if all were tried, there
1042 was no problem. Two fixes have been applied:
1044 (i) If there are any address or message errors in an SMTP delivery, none
1045 of the hosts (tried or untried) are now added to the list of hosts
1046 for which the message is waiting, so the message should not be a
1047 candidate for sending over the same connection that was used for a
1048 successful delivery of some other message. This seems entirely
1049 reasonable: after all the message is NOT "waiting for some host".
1050 This is so "obvious" that I'm not sure why it wasn't done
1051 previously. Hope I haven't missed anything, but it can't do any
1052 harm, as the worst effect is to miss an optimization.
1054 (ii) If, despite (i), such a delivery is accidentally attempted, the
1055 routing retry time is respected, so at least it doesn't keep
1056 hammering the server.
1058 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1059 in ${readsocket because some servers need this prod.
1061 PH/32 Added some extra debug output when updating a wait-xxx database.
1063 PH/33 The hint "could be header name not terminated by colon", which has been
1064 given for certain expansion errors for a long time, was not being given
1065 for the ${if def:h_colon_omitted{... case.
1067 PH/34 The spec says: "With one important exception, whenever a domain list is
1068 being scanned, $domain contains the subject domain." There was at least
1069 one case where this was not true.
1071 PH/35 The error "getsockname() failed: connection reset by peer" was being
1072 written to the panic log as well as the main log, but it isn't really
1073 panic-worthy as it just means the connection died rather early on. I have
1074 removed the panic log writing for the ECONNRESET error when getsockname()
1077 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1078 runs only) independently of the message's sender address. This meant
1079 that, if the 4xx error was in fact related to the sender, a different
1080 message to the same recipient with a different sender could confuse
1081 things. In particualar, this can happen when sending to a greylisting
1082 server, but other circumstances could also provoke similar problems.
1083 I have changed the default so that the retry time for these errors is now
1084 based a combination of the sender and recipient addresses. This change
1085 can be overridden by setting address_retry_include_sender=false in the
1088 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1089 remote server are returned as part of bounce messages. This was not
1090 happening for LMTP over a pipe (the lmtp transport), but now it is the
1091 same for both kinds of LMTP.
1093 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1094 in header lines that were in fact CNAMEs. This is no longer the case.
1096 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1097 and queue runs started by the daemon processed all messages. This has
1098 been fixed so that -R and -S can now usefully be given with -q<time>.
1100 PH/40 Import PCRE release 6.7 (fixes some bugs).
1102 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1104 PH/42 Give an error if -q is specified more than once.
1106 PH/43 Renamed the variables $interface_address and $interface_port as
1107 $received_ip_address and $received_port, to make it clear that these
1108 values apply to message reception, and not to the outgoing interface when
1109 a message is delivered. (The old names remain recognized, of course.)
1111 PH/44 There was no timeout on the connect() call when using a Unix domain
1112 socket in the ${readsocket expansion. There now is.
1114 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1115 be meaningful with "accept".
1117 SC/01 Eximstats V1.43
1118 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1120 SC/02 Eximstats V1.44
1121 Use a glob alias rather than an array ref in the generated
1122 parser. This improves both readability and performance.
1124 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1125 Collect SpamAssassin and rejection statistics.
1126 Don't display local sender or destination tables unless
1127 there is data to show.
1128 Added average volumes into the top table text output.
1130 SC/04 Eximstats V1.46
1131 Collect data on the number of addresses (recipients)
1132 as well as the number of messages.
1134 SC/05 Eximstats V1.47
1135 Added 'Message too big' to the list of mail rejection
1136 reasons (thanks to Marco Gaiarin).
1138 SC/06 Eximstats V1.48
1139 Mainlog lines which have GMT offsets and are too short to
1140 have a flag are now skipped.
1142 SC/07 Eximstats V1.49 (Alain Williams)
1143 Added the -emptyok flag.
1145 SC/08 Eximstats V1.50
1146 Fixes for obtaining the IP address from reject messages.
1148 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1149 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1150 whitesspace changes from 4.64-PH/27
1152 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1155 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1156 are found, allow negative numbers in numeric criteria)
1158 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1160 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1161 to match changes made in 4.64-PH/43
1163 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1165 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1166 "domain" is actually an IP address, and force a failure. This locks out
1167 those revolvers/nameservers that support "A-for-A" lookups, in
1168 contravention of the specifications.
1170 PH/48 When a host name was looked up from an IP address, and the subsequent
1171 forward lookup of the name timed out, the host name was left in
1172 $sender_host_name, contrary to the specification.
1174 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1175 restricted to single-key lookups, Exim was not diagnosing an error if
1176 * or *@ was used with a query-style lookup.
1178 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1180 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1181 long ago, but noone interested enough thought of it. Let's just say that
1182 the "1.1" means that there are some new functions that weren't there at
1183 some point in the past.
1185 PH/51 Error processing for expansion failure of helo_data from an smtp
1186 transport during callout processing was broken.
1188 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1189 tested/used via the -bh/-bhc/-bs options.
1191 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1192 bug, fixed in subsequent PCRE releases).
1194 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1195 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1197 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1203 SC/01 Use a glob alias rather than an array ref in eximstats generated
1204 parser. This improves both readability and performance.
1206 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1207 Don't display local sender or destination tables in eximstats unless
1208 there is data to show.
1209 Added average volumes into the eximstats top table text output.
1211 SC/03 Collect data on the number of addresses (recipients) as well
1212 as the number of messages in eximstats.
1214 TF/01 Correct an error in the documentation for the redirect router. Exim
1215 does (usually) call initgroups() when daemonizing.
1217 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1218 with consistent privilege compared to when running as a daemon.
1220 TF/03 Note in the spec that $authenticated_id is not set for local
1221 submissions from trusted users.
1223 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1224 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1226 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1227 by adding some example configuration directives to the default
1228 configuration file. A little bit of work is required to uncomment the
1229 directives and define how usernames and passwords are checked, but
1230 there is now a framework to start from.
1232 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1233 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1234 without this. I don't know how relevant this is to other LDAP libraries.
1236 PH/02 Add the verb name to the "unknown ACL verb" error.
1238 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1240 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1242 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1243 directory not expanded when it should be if an expanded home directory
1244 was set for the address (which is overridden by the transport).
1246 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1249 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1250 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1251 because it is too late at that time, and has no effect.
1253 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1254 security issue with \' (bugzilla #107). I could not use the
1255 PQescapeStringConn() function, because it needs a PGconn value as one of
1258 PH/08 When testing addresses using -bt, indicate those final addresses that
1259 are duplicates that would not cause an additional delivery. At least one
1260 person was confused, thinking that -bt output corresponded to deliveries.
1261 (Suppressing duplicates isn't a good idea as you lose the information
1262 about possibly different redirections that led to the duplicates.)
1264 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1265 systems where poll() doesn't work, in particular OS X.
1267 PH/10 Added more information to debugging output for retry time not reached.
1269 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1270 operations in malware.c.
1272 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1275 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1276 syslog_duplication set false, log lines that would normally be written
1277 both the the main log and to the reject log were not written to syslog at
1280 PH/14 In the default configuration, change the use of "message" in ACL warn
1281 statements to "add_header".
1283 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1284 not followed by a command (e.g. "seen endif").
1286 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1287 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1290 PH/17 Added extra conditions to the default value of delay_warning_condition
1294 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1295 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1296 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1299 The Auto-Submitted: and various List- headers are standardised, whereas I
1300 don't think Precedence: ever was.
1302 PH/18 Refactored debugging code in route_finduser() to show more information,
1303 in particular, the error code if getpwnam() issues one.
1305 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1306 This is apparently needed in addition to the PH/07 change above to avoid
1307 any possible encoding problems.
1309 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1310 but not after initializing Perl.
1312 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1313 output them only if debugging. By default they are written stderr,
1314 apparently, which is not desirable.
1316 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1319 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1322 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1324 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1325 authenticated or an ident call has been made. Suppress the default
1326 values for $authenticated_id and $authenticated_sender (but permit -oMai
1327 and -oMas) when testing with -bh.
1329 PH/24 Re-jigged the order of the tests in the default configuration so that the
1330 tests for valid domains and recipients precede the DNS black list and CSA
1331 tests, on the grounds that those ones are more expensive.
1333 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1334 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1335 This bug exists in every version of Exim that I still have, right back to
1338 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1339 However, an attempt to turn on case-sensitivity in a regex key by
1340 including (?-i) didn't work because the subject string was already
1341 lowercased, and the effects were non-intuitive. It turns out that a
1342 one-line patch can be used to allow (?-i) to work as expected.
1348 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1349 other effects) broke the use of negated acl sub-conditions.
1351 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1354 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1355 "Deny" causes Exim to reject the incoming connection with a 554 error.
1356 Unfortunately, if there is a major crisis, such as a disk failure,
1357 tcp-wrappers gives "deny", whereas what one would like would be some
1358 kind of temporary error. A kludge has been added to help with this.
1359 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1360 554 error is used if errno is still zero or contains ENOENT (which occurs
1361 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1364 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1366 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1367 errors. Otherwise a message that provokes a temporary error (when other
1368 messages do not) can cause a whole host to time out.
1370 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1371 addresses were routed directly to files or pipes from a redirect router.
1372 File deliveries just didn't batch; pipe deliveries might have suffered
1375 PH/06 A failure to get a lock for a hints database would erroneously always say
1376 "Failed to get write lock", even when it was really a read lock.
1378 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1379 of 0600. This has been changed to use the value of the lockfile_mode
1380 option (which defaults to 0600).
1382 PH/08 Applied small patch from the Sieve maintainer.
1384 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1385 folder from quota calculations, a direct delivery into this folder messed
1386 up the contents of the maildirsize file. This was because the regex was
1387 used only to exclude .Trash (or whatever) when the size of the mailbox
1388 was calculated. There was no check that a delivery was happening into an
1389 excluded directory. This bug has been fixed by ignoring all quota
1390 processing for deliveries into excluded directories.
1392 PH/10 Added the maildirfolder_create_regex option to appendfile.
1398 PH/01 The code for finding all the local interface addresses on a FreeBSD
1399 system running IPv6 was broken. This may well have applied to all BSD
1400 systems, as well as to others that have similar system calls. The broken
1401 code found IPv4 interfaces correctly, but gave incorrect values for the
1402 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1403 that it would not match correctly against @[] and not recognize the IPv6
1406 PH/02 The ipliteral router was not recognizing addresses of the form user@
1407 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1409 PH/03 Added disable_ipv6.
1411 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1412 decoded form, because it is most often used to construct To: headers
1413 lines in autoreplies, and the decoded form may well be syntactically
1414 invalid. However, $reply_address has leading white space removed, and all
1415 newlines turned into spaces so that the autoreply transport does not
1418 PH/05 If group was specified without a user on a router, and no group or user
1419 was specified on a transport, the group from the router was ignored.
1421 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1422 for visible compile-time settings that can be used to change these
1423 numbers, for those that want even more. Backwards compatibility with old
1424 spool files has been maintained. However, going back to a previous Exim
1425 release will lost any variables that are in spool files.
1427 PH/07 Two small changes when running in the test harness: increase delay when
1428 passing a TCP/IP connection to a new process, in case the original
1429 process has to generate a bounce, and remove special handling of
1430 127.0.0.2 (sic), which is no longer necessary.
1432 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1433 be the same on different OS.
1435 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1438 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1439 whether --show-vars was specified or not
1441 JJ/02 exipick: Added support for new ACL variable spool format introduced
1444 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1445 syntactically invalid From: or Reply-to: line, and a filter used this to
1446 generate an autoreply, and therefore failed to obtain an address for the
1447 autoreply, Exim could try to deliver to a non-existent relative file
1448 name, causing unrelated and misleading errors. What now happens is that
1449 it logs this as a hard delivery error, but does not attempt to create a
1452 PH/11 The exinext utility has a -C option for testing purposes, but although
1453 the given file was scanned by exinext itself; it wasn't being passed on
1454 when Exim was called.
1456 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1457 an end-of-file indication when reading a command response.
1459 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1460 compiled. In many other places in Exim, IPv6 addresses are always
1461 recognized, so I have changed this. It also means that IPv4 domain
1462 literals of the form [IPV4:n.n.n.n] are now always recognized.
1464 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1465 used if the router is not running as root, for example, when verifying at
1466 ACL time, or when using -bh. The debugging output from this situation was
1467 non-existent - all you got was a failure to exec. I have made two
1470 (a) Failures to set uid/gid, the current directory, or a process leader
1471 in a subprocess such as that created by queryprogram now generate
1472 suitable debugging ouput when -d is set.
1474 (b) The queryprogram router detects when it is not running as root,
1475 outputs suitable debugging information if -d is set, and then runs
1476 the subprocess without attempting to change uid/gid.
1478 PH/15 Minor change to Makefile for building test_host (undocumented testing
1481 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1482 additional section of a DNS packet that returns MX or SRV records.
1483 Instead, it always explicitly searches for A/AAAA records. This avoids
1484 major problems that occur when a DNS server includes only records of one
1485 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1486 fixed another bug: if SRV records were looked up and the corresponding
1487 address records were *not* found in the additional section, the port
1488 values from the SRV records were lost.
1490 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1491 using the correct key (the original address) when searching the retry
1492 rules in order to find which one to use for generating the retry hint.
1494 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1495 adding the default one. Similarly, if it contains a Reply-To: header, the
1496 errors_reply_to option, if set, is not used.
1498 PH/19 When calculating a retry time, Exim used to measure the "time since
1499 failure" by looking at the "first failed" field in the retry record. Now
1500 it does not use this if it is later than than the arrival time of the
1501 message. Instead it uses the arrival time. This makes for better
1502 behaviour in cases where some deliveries succeed, thus re-setting the
1503 "first failed" field. An example is a quota failure for a huge message
1504 when small messages continue to be delivered. Without this change, the
1505 "time since failure" will always be short, possible causing more frequent
1506 delivery attempts for the huge message than are intended.
1507 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1509 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1510 $1, $2, $3) because the numerical variables can be reset during some
1511 expansion items (e.g. "match"), thereby losing the authentication data.
1513 PH/21 Make -bV show the size of off_t variables so that the test suite can
1514 decide whether to run tests for quotas > 2G.
1516 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1517 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1518 filecount value is greater than 2G or if a quota value is greater than 2G
1519 on a system where the size of off_t is not greater than 4, a panic error
1522 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1523 never match. The debug and -bh output now contains an explicit error
1524 message indicating a malformed IPv4 address or mask.
1526 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1527 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1528 PH/23 above applies.
1530 PH/25 Do not write to syslog when running in the test harness. The only
1531 occasion when this arises is a failure to open the main or panic logs
1532 (for which there is an explicit test).
1534 PH/26 Added the /no_tell option to "control=freeze".
1536 PH/27 If a host name lookup failed very early in a connection, for example, if
1537 the IP address matched host_lookup and the reverse lookup yielded a name
1538 that did not have a forward lookup, an error message of the form "no IP
1539 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1540 could be logged. Now it outputs the IP address instead of "NULL".
1542 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1543 allows the sender and the authenticated sender to be set when
1544 submitting a message from within Exim. Since child_open_exim() is
1545 documented for local_scan(), the new function should be too.
1547 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1548 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1549 results in an empty string is now treated as unset.
1551 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1553 PH/31 Added sender_verify_fail logging option.
1555 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1556 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1557 tidied the source and removed it altogether.
1559 PH/33 When a queue run was abandoned because the load average was too high, a
1560 log line was always written; now it is written only if the queue_run log
1561 selector is set. In addition, the log line for abandonment now contains
1562 information about the queue run such as the pid. This is always present
1563 in "start" and "stop" lines but was omitted from the "abandon" line.
1565 PH/34 Omit spaces between a header name and the colon in the error message that
1566 is given when verify = headers_syntax fails (if there are lots of them,
1567 the message gets confusing).
1569 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1570 names, as there are now some PTR records that contain slashes. This check
1571 is only to protect against broken name servers that fall over on strange
1572 characters, so the fact that it applies to all lookups doesn't matter.
1574 PH/36 Now that the new test suite is complete, we can remove some of the
1575 special code in Exim that was needed for the old test suite. For example,
1576 sorting DNS records because real resolvers return them in an arbitrary
1577 order. The new test suite's fake resolver always returns records in the
1580 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1581 bounces) except when queue_only is set, to avoid logging races between
1582 the different processes.
1584 PH/38 Panic-die if .include specifies a non-absolute path.
1586 PH/39 A tweak to the "H" retry rule from its user.
1588 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1589 a label. They prevented compilation on older perls.
1591 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1592 a warning to be raised on newish perls.
1594 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1595 on queue. Changes to match documented behaviour of showing count of
1596 messages matching specified criteria.
1598 PH/40 Changed the default ident timeout from 30s to 5s.
1600 PH/41 Added support for the use of login_cap features, on those BSD systems
1601 that have them, for controlling the resources used by pipe deliveries.
1603 PH/42 The content-scanning code uses fopen() to create files in which to put
1604 message data. Previously it was not paying any attention to the mode of
1605 the files. Exim runs with umask(0) because the rest of the code creates
1606 files with open(), and sets the required mode explicitly. Thus, these
1607 files were ending up world-writeable. This was not a big issue, because,
1608 being within the spool directory, they were not world-accessible. I have
1609 created a function called modefopen, which takes an additional mode
1610 argument. It sets umask(777), creates the file, chmods it to the required
1611 mode, then resets the umask. All the relevant calls to fopen() in the
1612 content scanning code have been changed to use this function.
1614 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1615 to 24 hours. This avoids potential overflow problems when processing G
1616 and H retry rules. I suspect nobody ever tinkers with this value.
1618 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1620 PH/45 When the plaintext authenticator is running as a client, the server's
1621 challenges are checked to ensure they are valid base64 strings. By
1622 default, the authentication attempt is cancelled if an invalid string is
1623 received. Setting client_ignore_invalid_base64 true ignores these errors.
1624 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1625 they are received. Thus, the responses can be made to depend on the
1626 challenges. If an invalid string is ignored, an empty string is placed in
1629 PH/46 Messages that are created by the autoreply transport now contains a
1630 References: header, in accordance with RFCs 2822 and 3834.
1632 PH/47 Added authenticated_sender_force to the smtp transport.
1634 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1636 PH/49 Installed latest patch from the Sieve maintainer.
1638 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1639 was also set, the appendfile transport was unnecessarily scanning a
1640 directory of message files (e.g. for maildir delivery) to find the count
1641 of files (along with the size), even though it did not need this
1642 information. It now does the scan only if it needs to find either the
1643 size of the count of files.
1645 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1647 PH/52 Two bugs concerned with error handling when the smtp transport is
1650 (i) Exim was not creating retry information for temporary errors given
1651 for individual recipients after the DATA command when the smtp transport
1652 was used in LMTP mode. This meant that they could be retried too
1653 frequently, and not timed out correctly.
1655 (ii) Exim was setting the flag that allows error details to be returned
1656 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1657 recipients that were returned after the DATA command.
1659 PH/53 This is related to PH/52, but is more general: for any failing address,
1660 when detailed error information was permitted to be returned to the
1661 sender, but the error was temporary, then after the final timeout, only
1662 "retry timeout exceeded" was returned. Now it returns the full error as
1663 well as "retry timeout exceeded".
1665 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1666 do this, and (what is worse) MTAs that accept it.
1668 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1669 will now be deprecated.
1671 PH/56 New os.c-cygwin from the Cygwin maintainer.
1673 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1674 formats (previously only available in exim formats via -bpr, -bpru,
1675 and -bpra. Now also available in native and exiqgrep formats)
1677 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1678 with very large, slow to parse queues
1680 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1682 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1684 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1685 responses to authentication challenges, though it was showing the
1686 challenges; (ii) I've removed the CR characters from the debug output for
1689 PH/58 Allow for the insertion of a newline as well as a space when a string
1690 is turned into more than one encoded-word during RFC 2047 encoding. The
1691 Sieve code now uses this.
1693 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1694 data_4xx, lost_connection, tls_required.
1696 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1697 message was being sent as an SMTP response.
1699 PH/61 Add -l and -k options to exicyclog.
1701 PH/62 When verifying, if an address was redirected to one new address, so that
1702 verification continued, and the new address failed or deferred after
1703 having set something in $address_data, the value of $address_data was not
1704 passed back to the ACL. This was different to the case when no
1705 redirection occurred. The value is now passed back in both cases.
1707 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1708 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1709 use login_cap.h, so on its own it isn't the distinguishing feature. The
1710 new name refers directly to the setclassresources() function.
1712 PH/65 Added configuration files for NetBSD3.
1714 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1716 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1717 is preferred over IPv4.
1719 PH/68 The bounce_return_message and bounce_return_body options were not being
1720 honoured for bounces generated during the reception of non-SMTP messages.
1721 In particular, this applied to messages rejected by the ACL. This bug has
1722 been fixed. However, if bounce_return_message is true and bounce_return_
1723 body is false, the headers that are returned for a non-SMTP message
1724 include only those that have been read before the error was detected.
1725 (In the case of an ACL rejection, they have all been read.)
1727 PH/69 The HTML version of the specification is now built in a directory called
1728 spec_html instead of spec.html, because the latter looks like a path with
1729 a MIME-type, and this confuses some software.
1731 PH/70 Catch two compiler warnings in sieve.c.
1733 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1734 function verify_get_ident() calls ip_connect() to connect a socket, but
1735 if the "connect()" function timed out, ip_connect() used to close the
1736 socket. However, verify_get_ident() also closes the socket later, and in
1737 between Exim writes to the log, which may get opened at this point. When
1738 the socket was closed in ip_connect(), the log could get the same file
1739 descriptor number as the socket. This naturally causes chaos. The fix is
1740 not to close the socket in ip_connect(); the socket should be closed by
1741 the function that creates it. There was only one place in the code where
1742 this was missing, in the iplookup router, which I don't think anybody now
1743 uses, but I've fixed it anyway.
1745 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1746 well as to direct DNS lookups. Otherwise the handling of names in host
1747 lists is inconsistent and therefore confusing.
1753 PH/01 Two changes to the default runtime configuration:
1755 (1) Move the checks for relay_from_hosts and authenticated clients from
1756 after to before the (commented out) DNS black list checks.
1758 (2) Add control=submission to the relay_from_hosts and authenticated
1759 clients checks, on the grounds that messages accepted by these
1760 statements are most likely to be submissions.
1762 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1764 (1) Generate an error if the third argument for the ${prvs expansion is
1767 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1770 (3) Reset the variables that are obtained from the first argument of
1771 ${prvscheck and used in the second argument before leaving the code,
1772 because their memory is reclaimed, so using them afterwards may do
1775 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1776 one (it's much easier than Tom thought :-).
1778 (5) Because of (4), we can now allow for the use of $prvscheck_result
1779 inside the third argument.
1781 PH/03 For some reason, the default setting of PATH when running a command from
1782 a pipe transport was just "/usr/bin". I have changed it to
1785 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1786 anything to be listed in the output from -bV.
1788 PH/05 When a filter generated an autoreply, the entire To: header line was
1789 quoted in the delivery log line, like this:
1791 => >A.N.Other <ano@some.domain> <original@ddress> ...
1793 This has been changed so that it extracts the operative address. There
1794 may be more than one such address. If so, they are comma-separated, like
1797 => >ano@some.domain,ona@other.domain <original@ddress> ...
1799 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1800 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1801 not being looked up in the rDNS to get a host name, Exim was showing the
1802 IP address twice in Received: lines, even though the IP addresses were
1803 identical. For example:
1805 Received: from [1.2.3.4] (helo=[1.2.3.4])
1807 However, if the real host name was known, it was omitting the HELO data
1808 if it matched the actual IP address. This has been tidied up so that it
1809 doesn't show the same IP address twice.
1811 PH/07 When both +timestamp and +memory debugging was on, the value given by
1812 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1813 called by the debug printing, thereby overwriting the timestamp buffer.
1814 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1816 PH/08 When the original message was included in an autoreply transport, it
1817 always said "this is a copy of the message, including all the headers",
1818 even if body_only or headers_only was set. It now gives an appropriate
1821 PH/09 Applied a patch from the Sieve maintainer which:
1823 o fixes some comments
1824 o adds the (disabled) notify extension core
1825 o adds some debug output for the result of if/elsif tests
1826 o points to the current vacation draft in the documentation
1827 and documents the missing references header update
1831 o fixes a bug in processing the envelope test (when testing
1832 multiple envelope elements, the last element determinted the
1835 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1836 Electronic Mail") by including:
1838 Auto-submitted: auto-generated
1840 in the messages that it generates (bounce messages and others, such as
1841 warnings). In the case of bounce messages for non-SMTP mesages, there was
1842 also a typo: it was using "Auto_submitted" (underscore instead of
1843 hyphen). Since every message generated by Exim is necessarily in response
1844 to another message, thes have all been changed to:
1846 Auto-Submitted: auto-replied
1848 in accordance with these statements in the RFC:
1850 The auto-replied keyword:
1852 - SHOULD be used on messages sent in direct response to another
1853 message by an automatic process,
1855 - MUST NOT be used on manually-generated messages,
1857 - MAY be used on Delivery Status Notifications (DSNs) and Message
1858 Disposition Notifications (MDNs),
1860 - MUST NOT be used on messages generated by automatic or periodic
1861 processes, except for messages which are automatic responses to
1864 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1865 to the default Received: header definition.
1867 PH/12 Added log selector acl_warn_skipped (default on).
1869 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1870 variables because (a) they are in the wrong storage pool and (b) even if
1871 they were copied, it wouldn't work properly because of the caching.
1873 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1874 checking when decoding. Apparently there are clients that generate
1875 overlong encoded strings. Why am I not surprised?
1877 PH/15 If the first argument of "${if match_address" was not empty, but did not
1878 contain an "@" character, Exim crashed. Now it writes a panic log message
1879 and treats the condition as false.
1881 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1883 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1884 extension "envelope-auth". The code is finished and in agreement with
1885 other implementations, but there is no documentation so far and in fact,
1886 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1887 not changing the active code.
1889 Print executed "if" and "elsif" statements when debugging is used. This
1890 helps a great deal to understand what a filter does.
1892 Document more things not specified clearly in RFC3028. I had all this
1893 sorted out, when out of a sudden new issues came to my mind. Oops."
1895 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1898 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1899 canonical form (as documented). However, after a host name lookup from
1900 the IP address, check_host() was doing a simple string comparison with
1901 addresses acquired from the DNS when checking that the found name did
1902 have the original IP as one of its addresses. Since any found IPv6
1903 addresses are likely to be in abbreviated form, the comparison could
1904 fail. Luckily, there already exists a function for doing the comparison
1905 by converting both addresses to binary, so now that is used instead of
1906 the text comparison.
1908 PH/20 There was another similar case to PH/19, when a complete host name was
1909 given in a host list; looking up its IP address could give an abbreviated
1910 form, whereas the current host's name might or might not be abbreviated.
1911 The same fix has been applied.
1917 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1918 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1921 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1922 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1924 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1926 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1927 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1928 Also, for "Auto-Submitted", treat anything other than "no" as
1929 non-personal, in accordance with RFC 3834. (Previously it treated
1930 anything starting "auto-" as non-personal.)
1932 TF/01 The control=submission/name=... option had a problem with syntax
1933 errors if the name included a slash character. The /name= option
1934 now slurps the rest of the string, so it can include any characters
1935 but it must come last in the list of options (after /sender_retain
1938 PH/05 Some modifications to the interface to the fake nameserver for the new
1946 TK/01 Added the "success_on_redirect" address verification option. See
1947 NewStuff for rationale and an example.
1949 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1951 PH/02 Patch to exigrep to allow it to work on syslog lines.
1953 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1954 fread() to skip over the body file's header line, because in Cygwin the
1955 header line is locked and is inaccessible.
1957 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1958 co-exist for some time) to make it clear that it is the Exim ID that is
1959 referenced, not the Message-ID: header line.
1961 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1962 string_format() function, because snprintf() does not exist on all
1965 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1966 ${stat: expansion item.
1968 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1969 protocol synchronization error", to keep the pedants happy.
1971 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1972 well as for IRIX systems, when gcc is being used. See the host.c source
1975 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1977 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1980 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1981 transport and to the smtp transport in LMTP mode.
1983 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1985 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1986 run a filter in a subprocess. This could lead to confusion in subsequent
1987 lookups in the parent process. There should also be a search_tidyup() at
1988 the end of the subprocess.
1990 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1991 only if the host matched helo_try_verify_hosts, which caused the
1992 verification to occur when the EHLO/HELO command was issued. The ACL just
1993 tested the remembered result. Now, if a previous verification attempt has
1994 not happened, "verify = helo" does it there and then.
1996 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1998 TK/03 Fix log output including CR from clamd.
2000 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2001 exist provoked a memory error which could cause a segfault.
2003 PH/15 Installed PCRE 6.2
2005 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2007 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2008 of the problem. Specifically, suggested +O2 rather than +O1 for the
2011 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2013 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2014 record, the port was not correctly logged when the outgoing_port log
2015 selector was set (it logged the transort's default port).
2017 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2018 fallback_hosts, and "hosts" in the smtp transport.
2020 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2021 host errors such as "Connection refused".
2023 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2024 authentication with radiusclient 0.4.9:
2026 - Error returned from rc_read_config was caught wrongly
2027 - Username/password not passed on to radius server due to wrong length.
2029 The presumption is that some radiusclient API changes for 4.51/PH/17
2030 were not taken care of correctly. The code is still untested by me (my
2031 Linux distribution still has 0.3.2 of radiusclient), but it was
2032 contributed by a Radius user.
2034 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2035 expanding the "port" option of the smtp transport.
2037 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2038 while reading a MIME header. Thanks to Tom Hughes for a patch.
2040 PH/24 Include config.h inside local_scan.h so that configuration settings are
2043 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2044 that in an ACL for RCPT (for example), you can examine exactly what was
2047 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2048 commands, but it was not correctly comparing the address with the actual
2049 client host address. Thus, it would show the EHLO address in Received:
2050 header lines when this was not necessary.
2052 PH/27 Added the % operator to ${eval:}.
2054 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2055 it should be ignoring failures (because with -C, for example, it has lost
2056 privilege). It wasn't ignoring creation failures other than "already
2059 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2062 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2063 because an input line was too long, either on its own, or by virtue of
2064 too many continuations, the temporary file was not being removed, and the
2065 return code was incorrect.
2067 PH/31 Missing "BOOL" in function definition in filtertest.c.
2069 PH/32 Applied Sieve patches from the maintainer.
2071 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2073 PH/33 Added "verify = not_blind".
2075 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2076 Local/Makefile (with some defaults set). These are used in built scripts
2077 such as exicyclog, but they have never been used in the exim_install
2078 script (though there are many overriding facilities there). I have
2079 arranged that the exim_install script now takes note of these two
2082 PH/35 Installed configuration files for Dragonfly.
2084 PH/36 When a locally submitted message by a trusted user did not contain a
2085 From: header, and the sender address was obtained from -f or from an SMTP
2086 MAIL command, and the trusted user did not use -F to supply a sender
2087 name, $originator_name was incorrectly used when constructing a From:
2088 header. Furthermore, $originator_name was used for submission mode
2089 messages from external hosts without From: headers in a similar way,
2090 which is clearly wrong.
2092 PH/37 Added control=suppress_local_fixups.
2094 PH/38 When log_selector = +received_sender was set, and the addition of the
2095 sender made the log line's construction buffer exactly full, or one byte
2096 less than full, an overflow happened when the terminating "\n" was
2099 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2100 when the result of a list match is failure because a DNS lookup failed.
2102 PH/40 RM_COMMAND is now used in the building process.
2104 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2105 the "build-* directories that it finds.
2107 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2108 address in a domain literal was a prefix of an interface address.
2110 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2111 when verifying a sender address, unless rewrite_headers is false.
2113 PH/44 Wrote a long comment about why errors_to addresses are verified as
2114 recipients, not senders.
2116 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2117 the ratelimit ACL was added.
2119 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2121 PH/46 Added extra information about PostgreSQL errors to the error string.
2123 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2124 avoiding the need to install special zones in a real server. This is
2125 backwards compatible; if it can't find the fake resolver, it drops back.
2126 Thus, both old and new test suites can be run.
2128 TF/02 Added util/ratelimit.pl
2130 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2131 clock is set back in time.
2133 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2134 Candler <B.Candler@pobox.com>.
2136 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2137 OK for addresses that are the result of redirections.
2139 PH/48 A number of further additions for the benefit of the new test suite,
2140 including a fake gethostbyname() that interfaces to the fake DNS resolver
2143 TF/06 The fix for widen_domains has also been applied to qualify_single and
2144 search_parents which are the other dnslookup options that can cause
2147 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2150 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2152 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2153 service" if it encountered an MX record with an empty target hostname.
2154 The message is now "an MX or SRV record indicated no SMTP service".
2156 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2157 if the DNS of the sending site is misconfigured. This is quite a
2158 common situation. This change restores the behaviour of treating a
2159 helo verification defer as a failure.
2161 PH/51 If self=fail was set on a router, the bounce message did not include the
2162 actual error message.
2168 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2170 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2171 command itself ended in error, the underlying message about the transport
2172 filter timeout was being overwritten with the pipe command error. Now the
2173 underlying error message should be appended to the second error message.
2175 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2177 PH/02 Reduce the amount of output that "make" produces by default. Full output
2178 can still be requested.
2180 PH/03 The warning log line about a condition test deferring for a "warn" verb
2181 was being output only once per connection, rather than after each
2182 occurrence (because it was using the same function as for successful
2183 "warn" verbs). This seems wrong, so I have changed it.
2185 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2186 it should not have, which might have caused a crash in the right
2187 circumstances, but probably never did.
2189 PH/04 Installed a modified version of Tony Finch's patch to make submission
2190 mode fix the return path as well as the Sender: header line, and to
2191 add a /name= option so that you can make the user's friendly name appear
2194 TF/03 Added the control = fakedefer ACL modifier.
2196 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2197 Mark Lowes for thorough testing.
2199 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2201 TK/03 Merged latest SRS patch from Miles Wilton.
2203 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2204 with the definition in sysexits.h (which is #included earlier).
2205 Fortunately, Exim does not actually use EX_OK. The code used to try to
2206 preserve the sysexits.h value, by assumimg that macro definitions were
2207 scanned for macro replacements. I have been disabused of this notion,
2208 so now the code just undefines EX_OK before #including unistd.h.
2210 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2211 in the smtp transport. When a block could not be written in a single
2212 write() function, the timeout was being re-applied to each part-write.
2213 This seems wrong - if the receiver was accepting one byte at a time it
2214 would take for ever. The timeout is now adjusted when this happens. It
2215 doesn't have to be particularly precise.
2217 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2218 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2220 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2221 without a domain if the check to <postmaster@domain> fails.
2223 SC/01 Eximstats: added -xls and the ability to specify output files
2224 (patch written by Frank Heydlauf).
2226 SC/02 Eximstats: use FileHandles for outputing results.
2228 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2230 SC/04 Eximstats: fixed display of large numbers with -nvr option
2232 SC/05 Eximstats: fixed merging of reports with empty tables.
2234 SC/06 Eximstats: added the -include_original_destination flag
2236 SC/07 Eximstats: removed tabs and trailing whitespace.
2238 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2240 TK/06 MBOX spool code: Add real "From " MBOX separator line
2241 so the .eml file is really in mbox format (even though
2242 most programs do not really care). Patch from Alex Miller.
2244 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2245 The latter is generated from $received_to and is only set if the
2246 message has one envelope recipient. SA can use these headers,
2247 obviously out-of-the-box. Patch from Alex Miller.
2249 PH/08 The ${def test on a variable was returning false if the variable's
2250 value was "0", contrary to what the specification has always said!
2251 The result should be true unless the variable is empty.
2253 PH/09 The syntax error of a character other than { following "${if
2254 def:variable_name" (after optional whitespace) was not being diagnosed.
2255 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2256 accidental colon was present, for example, could give incorrect results.
2258 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2259 result is used (not including appendfile, where other changes are about
2262 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2263 This involved changing a lot of size variables from int to off_t. It
2264 should work with maildirs and everything.
2266 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2267 spamd dying while we are connected to it.
2269 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2272 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2273 function for BDB 4.3.
2275 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2277 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2278 the sources for PCRE have been re-arranged and more files are now
2281 PH/15 The code I had for printing potentially long long variables in PH/11
2282 above was not the best (it lost precision). The length of off_t variables
2283 is now inspected at build time, and an appropriate printing format (%ld
2284 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2285 to be "long long int" or "long int". This is needed for the internal
2286 formatting function string_vformat().
2288 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2289 the configuration file to be ":syslog", then the script "guesses" where
2290 the logs files are, rather than using the compiled in default. In our
2291 case the guess is not the same as the compiled default, so the script
2292 suddenly stopped working when I started to use syslog. The patch checks
2293 to see if log_file_path is "". If so, it attempts to read it from exim
2294 with no configuration file to get the compiled in version, before it
2295 falls back to the previous guessing code."
2297 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2298 implementing BATV in an Exim configuration. See NewStuff for the gory
2301 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2302 Makefile that are specific to HP-UX.
2304 PH/18 If the "use_postmaster" option was set for a recipient callout together
2305 with the "random" option, the postmaster address was used as the MAIL
2306 FROM address for the random test, but not for the subsequent recipient
2307 test. It is now used for both.
2309 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2310 patch removes a few documentation additions to RFC 3028, because the
2311 latest draft now contains them. It adds the new en;ascii-case comparator
2312 and a new error check for 8bit text in MIME parts. Comparator and
2313 require names are now matched exactly. I enabled the subaddress
2314 extension, but it is not well tested yet (read: it works for me)."
2316 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2317 rework some of the code of TK/09 above to avoid the hardwired use of
2318 "%lld" and "long long". Replaced the call to snprintf() with a call to
2321 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2322 records point to non-existent hosts", "retry timeout exceeded", and
2323 "retry time not reached for any host after a long failure period".
2325 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2326 experimental DomainKeys support:
2328 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2329 (2) On an error such as an illegally used "control", the wrong name for
2330 the control was given.
2332 These problems did NOT occur unless DomainKeys support was compiled.
2334 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2336 PH/24 Added ${if match_ip condition.
2338 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2339 databases so that it will be absolutely obvious if a crash occurs in the
2340 DB library. This is a regular occurrence (often caused by mis-matched
2343 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2344 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2345 picked up on a user's system that detects such things. There doesn't seem
2346 to be a gcc warning option for this - only an attribute that has to be
2347 put on the function's prototype. It seems that in Fedora Core 4 they have
2348 set this on a number of new functions. No doubt there will be more in due
2351 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2352 specify a transport. However, if an address that was verified by such a
2353 router was the subject of a callout, Exim crashed because it tried to
2354 read the rcpt_include_affixes from the non-existent transport. Now it
2355 just assumes that the setting of that option is false. This bug was
2356 introduced by 4.51/PH/31.
2358 PH/28 Changed -d+all to exclude +memory, because that information is very
2359 rarely of interest, but it makes the output a lot bigger. People tend to
2360 do -d+all out of habit.
2362 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2363 code in os-type was giving problems when libc.so lives in lib64, like on
2366 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2367 aren't the modern standard, and it seems that some systems' include files
2368 don't always have them. Exim was already checking for some of the newer
2369 ones like T_AAAA, and defining it itself. I've added checks for all the
2370 record types that Exim uses.
2372 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2373 not automatically generating a new one, as it is supposed to. This
2374 prevented TLS from working. If the file did exist, but contained invalid
2375 data, a new version was generated, as expected. It was only the case of a
2376 non-existent file that was broken.
2378 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2379 with a change in libdomainkeys > 0.64.
2381 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2382 from DNS. If the selector record carries the flag, it now has
2383 precedence over the domain-wide flag.
2385 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2387 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2388 the use of a port name that isn't defined in /etc/services) occurred, the
2389 message was deferred as in a normal delivery, and thus remained on the
2390 spool, instead of being failed because of the mua_wrapper setting. This
2391 is now fixed, and I tidied up some of the mua_wrapper messages at the
2394 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2395 the messages in a hash of arrays rather than using individual hashes.
2396 This is a bit cleaner and results in dramatic memory savings, albeit
2397 at a slight CPU cost.
2399 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2400 as requested by Marc Sherman.
2402 SC/10 Eximstats: added histograms for user specified patterns as requested
2405 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2407 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2408 fopen() in the content-scanning modules that did not already have it.
2414 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2415 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2417 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2419 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2421 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2422 to test Sieve filters that use "vacation".
2424 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2425 that changes the way the GnuTLS parameters are stored in the cache file.
2426 The new format can be generated externally. For backward compatibility,
2427 if the data in the cache doesn't make sense, Exim assumes it has read an
2428 old-format file, and it generates new data and writes a new file. This
2429 means that you can't go back to an older release without removing the
2432 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2433 work if there are any delivery delays because "one_time" forces the
2434 parent to be marked "delivered", so its unseen clone is never tried
2435 again. For this reason, Exim now forbids the simultaneous setting of
2438 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2439 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2440 entry if you want to know the details. The fix, however, neglected to
2441 consider the case when local delivery batching is involved. The test for
2442 "previously delivered" was not happening when checking to see if an
2443 address could be batched with a previous (undelivered) one; under
2444 certain circumstances this could lead to multiple deliveries to the same
2447 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2448 in its include files, and this causes problems building Exim.
2450 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2451 header_syntax) but Exim was just ignoring anything given after a slash.
2452 In particular, this caused confusion with an attempt to use "verify =
2453 reverse_host_lookup/defer_ok". An error is now given when options are
2454 supplied for verify items that do not have them. (Maybe reverse_host_
2455 lookup should have a defer_ok option, but that's a different point.)
2457 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2458 defined by RFC 821) to 2048, because there were problems with some AUTH
2459 commands, and RFC 1869 says the size should be increased for extended
2460 SMTP commands that take arguments.
2462 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2465 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2466 "unknown" error; now it says that the functionality isn't in the binary.
2468 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2469 an address' error message when a string expansion fails (syntax or
2470 whatever). Otherwise the password may appear in the log. Following change
2471 PH/42 below, there is no longer a chance of it appearing in a bounce
2474 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2476 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2477 of its addresses was ever tried. (Bugzilla bug #2.)
2479 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2480 the result incorrectly in the debug output. (It correctly added a newline
2481 to what was transported.)
2483 TF/01 Added $received_time.
2485 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2486 commented out examples of how to interface to a virus scanner and to
2487 SpamAssassin. Also added commented examples of av_scanner and
2488 spamd_address settings.
2490 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2491 and controls are allowed in which ACLs. There were a couple of minor
2492 errors. Some of the entries in the conditions table (which is a table of
2493 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2494 negation of where the condition IS allowed.
2496 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2498 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2499 header file does not have a version number, so I've had to invent a new
2500 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2501 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2502 radiusclient), but it was contributed by a Radius user.
2504 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2505 files or directories, for OpenSSL.
2507 PH/19 When an Exim process that is running as root has to create an Exim log
2508 file, it does so in a subprocess that runs as exim:exim so as to get the
2509 ownership right at creation (otherwise, other Exim processes might see
2510 the file with the wrong ownership). There was no test for failure of this
2511 fork() call, which would lead to the process getting stuck as it waited
2512 for a non-existent subprocess. Forks do occasionally fail when resources
2513 run out. I reviewed all the other calls to fork(); they all seem to check
2516 PH/20 When checking for unexpected SMTP input at connect time (before writing
2517 the banner), Exim was not dealing correctly with a non-positive return
2518 from the read() function. If the client had disconnected by this time,
2519 the result was a log entry for a synchronization error with an empty
2520 string after "input=" when read() returned zero. If read() returned -1
2521 (an event I could not check), uninitialized data bytes were printed.
2522 There were reports of junk text (parts of files, etc) appearing after
2525 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2527 PH/22 Added support for macro redefinition, and (re)definition in between
2528 driver and ACL definitions.
2530 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2531 forgetting to use the resulting value; it was using the unexpanded value.
2533 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2534 hadn't been configured. The fix is from Juergen Kreileder, who
2535 understands it better than I do:
2537 "Here's what I see happening with three configured cyrus_sasl
2538 authenticators configured (plain, login, cram-md5):
2540 On startup auth_cyrus_sasl_init() gets called for each of these.
2541 This means three calls to sasl_listmech() without a specified mech_list.
2542 => SASL tests which mechs of all available mechs actually work
2543 => three warnings about OTP not working
2544 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2546 With the patch, sasl_listmech() also gets called three times. But now
2547 SASL's mech_list option is set to the server_mech specified in the the
2548 authenticator. Or in other words, the answer from sasl_listmech()
2549 gets limited to just the mech you're testing for (which is different
2551 => the return list contains just 'plain' or 'login', 'cram-md5' or
2552 nothing depending on the value of ob->server_mech.
2554 I've just tested the patch: Authentication still works fine,
2555 unavailable mechs specified in the exim configuration are still
2556 caught, and the auth.log warnings about OTP are gone."
2558 PH/25 When debugging is enabled, the contents of the command line are added
2559 to the debugging output, even when log_selector=+arguments is not
2562 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2563 answer is "GNU", and only if the return is "GNU/something" is the answer
2566 PH/27 $acl_verify_message is now set immediately after the failure of a
2567 verification in an ACL, and so is available in subsequent modifiers. In
2568 particular, the message can be preserved by coding like this:
2570 warn !verify = sender
2571 set acl_m0 = $acl_verify_message
2573 Previously, $acl_verify_message was set only while expanding "message"
2574 and "log_message" when a very denied access.
2576 PH/28 Modified OS/os.c-Linux with
2578 -#ifndef OS_LOAD_AVERAGE
2579 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2581 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2582 nomenclature these days.)
2584 PH/29 Installed patch from the Sieve maintainer that adds the options
2585 sieve_useraddress and sieve_subaddress to the redirect router.
2587 PH/30 In these circumstances:
2588 . Two addresses routed to the same list of hosts;
2589 . First host does not offer TLS;
2590 . First host accepts first address;
2591 . First host gives temporary error to second address;
2592 . Second host offers TLS and a TLS session is established;
2593 . Second host accepts second address.
2594 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2595 and peerdn, if requested) that were in fact used only for the second
2598 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2599 attention to any local part prefix or suffix that was matched by the
2600 router that accepted the address. It now behaves in the same way as it
2601 does for delivery: the affixes are removed from the local part unless
2602 rcpt_include_affixes is set on the transport.
2604 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2605 timeout during the DATA phase of an incoming message.
2607 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2608 applied a patch sanctioned by the Sieve maintainer.
2610 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2611 the uid or gid is negative. A case of a negative gid caused this to be
2612 noticed. The fix allows for either to be negative.
2614 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2615 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2618 PH/36 The change PH/12 above was broken. Fixed it.
2620 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2621 the grounds that routing the same address twice would always produce the
2622 same answer. This might have been true once, but it is certainly no
2623 longer true now. Routing a child address may depend on the previous
2624 routing that produced that child. Some complicated redirection strategies
2625 went wrong when messages had multiple recipients, and made Exim's
2626 behaviour dependent on the order in which the addresses were given.
2628 I have moved the duplicate checking until after the routing is complete.
2629 Exim scans the addresses that are assigned to local and remote
2630 transports, and removes any duplicates. This means that more work will be
2631 done, as duplicates will always all be routed, but duplicates are
2632 presumably rare, so I don't expect this is of any significance.
2634 For deliveries to pipes, files, and autoreplies, the duplicate checking
2635 still happens during the routing process, since they are not going to be
2638 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2639 It corrects a timeout issue with spamd. This is Ian's comment: "The
2640 background is that sometimes spamd either never reads data from a
2641 connection it has accepted, or it never writes response data. The exiscan
2642 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2643 blindly assumes that writes won't block so it may never time out."
2645 PH/39 Allow G after quota size as well as K and M.
2647 PH/40 The value set for $authenticated_id in an authenticator may not contain
2648 binary zeroes or newlines because the value is written to log lines and
2649 to spool files. There was no check on this. Now the value is run through
2650 the string_printing() function so that such characters are converted to
2651 printable escape sequences.
2653 PH/41 $message_linecount is a new variable that contains the total number of
2654 lines in the message. Compare $body_linecount, which is the count for the
2657 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2658 bounce and delay warning messages, except in certain special cases, which
2661 (a) An SMTP error message from a remote host;
2662 (b) A message specified in a :fail: redirection;
2663 (c) A message specified in a "fail" command in a system filter;
2664 (d) A message specified in a FAIL return from the queryprogram router;
2665 (e) A message specified by the cannot_route_message router option.
2667 In these cases only, Exim does include the error details in bounce and
2668 warning messages. There are also a few cases where bland messages such
2669 as "unrouteable address" or "local delivery error" are given.
2671 PH/43 $value is now also set for the "else" part of a ${run expansion.
2673 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2674 being worked on, but at least Exim now implements the latest version to
2677 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2678 process to complete was treated as a delivery failure, a timeout while
2679 writing the message to the pipe was logged, but erroneously treated as a
2680 successful delivery. Such timeouts include transport filter timeouts. For
2681 consistency with the overall process timeout, these timeouts are now
2682 treated as errors, giving rise to delivery failures by default. However,
2683 there is now a new Boolean option for the pipe transport called
2684 timeout_defer, which, if set TRUE, converts the failures into defers for
2685 both kinds of timeout. A transport filter timeout is now identified in
2688 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2689 systems where "make" and "gmake" are different, calling "gmake" at top
2690 level broke things. I've arranged for the value of $(MAKE) to be passed
2691 from the Makefile to this script so that it can call the same version of
2695 A note about Exim versions 4.44 and 4.50
2696 ----------------------------------------
2698 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2699 changes of various kinds. As a consequence, a big documentation update was
2700 needed. This delayed the release for rather longer than seemed good, especially
2701 in the light of a couple of (minor) security issues. Therefore, the changes
2702 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2703 release. So 4.44 and 4.50 are in effect two different branches that both start
2706 I have left the 4.50 change log unchanged; it contains all the changes since
2707 4.43. The change log for 4.44 is below; many of its items are identical to
2708 those for 4.50. This seems to be the most sensible way to preserve the
2709 historical information.
2715 1. Minor wording change to the doc/README.SIEVE file.
2717 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2718 computation of the current number of files was incorrect.
2720 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2721 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2724 4. Give more explanation in the error message when the command for a transport
2725 filter fails to execute.
2727 5. There are several places where Exim runs a non-Exim command in a
2728 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2729 was being done only for the command run by the queryprogram router. It is
2730 now done for all such subprocesses. The other cases are: ${run, transport
2731 filters, and the commands run by the lmtp and pipe transports.
2733 6. Added CONFIGURE_GROUP build-time option.
2735 7. Some older OS have a limit of 256 on the maximum number of file
2736 descriptors. Exim was using setrlimit() to set 1000 as a large value
2737 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2738 systems. I've change it so that if it can't get 1000, it tries for 256.
2740 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2741 was an oversight, and furthermore, ever since the addition of extra
2742 controls (e.g. 4.43/32), the checks on when to allow different forms of
2743 "control" were broken. There should now be diagnostics for all cases when a
2744 control that does not make sense is encountered.
2746 9. Added the /retain_sender option to "control=submission".
2748 10. $recipients is now available in the predata ACL (oversight).
2750 11. Tidy the search cache before the fork to do a delivery from a message
2751 received from the command line. Otherwise the child will trigger a lookup
2752 failure and thereby defer the delivery if it tries to use (for example) a
2753 cached ldap connection that the parent has called unbind on.
2755 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2756 of $address_data from the recipient verification was clobbered by the
2757 sender verification.
2759 13. The value of address_data from a sender verification is now available in
2760 $sender_address_data in subsequent conditions in the ACL statement.
2762 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2764 15. Added a new option "connect=<time>" to callout options, to set a different
2767 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2768 was its contents. (It was OK if the option was not defined at all.)
2770 17. A "Completed" log line is now written for messages that are removed from
2771 the spool by the -Mrm option.
2773 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2774 information about exactly what failed.
2776 19. Added -dd to debug only the daemon process.
2778 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2779 handles timeouts, both on the server side and network timeouts. Renamed the
2780 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2782 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2783 It is now set to "smtps".
2785 22. $host_address is now set to the target address during the checking of
2786 ignore_target_hosts.
2788 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2789 being passed; this would have caused $sender_host_name to have been used if
2790 matching the list had actually called for a host name (not very likely,
2791 since this list is usually IP addresses). A host name is now passed as
2794 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2795 code that specifies a non-restarting handler (typically sigaction() in
2796 modern systems) in an attempt to fix a rare and obscure crash bug.
2798 25. Narrowed the window for a race in the daemon that could cause it to ignore
2799 SIGCHLD signals. This is not a major problem, because they are used only to
2800 wake it up if nothing else does.
2802 26. A malformed maildirsize file could cause Exim to calculate negative values
2803 for the mailbox size or file count. Odd effects could occur as a result.
2804 The maildirsize information is now recalculated if the size or filecount
2807 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2808 support for a long time. Removed HAVE_SYS_VFS_H.
2810 28. Installed the latest version of exipick from John Jetmore.
2812 29. In an address list, if the pattern was not a regular expression, an empty
2813 subject address (from a bounce message) matched only if the pattern was an
2814 empty string. Non-empty patterns were not even tested. This was the wrong
2815 because it is perfectly reasonable to use an empty address as part of a
2816 database query. An empty address is now tested by patterns that are
2817 lookups. However, all the other forms of pattern expect the subject to
2818 contain a local part and a domain, and therefore, for them, an empty
2819 address still always fails if the pattern is not itself empty.
2821 30. Exim went into a mad DNS loop when attempting to do a callout where the
2822 host was specified on an smtp transport, and looking it up yielded more
2823 than one IP address.
2825 31. Re-factored the code for checking spool and log partition space into a
2826 function that finds that data and another that does the check. The former
2827 is then used to implement four new variables: $spool_space, $log_space,
2828 $spool_inodes, and $log_inodes.
2830 32. The RFC2047 encoding function was originally intended for short strings
2831 such as real names; it was not keeping to the 75-character limit for
2832 encoded words that the RFC imposes. It now respects the limit, and
2833 generates multiple encoded words if necessary. To be on the safe side, I
2834 have increased the buffer size for the ${rfc2047: expansion operator from
2837 33. It is now permitted to omit both strings after an "if" condition; if the
2838 condition is true, the result is "true". As before, when the second string
2839 is omitted, a false condition yields an empty string. This makes it less
2840 cumbersome to write custom ACL and router conditions.
2842 34. Failure to deliver a bounce message always caused it to be frozen, even if
2843 there was an errors_to setting on the router. The errors_to setting is now
2846 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2847 canonical form (fully expanded) before being placed in
2848 $sender_host_address.
2850 36. The table in the code that translates DNS record types into text (T_A to
2851 "A" for instance) was missing entries for NS and CNAME. It is just possible
2852 that this could have caused confusion if both these types were looked up
2853 for the same domain, because the text type is used as part of Exim's
2854 per-process caching. But the chance of anyone hitting this buglet seems
2857 37. The dnsdb lookup has been extended in a number of ways.
2859 (1) There is a new type, "zns", which walks up the domain tree until it
2860 finds some nameserver records. It should be used with care.
2862 (2) There is a new type, "mxh", which is like "mx" except that it returns
2863 just the host names, not the priorities.
2865 (3) It is now possible to give a list of domains (or IP addresses) to be
2866 looked up. The behaviour when one of the lookups defers can be
2867 controlled by a keyword.
2869 (4) It is now possible to specify the separator character for use when
2870 multiple records are returned.
2872 38. The dnslists ACL condition has been extended: it is now possible to supply
2873 a list of IP addresses and/or domains to be looked up in a particular DNS
2876 39. Added log_selector=+queue_time_overall.
2878 40. When running the queue in the test harness, wait just a tad after forking a
2879 delivery process, to get repeatability of debugging output.
2881 41. Include certificate and key file names in error message when GnuTLS fails
2882 to set them up, because the GnuTLS error message doesn't include the name
2883 of the failing file when there is a problem reading it.
2885 42. Allow both -bf and -bF in the same test run.
2887 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2889 44. The "Exiscan patch" is now merged into the mainline Exim source.
2891 45. Sometimes the final signoff response after QUIT could fail to get
2892 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2893 < 0 before doing a fflush(). This bug looks as though it goes back to the
2894 introduction of TLS in release 3.20, but "sometimes" must have been rare
2895 because the tests only now provoked it.
2897 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2898 (this can affect the format of dates).
2900 47. exim_tidydb, when checking for the continued existence of a message for
2901 which it has found a message-specific retry record, was not finding
2902 messages that were in split spool directories. Consequently, it was
2903 deleting retry records that should have stayed in existence.
2905 48. Steve fixed some bugs in eximstats.
2907 49. The SPA authentication driver was not abandoning authentication and moving
2908 on to the next authenticator when an expansion was forced to fail,
2909 contradicting the general specification for all authenticators. Instead it
2910 was generating a temporary error. It now behaves as specified.
2912 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2913 (the order specifies the preference for clients). The order is now AES256,
2914 AES128, 3DES, ARCFOUR128.
2916 51. Small patch to Sieve code - explicitly set From: when generating an
2919 52. Exim crashed if a remote delivery caused a very long error message to be
2920 recorded - for instance if somebody sent an entire SpamAssassin report back
2921 as a large number of 550 error lines. This bug was coincidentally fixed by
2922 increasing the size of one of Exim's internal buffers (big_buffer) that
2923 happened as part of the Exiscan merge. However, to be on the safe side, I
2924 have made the code more robust (and fixed the comments that describe what
2927 53. Now that there can be additional text after "Completed" in log lines (if
2928 the queue_time_overall log selector is set), a one-byte patch to exigrep
2929 was needed to allow it to recognize "Completed" as not the last thing in
2932 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2933 patch that reportedly fixes this has been added. I am not expert enough to
2934 create a test for it. This is what the patch creator wrote:
2936 "I found a little strange behaviour of ldap code when working with
2937 Windows 2003 AD Domain, where users was placed in more than one
2938 Organization Units. When I tried to give exim partial DN, the exit code
2939 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2940 But simultaneously result of request was absolutely normal ldap result,
2941 so I produce this patch..."
2943 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2944 REFERENCE, so I have modified the code to exclude the patch when that macro
2947 55. Some experimental protocols are using DNS PTR records for new purposes. The
2948 keys for these records are domain names, not reversed IP addresses. The
2949 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2950 leaves it alone. Component reversal etc. now happens only for IP addresses.
2953 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2955 57. Double the size of the debug message buffer (to 2048) so that more of very
2956 long debug lines gets shown.
2958 58. The exicyclog utility now does better if the number of log files to keep
2959 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2961 59. Two changes related to the smtp_active_hostname option:
2963 (1) $smtp_active_hostname is now available as a variable.
2964 (2) The default for smtp_banner uses $smtp_active_hostname instead
2965 of $primary_hostname.
2967 60. The host_aton() function is supposed to be passed a string that is known
2968 to be a valid IP address. However, in the case of IPv6 addresses, it was
2969 not checking this. This is a hostage to fortune. Exim now panics and dies
2970 if the condition is not met. A case was found where this could be provoked
2971 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2972 components; fortuitously, this particular loophole had already been fixed
2973 by change 4.50/55 above.
2975 If there are any other similar loopholes, the new check in host_aton()
2976 itself should stop them being exploited. The report I received stated that
2977 data on the command line could provoke the exploit when Exim was running as
2978 exim, but did not say which command line option was involved. All I could
2979 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2980 running as the user.
2983 61. There was a buffer overflow vulnerability in the SPA authentication code
2984 (which came originally from the Samba project). I have added a test to the
2985 spa_base64_to_bits() function which I hope fixes it.
2988 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2989 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2991 63. The daemon start-up calls getloadavg() while still root for those OS that
2992 need the first call to be done as root, but it missed one case: when
2993 deliver_queue_load_max is set with deliver_drop_privilege. This is
2994 necessary for the benefit of the queue runner, because there is no re-exec
2995 when deliver_drop_privilege is set.
2997 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2998 This has been fixed.
3000 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3001 was in use, was not putting the data itself into the right store pool;
3002 consequently, it could be overwritten for a subsequent message in the same
3003 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3006 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3008 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3009 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3010 the function were treating the return as a boolean value, which happened to
3011 work because 0=false and not-0=true, but is not correct code.
3013 68. The host_aton() function was not handling scoped IPv6 addresses (those
3014 with, for example, "%eth0" on the end) correctly.
3016 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3017 negated items (that is, ~something) in unsigned ints. Some compilers
3018 apparently mutter when there is no cast.
3020 70. If an address verification called from an ACL failed, and did not produce a
3021 user-specific message (i.e. there was only a "system" message), nothing was
3022 put in $acl_verify_message. In this situation, it now puts the system
3025 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3026 session; change 4.31/43 added the unwanted input to the log line - except
3027 that it did not do this in the start of session case. It now does.
3029 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3030 This is wrong and can cause the other end to generate a synchronization
3031 error if it is another Exim or anything else that does the synchronization
3032 check. A QUIT command is no longer sent after a timeout.
3034 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3035 during host lookups.
3037 74. The defer_ok option of callout verification was not working if it was used
3038 when verifying addresses in header lines, that is, for this case:
3040 verify = header_sender/callout=defer_ok
3042 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3043 those file descriptors could be used for SMTP connections. If anything
3044 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3045 could be sent to the SMTP client, causing chaos. The daemon now opens
3046 stdin, stdout, and stderr to /dev/null when it puts itself into the
3049 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3050 log by default. The user can override this with suitable Perl magic.
3052 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3053 the log message when discard triggers, was not working for the DATA ACL or
3054 for the non-SMTP ACL.
3056 78. Error message wording change in sieve.c.
3058 79. If smtp_accept_max_per_host was set, the number of connections could be
3059 restricted to fewer than expected, because the daemon was trying to set up
3060 a new connection before checking whether the processes handling previous
3061 connections had finished. The check for completed processes is now done
3062 earlier. On busy systems, this bug wouldn't be noticed because something
3063 else would have woken the daemon, and it would have reaped the completed
3066 80. If a message was submitted locally by a user whose login name contained one
3067 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3068 It caused a spool format error. I have fixed the spool reading code. A
3069 related problem was that the "from" clause in the Received: line became
3070 illegal because of the space(s). It is now covered by ${quote_local_part.
3072 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3075 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3077 83. Patch from Sieve maintainer to fix unterminated string problem in
3078 "vacation" handling.
3080 84. Some minor changes to the Linux configuration files to help with other
3081 OS variants using glibc.
3083 85. One more patch for Sieve to update vacation handling to latest spec.
3086 ----------------------------------------------------
3087 See the note above about the 4.44 and 4.50 releases.
3088 ----------------------------------------------------
3094 1. Change 4.43/35 introduced a bug that caused file counts to be
3095 incorrectly computed when quota_filecount was set in an appendfile
3098 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3099 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3102 3. Give more explanation in the error message when the command for a transport
3103 filter fails to execute.
3105 4. There are several places where Exim runs a non-Exim command in a
3106 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3107 was being done only for the command run by the queryprogram router. It is
3108 now done for all such subprocesses. The other cases are: ${run, transport
3109 filters, and the commands run by the lmtp and pipe transports.
3111 5. Some older OS have a limit of 256 on the maximum number of file
3112 descriptors. Exim was using setrlimit() to set 1000 as a large value
3113 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3114 systems. I've change it so that if it can't get 1000, it tries for 256.
3116 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3117 was an oversight, and furthermore, ever since the addition of extra
3118 controls (e.g. 4.43/32), the checks on when to allow different forms of
3119 "control" were broken. There should now be diagnostics for all cases when a
3120 control that does not make sense is encountered.
3122 7. $recipients is now available in the predata ACL (oversight).
3124 8. Tidy the search cache before the fork to do a delivery from a message
3125 received from the command line. Otherwise the child will trigger a lookup
3126 failure and thereby defer the delivery if it tries to use (for example) a
3127 cached ldap connection that the parent has called unbind on.
3129 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3130 of $address_data from the recipient verification was clobbered by the
3131 sender verification.
3133 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3134 was its contents. (It was OK if the option was not defined at all.)
3136 11. A "Completed" log line is now written for messages that are removed from
3137 the spool by the -Mrm option.
3139 12. $host_address is now set to the target address during the checking of
3140 ignore_target_hosts.
3142 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3143 being passed; this would have caused $sender_host_name to have been used if
3144 matching the list had actually called for a host name (not very likely,
3145 since this list is usually IP addresses). A host name is now passed as
3148 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3149 code that specifies a non-restarting handler (typically sigaction() in
3150 modern systems) in an attempt to fix a rare and obscure crash bug.
3152 15. Narrowed the window for a race in the daemon that could cause it to ignore
3153 SIGCHLD signals. This is not a major problem, because they are used only to
3154 wake it up if nothing else does.
3156 16. A malformed maildirsize file could cause Exim to calculate negative values
3157 for the mailbox size or file count. Odd effects could occur as a result.
3158 The maildirsize information is now recalculated if the size or filecount
3161 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3162 support for a long time. Removed HAVE_SYS_VFS_H.
3164 18. Updated exipick to current release from John Jetmore.
3166 19. Allow an empty sender to be matched against a lookup in an address list.
3167 Previously the only cases considered were a regular expression, or an
3170 20. Exim went into a mad DNS lookup loop when doing a callout where the
3171 host was specified on the transport, if the DNS lookup yielded more than
3174 21. The RFC2047 encoding function was originally intended for short strings
3175 such as real names; it was not keeping to the 75-character limit for
3176 encoded words that the RFC imposes. It now respects the limit, and
3177 generates multiple encoded words if necessary. To be on the safe side, I
3178 have increased the buffer size for the ${rfc2047: expansion operator from
3181 22. Failure to deliver a bounce message always caused it to be frozen, even if
3182 there was an errors_to setting on the router. The errors_to setting is now
3185 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3186 canonical form (fully expanded) before being placed in
3187 $sender_host_address.
3189 24. Updated eximstats to version 1.33
3191 25. Include certificate and key file names in error message when GnuTLS fails
3192 to set them up, because the GnuTLS error message doesn't include the name
3193 of the failing file when there is a problem reading it.
3195 26. Expand error message when OpenSSL has problems setting up cert/key files.
3198 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3199 (this can affect the format of dates).
3201 28. exim_tidydb, when checking for the continued existence of a message for
3202 which it has found a message-specific retry record, was not finding
3203 messages that were in split spool directories. Consequently, it was
3204 deleting retry records that should have stayed in existence.
3206 29. eximstats updated to version 1.35
3207 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3208 1.35 - bugfix such that pie charts by volume are generated correctly
3210 30. The SPA authentication driver was not abandoning authentication and moving
3211 on to the next authenticator when an expansion was forced to fail,
3212 contradicting the general specification for all authenticators. Instead it
3213 was generating a temporary error. It now behaves as specified.
3215 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3216 (the order specifies the preference for clients). The order is now AES256,
3217 AES128, 3DES, ARCFOUR128.
3219 31. Small patch to Sieve code - explicitly set From: when generating an
3222 32. Exim crashed if a remote delivery caused a very long error message to be
3223 recorded - for instance if somebody sent an entire SpamAssassin report back
3224 as a large number of 550 error lines. This bug was coincidentally fixed by
3225 increasing the size of one of Exim's internal buffers (big_buffer) that
3226 happened as part of the Exiscan merge. However, to be on the safe side, I
3227 have made the code more robust (and fixed the comments that describe what
3230 33. Some experimental protocols are using DNS PTR records for new purposes. The
3231 keys for these records are domain names, not reversed IP addresses. The
3232 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3233 leaves it alone. Component reversal etc. now happens only for IP addresses.
3236 34. The host_aton() function is supposed to be passed a string that is known
3237 to be a valid IP address. However, in the case of IPv6 addresses, it was
3238 not checking this. This is a hostage to fortune. Exim now panics and dies
3239 if the condition is not met. A case was found where this could be provoked
3240 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3241 components; fortuitously, this particular loophole had already been fixed
3242 by change 4.50/55 or 4.44/33 above.
3244 If there are any other similar loopholes, the new check in host_aton()
3245 itself should stop them being exploited. The report I received stated that
3246 data on the command line could provoke the exploit when Exim was running as
3247 exim, but did not say which command line option was involved. All I could
3248 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3249 running as the user.
3252 35. There was a buffer overflow vulnerability in the SPA authentication code
3253 (which came originally from the Samba project). I have added a test to the
3254 spa_base64_to_bits() function which I hope fixes it.
3257 36. The daemon start-up calls getloadavg() while still root for those OS that
3258 need the first call to be done as root, but it missed one case: when
3259 deliver_queue_load_max is set with deliver_drop_privilege. This is
3260 necessary for the benefit of the queue runner, because there is no re-exec
3261 when deliver_drop_privilege is set.
3263 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3264 was in use, was not putting the data itself into the right store pool;
3265 consequently, it could be overwritten for a subsequent message in the same
3266 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3269 38. Sometimes the final signoff response after QUIT could fail to get
3270 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3271 < 0 before doing a fflush(). This bug looks as though it goes back to the
3272 introduction of TLS in release 3.20, but "sometimes" must have been rare
3273 because the tests only now provoked it.
3279 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3280 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3281 Now it returns an integer. A number of calls were still expecting a T/F
3282 return. Fortuitously, in all cases, the tests worked in OK situations,
3283 which is the norm. However, things would have gone wrong on any write
3284 failures on the smtp file descriptor. This function is used when sending
3285 messages over SMTP and also when doing verify callouts.
3287 2. When Exim is called to do synchronous delivery of a locally submitted
3288 message (the -odf or -odi options), it no longer closes stderr before doing
3291 3. Implemented the mua_wrapper option.
3293 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3295 5. Implemented the functions header_remove(), header_testname(),
3296 header_add_at_position(), and receive_remove_recipient(), and exported them
3299 6. If an ACL "warn" statement specified the addition of headers, Exim already
3300 inserted X-ACL-Warn: at the start if there was no header name. However, it
3301 was not making this test for the second and subsequent header lines if
3302 there were newlines in the string. This meant that an invalid header could
3303 be inserted if Exim was badly configured.
3305 7. Allow an ACL "warn" statement to add header lines at the start or after all
3306 the Received: headers, as well as at the end.
3308 8. Added the rcpt_4xx retry error code.
3310 9. Added postmaster_mailfrom=xxx to callout verification option.
3312 10. Added mailfrom=xxxx to the callout verification option, for verify=
3315 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3316 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3318 12. Inserted some casts to stop certain compilers warning when using pointer
3319 differences as field lengths or precisions in printf-type calls (mostly
3320 affecting debugging statements).
3322 13. Added optional readline() support for -be (dynamically loaded).
3324 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3325 same clock tick as a message's arrival, so that its received time was the
3326 same as the "first fail" time on the retry record, and that message
3327 remained on the queue past the ultimate address timeout, every queue runner
3328 would try a delivery (because it was past the ultimate address timeout) but
3329 after another failure, the ultimate address timeout, which should have then
3330 bounced the address, did not kick in. This was a "< instead of <=" error;
3331 in most cases the first failure would have been in the next clock tick
3332 after the received time, and all would be well.
3334 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3335 being recognized when the domain list was tested by the match_domain
3336 condition in an expansion string.
3338 16. Added the ${str2b64: operator.
3340 17. Exim was always calling setrlimit() to set a large limit for the number of
3341 processes, without checking whether the existing limit was already
3342 adequate. (It did check for the limit on file descriptors.) Furthermore,
3343 errors from getrlimit() and setrlimit() were being ignored. Now they are
3344 logged to the main and panic logs, but Exim does carry on, to try to do its
3345 job under whatever limits there are.
3347 18. Imported PCRE 5.0.
3349 19. Trivial typo in log message " temporarily refused connection" (the leading
3352 20. If the log selector return_path_on_delivery was set and an address was
3353 redirected to /dev/null, the delivery process crashed because it assumed
3354 that a return path would always be set for a "successful" delivery. In this
3355 case, the whole delivery is bypassed as an optimization, and therefore no
3358 21. Internal re-arrangement: the function for sending a challenge and reading
3359 a response while authentication was assuming a zero-terminated challenge
3360 string. It's now changed to take a pointer and a length, to allow for
3361 binary data in such strings.
3363 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3365 23. Exim was not respecting finduser_retries when seeking the login of the
3366 uid under which it was called; it was always trying 10 times. (The default
3367 setting of finduser_retries is zero.) Also, it was sleeping after the final
3368 failure, which is pointless.
3370 24. Implemented tls_on_connect_ports.
3372 25. Implemented acl_smtp_predata.
3374 26. If the domain in control=submission is set empty, Exim assumes that the
3375 authenticated id is a complete email address when it generates From: or
3376 Sender: header lines.
3378 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3379 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3380 chown and chgrp in /bin and hostname in /usr/bin.
3382 28. Exim was keeping the "process log" file open after each use, just as it
3383 does for the main log. This opens the possibility of it remaining open for
3384 long periods when the USR1 signal hits a daemon. Occasional processlog
3385 errors were reported, that could have been caused by this. Anyway, it seems
3386 much more sensible not to leave this file open at all, so that is what now
3389 29. The long-running daemon process does not normally write to the log once it
3390 has entered its main loop, and it closes the log before doing so. This is
3391 so that log files can straightforwardly be renamed and moved. However,
3392 there are a couple of unusual error situations where the daemon does write
3393 log entries, and I had neglected to close the log afterwards.
3395 30. The text of an SMTP error response that was received during a remote
3396 delivery was being truncated at 512 bytes. This is too short for some of
3397 the long messages that one sometimes sees. I've increased the limit to
3400 31. It is now possible to make retry rules that apply only when a message has a
3401 specific sender, in particular, an empty sender.
3403 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3404 it possible to be selective about when SMTP synchronization is enforced.
3406 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3408 32. Implemented hosts_connection_nolog.
3410 33. Added an ACL for QUIT.
3412 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3415 35. Added mailbox_size and mailbox_filecount to appendfile.
3417 36. Added control = no_multiline_responses to ACLs.
3419 37. There was a bug in the logic of the code that waits for the clock to tick
3420 in the case where the clock went backwards by a substantial amount such
3421 that the microsecond fraction of "now" was more than the microsecond
3422 fraction of "then" (but the whole seconds number was less).
3424 38. Added support for the libradius Radius client library this is found on
3425 FreeBSD (previously only the radiusclient library was supported).
3431 1. When certain lookups returned multiple values in the form name=value, the
3432 quoting of the values was not always being done properly. Specifically:
3433 (a) If the value started with a double quote, but contained no whitespace,
3435 (b) If the value contained whitespace other than a space character (i.e.
3436 tabs or newlines or carriage returns) it was not quoted.
3437 This fix has been applied to the mysql and pgsql lookups by writing a
3438 separate quoting function and calling it from the lookup code. The fix
3439 should probably also be applied to nisplus, ibase and oracle lookups, but
3440 since I cannot test any of those, I have not disturbed their existing code.
3442 2. A hit in the callout cache for a specific address caused a log line with no
3443 reason for rejecting RCPT. Now it says "Previous (cached) callout
3444 verification failure".
3446 3. There was an off-by-one bug in the queryprogram router. An over-long
3447 return line was truncated at 256 instead of 255 characters, thereby
3448 overflowing its buffer with the terminating zero. As well as fixing this, I
3449 have increased the buffer size to 1024 (and made a note to document this).
3451 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3452 when Exim is waiting for an SMTP response from a remote server, Exim
3453 restarts its select() call on the socket, thereby resetting its timeout.
3454 This is not a problem when such interrupts are rare. Somebody set up a cron
3455 job to run exiwhat every 2 minutes, which is less than the normal select()
3456 timeout (5 or 10 minutes). This meant that the select() timeout never
3457 kicked in because it was always reset. I have fixed this by comparing the
3458 time when an interrupt arrives with the time at the start of the first call
3459 to select(). If more time than the timeout has elapsed, the interrupt is
3460 treated as a timeout.
3462 5. Some internal re-factoring in preparation for the addition of Sieve
3463 extensions (by MH). In particular, the "personal" test is moved to a
3464 separate function, and given an option for scanning Cc: and Bcc: (which is
3465 not set for Exim filters).
3467 6. When Exim created an email address using the login of the caller as the
3468 local part (e.g. when creating a From: or Sender: header line), it was not
3469 quoting the local part when it contained special characters such as @.
3471 7. Installed new OpenBSD configuration files.
3473 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3474 try to make them clearer.
3476 9. Callout options, other than the timeout value, were being ignored when
3477 verifying sender addresses in header lines. For example, when using
3479 verify = header_sender/callout=no_cache
3481 the cache was (incorrectly) being used.
3483 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3484 only the Cygwin environment.
3486 11. When return_path_on_delivery was set as a log selector, if different remote
3487 addresses in the same message used different return paths and parallel
3488 remote delivery occurred, the wrong values would sometimes be logged.
3489 (Whenever a remote delivery process finished, the return path value from
3490 the most recently started remote delivery process was logged.)
3492 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3493 lines when AUTH and/or TLS are in use. This is the "received protocol"
3494 field. Exim used to use "asmtp" for authenticated SMTP, without any
3495 indication (in the protocol name) for TLS use. Now it follows the RFC and
3496 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3497 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3498 names appear in log lines as well as in Received: header lines.
3500 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3501 extensions, and comparison tests, and to fix some bugs.
3503 14. Changes to the "personal" filter test:
3505 (1) The test was buggy in that it was just doing the equivalent of
3506 "contains" tests on header lines. For example, if a user's address was
3507 anne@some.where, the "personal" test would incorrectly be true for
3509 To: susanne@some.where
3511 This test is now done by extracting each address from the header in turn,
3512 and checking the entire address. Other tests that are part of "personal"
3513 are now done using regular expressions (for example, to check local parts
3514 of addresses in From: header lines).
3516 (2) The list of non-personal local parts in From: addresses has been
3517 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3518 taken from the Sieve specification recommendations.
3520 (3) If the message contains any header line starting with "List-" it is
3521 treated as non-personal.
3523 (4) The test for "circular" in the Subject: header line has been removed
3524 because it now seems ill-conceived.
3526 15. Minor typos in src/EDITME comments corrected.
3528 16. Installed latest exipick from John Jetmore.
3530 17. If headers_add on a router specified a text string that was too long for
3531 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3532 of string_sprintf() is now avoided.
3534 18. $message_body_size was not set (it was always zero) when running the DATA
3535 ACL and the local_scan() function.
3537 19. For the "mail" command in an Exim filter, no default was being set for
3538 the once_repeat time, causing a random time value to be used if "once" was
3539 specified. (If the value happened to be <= 0, no repeat happened.) The
3540 default is now 0s, meaning "never repeat". The "vacation" command was OK
3541 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3542 (I found it when inspecting the code).
3544 20. There is now an overall timeout for performing a callout verification. It
3545 defaults to 4 times the callout timeout, which applies to individual SMTP
3546 commands during the callout. The overall timeout applies when there is more
3547 than one host that can be tried. The timeout is checked before trying the
3548 next host. This prevents very long delays if there are a large number of
3549 hosts and all are timing out (e.g. when the network connections are timing
3550 out). The value of the overall timeout can be changed by specifying an
3551 additional sub-option for "callout", called "maxwait". For example:
3553 verify = sender/callout=5s,maxwait=20s
3555 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3556 to the end before writing, but this should make it even safer).
3558 22. Exim was forgetting that it had advertised PIPELINING for the second and
3559 subsequent messages on an SMTP connection. It was also not resetting its
3560 memory on STARTTLS and an internal HELO.
3562 23. When Exim logs an SMTP synchronization error within a session, it now
3563 records whether PIPELINING has been advertised or not.
3565 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3566 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3567 rather than long int.
3569 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3571 26. Added the never_mail option to autoreply.
3577 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3578 crash if the getsockname() call failed; this can happen if a connection is
3579 closed very soon after it is established. The problem was simply in the
3580 order in which certain operations were done, causing Exim to try to write
3581 to the SMTP stream before it had set up the file descriptor. The bug has
3582 been fixed by making things happen in the correct order.
3588 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3589 before the connection was closed, thus losing the rejection response.
3591 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3592 some early Solaris releases, but causes trouble in current releases where
3593 socklen_t is defined.
3595 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3598 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3601 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3602 that releasing the top of it at the end releases what was used for sub-
3603 expansions (unless the block got too big). However, discard this block if
3604 the first thing is a variable or header, so that we can use its block when
3605 it is dynamic (useful for very large $message_headers, for example).
3607 6. Lookups now cache *every* query, not just the most recent. A new, separate
3608 store pool is used for this. It can be recovered when all lookup caches are
3609 flushed. Lookups now release memory at the end of their result strings.
3610 This has involved some general refactoring of the lookup sources.
3612 7. Some code has been added to the store_xxx() functions to reduce the amount
3613 of flapping under certain conditions.
3615 8. log_incoming_interface used to affect only the <= reception log lines. Now
3616 it causes the local interface and port to be added to several more SMTP log
3617 lines, for example "SMTP connection from", and rejection lines.
3619 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3621 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3623 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3624 could be overwritten at the end of the current message (or the start of a
3625 new message if it was set in a HELO ACL). The value is now preserved for
3626 the duration of the SMTP connection.
3628 12. If a transport had a headers_rewrite setting, and a matching header line
3629 contained an unqualified address, that address was qualified, even if it
3630 did not match any rewriting rules. The underlying bug was that the values
3631 of the flags that permit the existence of unqualified sender and recipient
3632 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3633 non-local messages, and by -bnq for local messages) were not being
3634 preserved with the message after it was received.
3636 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3637 "next input=" as part of the text comprising the host identity instead of
3638 the correct text. The code was using the same buffer for two different
3639 strings. However, depending on which order the printing function evaluated
3640 its arguments, the bug did not always show up. Under Linux, for example, my
3641 test suite worked just fine.
3643 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3644 doesn't work with some older versions of Perl. It has been changed to "my",
3645 which in any case is probably the better facility to use.
3647 15. A really picky compiler found some instances of statements for creating
3648 error messages that either had too many or two few arguments for the format
3651 16. The size of the buffer for calls to the DNS resolver has been increased
3652 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3653 for addresses that have a lot of PTR records. This alleviates a problem; it
3654 does not fully solve it.
3656 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3657 buffer now truncates the list and logs the incident, which is the same
3658 action as happens when Exim is looking up a host name and its aliases.
3659 Previously in this situation something unpredictable would happen;
3660 sometimes it was "internal error: store_reset failed".
3662 18. If a server dropped the connection unexpectedly when an Exim client was
3663 using GnuTLS and trying to read a response, the client delivery process
3664 crashed while trying to generate an error log message.
3666 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3667 string, for example:
3669 warn message = H1: something\nH2: something
3671 the text was added as a single header line from Exim's point of view
3672 though it ended up OK in the delivered message. However, searching for the
3673 second and subsequent header lines using $h_h2: did not work. This has been
3674 fixed. Similarly, if a system filter added multiple headers in this way,
3675 the routers could not see them.
3677 20. Expanded the error message when iplsearch is called with an invalid key to
3678 suggest using net-iplsearch in a host list.
3680 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3681 ACLs are no longer actually imposed (and a message to that effect is
3684 22. If a "gecos" field in a passwd entry contained escaped characters, in
3685 particular, if it contained a \" sequence, Exim got it wrong when building
3686 a From: or a Sender: header from that name. A second bug also caused
3687 incorrect handling when an unquoted " was present following a character
3688 that needed quoting.
3690 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3691 was not being matched caselessly.
3693 24. Arranged for all hyphens in the exim.8 source to be escaped with
3696 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3697 properly. Recipient callout cache records were still being keyed to include
3698 the sender, even when use_sender was set false. This led to far more
3699 callouts that were necessary. The sender is no longer included in the key
3700 when use_sender is false.
3702 26. Added "control = submission" modifier to ACLs.
3704 27. Added the ${base62d: operator to decode base 62 numbers.
3706 28. dnsdb lookups can now access SRV records.
3708 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3709 the configuration file.
3711 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3712 (-v) mode. This makes the output for a verbose queue run more intelligible.
3714 31. Added a use_postmaster feature to recipient callouts.
3716 32. Added the $body_zerocount variable, containing the number of binary zero
3717 bytes in the message body.
3719 33. The time of last modification of the "new" subdirectory is now used as the
3720 "mailbox time last read" when there is a quota error for a maildir
3723 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3725 35. Added +ignore_unknown as a special item in host lists.
3727 36. Code for decoding IPv6 addresses in host lists is now included, even if
3728 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3729 address was recognized as an IP address, but was then not correctly decoded
3730 into binary, causing unexpected and incorrect effects when compared with
3737 1. Very minor rewording of debugging text in manualroute to say "list of
3738 hosts" instead of "hostlist".
3740 2. If verify=header_syntax was set, and a header line with an unqualified
3741 address (no domain) and a large number of spaces between the end of the
3742 name and the colon was received, the reception process suffered a buffer
3743 overflow, and (when I tested it) crashed. This was caused by some obsolete
3744 code that should have been removed. The fix is to remove it!
3746 3. When running in the test harness, delay a bit after writing a bounce
3747 message to get a bit more predictability in the log output.
3749 4. Added a call to search_tidyup() just before forking a reception process. In
3750 theory, someone could use a lookup in the expansion of smtp_accept_max_
3751 per_host which, without the tidyup, could leave open a database connection.
3753 5. Added the variables $recipient_data and $sender_data which get set from a
3754 lookup success in an ACL "recipients" or "senders" condition, or a router
3755 "senders" option, similar to $domain_data and $local_part_data.
3757 6. Moved the writing of debug_print from before to after the "senders" test
3760 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3761 problems for message scanning, either using a data ACL, or using
3762 local_scan() because the Received: header was not generated till after they
3763 were called (in order to set the time as the time of reception completion).
3764 I have revised the way this works. The header is now generated after the
3765 body is received, but before the ACL or local_scan() are called. After they
3766 are run, the timestamp in the header is updated.
3772 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3773 before starting a queue runner without re-exec. This happened only when
3774 deliver_drop_privilege was set or when the Exim user was set to root. The
3775 effect of the bug was that timeouts during subsequent deliveries caused
3776 crashes instead of being properly handled. The handler is now left at its
3777 default (and expected) setting.
3779 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3780 message, again when deliver_drop_privilege is set or Exim is run as root.
3781 The bug described in (1) was not present in this case, but the tidying up
3782 of the other signals was missing. I have made the two cases consistent.
3784 3. The ignore_target_hosts setting on a manualroute router was being ignored
3785 for hosts that were looked up using the /MX notation.
3787 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3790 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3791 operated on the sender address. After changing the $sender_address to <>
3792 for the sender address verify, Exim was re-instated it as the original
3793 (before rewriting) address, but remembering that it had rewritten it, so it
3794 wasn't rewriting it again. This bug also had the effect of breaking the
3795 sender address verification caching when the sender address was rewritten.
3797 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3798 This has been changed so that if the ip literal address matches
3799 ignore_target_hosts, the router declines.
3801 7. Added expansion conditions match_domain, match_address, and match_local_
3802 part (NOT match_host).
3804 8. The placeholder for the Received: header didn't have a length field set.
3806 9. Added code to Exim itself and to exim_lock to test for a specific race
3807 condition that could lead to file corruption when using MBX delivery. The
3808 issue is with the lockfile that is created in /tmp. If this file is removed
3809 after a process has opened it but before that process has acquired a lock,
3810 there is the potential for a second process to recreate the file and also
3811 acquire a lock. This could lead to two Exim processes writing to the file
3812 at the same time. The added code performs the same test as UW imapd; it
3813 checks after acquiring the lock that its file descriptor still refers to
3814 the same named file.
3816 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3817 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3820 11. Added the smtp_active_hostname option. If used, this will typically be made
3821 to depend on the incoming interface address. Because $interface_address is
3822 not set up until the daemon has forked a reception process, error responses
3823 that can happen earlier (such as "too many connections") no longer contain
3826 12. If an expansion in a condition on a "warn" statement fails because a lookup
3827 defers, the "warn" statement is abandoned, and the next ACL statement is
3828 processed. Previously this caused the whole ACL to be aborted.
3830 13. Added the iplsearch lookup type.
3832 14. Added ident_timeout as a log selector.
3834 15. Added tls_certificate_verified as a log selector.
3836 16. Added a global option tls_require_ciphers (compare the smtp transport
3837 option of the same name). This controls incoming TLS connections.
3839 17. I finally figured out how to make tls_require_ciphers do a similar thing
3840 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3841 before starting the TLS session.
3843 18. Tabs are now shown as \t in -bP output.
3845 19. If the log selector return_path_on_delivery was set, Exim crashed when
3846 bouncing a message because it had too many Received: header lines.
3848 20. If two routers both had headers_remove settings, and the first one included
3849 a superfluous trailing colon, the final name in the first list and the
3850 first name in the second list were incorrectly joined into one item (with a
3851 colon in the middle).
3857 1. Added -C and -D options to the exinext utility, mainly to make it easier
3858 to include in the automated testing, but these could be helpful when
3859 multiple configurations are in use.
3861 2. The exinext utility was not formatting the output nicely when there was
3862 an alternate port involved in the retry record key, nor when there was a
3863 message id as well (for retries that were specific to a specific message
3864 and a specific host). It was also confused by IPv6 addresses, because of
3865 the additional colons they contain. I have fixed the IPv4 problem, and
3866 patched it up to do a reasonable job for IPv6.
3868 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3869 delivery, the log line now contains "pipelined" if PIPELINING was used.
3871 4. An SMTP transport process used to panic and die if the bind() call to set
3872 an explicit outgoing interface failed. This has been changed; it is now
3873 treated in the same way as a connect() failure.
3875 5. A reference to $sender_host_name in the part of a conditional expansion
3876 that was being skipped was still causing a DNS lookup. This no longer
3879 6. The def: expansion condition was not recognizing references to header lines
3880 that used bh_ and bheader_.
3882 7. Added the _cache feature to named lists.
3884 8. The code for checking quota_filecount in the appendfile transport was
3885 allowing one more file than it should have been.
3887 9. For compatibility with Sendmail, the command line option
3895 and sets the incoming protocol and host name (for trusted callers). The
3896 host name and its colon can be omitted when only the protocol is to be set.
3897 Note the Exim already has two private options, -pd and -ps, that refer to
3898 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3899 "s", but I don't think that's a major issue.
3901 10. A number of refactoring changes to the code, none of which should affect
3904 (a) The number of logging options was getting close to filling up the
3905 32-bit word that was used as a bit map. I have split them into two classes:
3906 those that are passed in the argument to log_write(), and those that are
3907 only ever tested independently outside of that function. These are now in
3908 separate 32-bit words, so there is plenty of room for expansion again.
3909 There is no change in the user interface or the logging behaviour.
3911 (b) When building, for example, log lines, the code previously used a
3912 macro that called string_cat() twice, in order to add two strings. This is
3913 not really sufficiently general. Furthermore, there was one instance where
3914 it was actually wrong because one of the argument was used twice, and in
3915 one call a function was used. (As it happened, calling the function twice
3916 did not affect the overall behaviour.) The macro has been replaced by a
3917 function that can join an arbitrary number of extra strings onto a growing
3920 (c) The code for expansion conditions now uses a table and a binary chop
3921 instead of a serial search (which was left over from when there were very
3922 few conditions). Also, it now recognizes conditions like "pam" even when
3923 the relevant support is not compiled in: a suitably worded error message is
3924 given if an attempt is made to use such a condition.
3926 11. Added ${time_interval:xxxxx}.
3928 12. A bug was causing one of the ddress fields not to be passed back correctly
3929 from remote delivery subprocesses. The field in question was not being
3930 subsequently used, so this caused to problems in practice.
3932 13. Added new log selectors queue_time and deliver_time.
3934 14. Might have fixed a bug in maildirsizefile handling that threw up
3935 "unexpected character" debug warnings, and recalculated the data
3936 unnecessarily. In any case, I expanded the warning message to give more
3939 15. Added the message "Restricted characters in address" to the statements in
3940 the default ACL that block characters like @ and % in local parts.
3942 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3943 Three changes have been made:
3945 (a) There was a serious bug; a negative response to MAIL caused the whole
3946 recipient domain to be cached as invalid, thereby blocking all messages
3947 to all local parts at the same domain, from all senders. This bug has
3948 been fixed. The domain is no longer cached after a negative response to
3949 MAIL if the sender used is not empty.
3951 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3954 (c) A new callout option, "use_sender" has been added for people who want
3955 the modified behaviour.
3961 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3964 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3965 indeed breaks things for older releases.
3967 3. Added additional logging to the case where there is a problem reading data
3968 from a filter that is running in a subprocess using a pipe, in order to
3969 try to track down a specific problem.
3971 4. Testing facility fudge: when running in the test harness and attempting
3972 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3973 getting "No route to host". Convert this to a timeout.
3975 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3978 6. Some OS don't have socklen_t but use size_t instead. This affects the
3979 fifth argument of getsockopt() amongst other things. This is now
3980 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3981 can be set for individual OS. I have set it for SunOS5, OSF1, and
3982 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3983 some earlier ones do not.
3985 7. Change 4.30/15 was not doing the test caselessly.
3987 8. The standard form for an IPv6 address literal was being rejected by address
3988 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3989 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3990 this, as well as the form without the "IPv6" on the front (but only when
3991 address literals are enabled, of course).
3993 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3995 10. Exim crashed if a message with an empty sender address specified by -f
3996 encountered a router with an errors_to setting. This could be provoked only
3997 by a command such as
4001 where an empty string was supplied; "<>" did not hit this bug.
4003 11. Installed PCRE release 4.5.
4005 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4006 remained set. It is now erased.
4008 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4009 times from message ids (which are base 36 rather than the normal 62).
4011 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4012 were being counted as actual protocol errors, and logged if the log
4013 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4014 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4015 and DATA following a set of rejected RCPTs do not count as protocol errors.
4016 In other words, Exim assumes they were pipelined, though this may not
4017 actually be the case. Of course, in all cases the client gets an
4018 appropriate error code.
4020 15. If a lookup fails in an ACL condition, a message about the failure may
4021 be available; it is used if testing the ACL cannot continue, because most
4022 such messages specify what the cause of the deferral is. However, some
4023 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4024 that caused an old message to be retained and used if a later statement
4025 caused a defer, replacing the real cause of the deferral.
4027 16. If an IP address had so many PTR records that the DNS lookup buffer
4028 was not large enough to hold them, Exim could crash while trying to process
4029 the truncated data. It now detects and logs this case.
4031 17. Further to 4.21/58, another change has been made: if (and only if) the
4032 first line of a message (the first header line) ends with CRLF, a bare LF
4033 in a subsequent header line has a space inserted after it, so as not to
4034 terminate the header.
4036 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4037 unnecessarily, used atoi() instead of strtol(), and didn't check the
4038 termination when getting file sizes from file names by regex.
4040 19. Completely re-implemented the support for maildirsize files, in the light
4041 of a number of problems with the previous contributed implementation
4042 (4.30/29). In particular:
4044 . If the quota is zero, the maildirsize file is maintained, but no quota is
4047 . If the maildir directory does not exist, it is created before any attempt
4048 to write a maildirsize file.
4050 . The quota value in the file is just a cache; if the quota is changed in
4051 the transport, the new value overrides.
4053 . A regular expression is available for excluding directories from the
4056 20. The autoreply transport checks the characters in options that define the
4057 message's headers; it allows continued headers, but it was checking with
4058 isspace() after an embedded newline instead of explicitly looking for a
4061 21. If all the "regular" hosts to which an address was routed had passed their
4062 expiry times, and had not reached their retry times, the address was
4063 bounced, even if fallback hosts were defined. Now Exim should go on to try
4066 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4067 equivalent code in the SMTP transport. Some hosts send humungous responses
4068 to HELO/EHLO, more than 1024 it seems.
4070 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4071 gives compiler warnings in some environments. I've now done it "properly",
4074 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4075 (because of problems with the built-in one) was declared to return uschar *
4076 instead of char *, causing compiler failure.
4078 25. Fixed a file descriptor leak when processing alias/forward files.
4080 26. Fixed a minor format string issue in dbfn.c.
4082 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4084 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4085 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4088 29. When a sender address is verified, it is cached, to save repeating the test
4089 when there is more than one recipient in a message. However, when the
4090 verification involves a callout, it is possible for different callout
4091 options to be set for different recipients. It is too complicated to keep
4092 track of this in the cache, so now Exim always runs a verification when a
4093 callout is required, relying on the callout cache for the optimization.
4094 The overhead is duplication of the address routing, but this should not be
4097 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4098 to be verified with callout=postmaster, and the main callout worked but the
4099 postmaster check failed, the verification correctly failed. However, if a
4100 subsequent RCPT command asked for sender verification *without* the
4101 postmaster check, incorrect caching caused this verification also to fail,
4104 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4105 it was not caching the DNS options (qualify_single, search_parents) that
4106 were used when the lookup failed. A subsequent lookup with different
4107 options therefore always gave the same answer, though there were cases
4108 where it should not have. (Example: a "domains = !$mx_any" option on a
4109 dnslookup router: the "domains" option is always processed without any
4110 widening, but the router might have qualify_single set.) Now Exim uses the
4111 cached value only when the same options are set.
4113 32. Added John Jetmore's "exipick" utility to the distribution.
4115 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4116 than a timeout (e.g. a certificate is required, and is not provided), an
4117 Exim server now closes the connection immediately. Previously it waited for
4118 the client to close - but if the client is SSL, it seems that they each
4119 wait for each other, leading to a delay before one of them times out.
4121 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4122 maintained 0.8.x compatibility because I don't think many are using it, and
4123 it is clearly obsolete.
4125 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4128 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4129 client certificate was expired. A simple patch fixes this, though I don't
4130 understand the full logic of why the verify callback is called multiple
4133 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4134 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4135 which causes problems with some clients (such as the Certicom SSL Plus
4136 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4137 disables the coutermeasure allowing Eudora to connect."
4139 38. Exim was not checking that a write() to a log file succeeded. This could
4140 lead to Bad Things if a log got too big, in particular if it hit a file
4141 size limit. Exim now panics and dies if it cannot write to a log file, just
4142 as it does if it cannot open a log file.
4144 39. Modified OS/Makefile-Linux so that it now contains
4146 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4148 The two -D definitions ensure that Exim is compiled with large file
4149 support, which makes it possible to handle log files that are bigger than
4152 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4153 instance) a domain was checked against a named list that involved a lookup,
4154 causing $domain_data to be set, then another domain was checked against the
4155 same list, then the first domain was re-checked, the value of $domain_data
4156 after the final check could be wrong. In particular, if the second check
4157 failed, it could be set empty. This bug probably also applied to
4160 41. The strip_trailing_dot option was not being applied to the address given
4161 with the -f command-line option.
4163 42. The code for reading a message's header from the spool was incrementing
4164 $received_count, but never initializing it. This meant that the value was
4165 incorrect (doubled) while delivering a message in the same process in which
4166 it was received. In the most common configuration of Exim, this never
4167 happens - a fresh exec is done - but it can happen when
4168 deliver_drop_privilege is set.
4170 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4171 it now includes up to 150 characters of the unexpected data in the log
4174 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4175 and building data strings. The size of both of these buffers was 10 000
4176 bytes - far larger than anybody would *ever* want, thought I. Needless to
4177 say, somebody hit the limit. I have increased the maximum line length to
4178 20 000 and the maximum data length of concatenated lines to 100 000. I have
4179 also fixed two bugs, because there was no checking on these buffers. Tsk,
4180 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4181 buffer is too small.
4183 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4184 lsearch lookups. Now it does.
4186 46. When parsing a route_list item in a manualroute router, a fixed-length
4187 buffer was used for the list of hosts. I made this 1024 bytes long,
4188 thinking that nobody would ever have a list of hosts that long. Wrong.
4189 Somebody had a whole pile of complicated expansion conditions, and the
4190 string was silently truncated, leading to an expansion error. It turns out
4191 that it is easier to change to an unlimited length (owing to other changes
4192 that have happened since this code was originally written) than to build
4193 structure for giving a limitation error. The length of the item that
4194 expands into the list of hosts is now unlimited.
4196 47. The lsearch lookup could not handle data where the length of text line was
4197 more than 4095 characters. Such lines were truncated, leading to shortened
4198 data being returned. It should now handle lines of any length.
4200 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4201 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4204 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4205 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4206 no longer happens in comment lines. A list of replacements is now placed
4207 at the head of all of the source files, except those whose only change is
4208 to replace PERL_COMMAND in the very first #! line.
4210 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4211 messages on the queue, with a bottom-up merge sort, using code contributed
4212 by Michael Haardt. This should make operations like -bp somewhat faster on
4213 large queues. It won't affect queue runners, except when queue_run_in_order
4216 51. Installed eximstats 1.31 in the distribution.
4218 52. Added support for SRV lookups to the dnslookup router.
4220 53. If an ACL referred to $message_body or $message_body_end, the value was not
4221 reset for any messages that followed in the same SMTP session.
4223 54. The store-handling optimization for building very long strings was not
4224 differentiating between the different store pools. I don't think this
4225 actually made any difference in practice, but I've tidied it.
4227 55. While running the routers to verify a sender address, $sender_address
4228 was still set to the sender address. This is wrong, because when routing to
4229 send a bounce to the sender, it would be empty. Therefore, I have changed
4230 it so that, while verifying a sender address, $sender_address is set to <>.
4231 (There is no change to what happens when verifying a recipient address.)
4233 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4234 target A or AAAA records (if not already returned) without resetting the
4235 qualify_single or search_parents options of the DNS resolver. These are
4236 inappropriate in this case because the targets of MX and SRV records must
4237 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4238 target that, when qualified, matched something in the local domain. These
4239 two options are now turned off when doing these lookups.
4241 57. It seems that at least some releases of Reiserfs (which does not have the
4242 concept of a fixed number of inodes) returns zero and not -1 for the
4243 number of available inodes. This interacted badly with check_spool_inodes,
4244 which assumed that -1 was the "no such thing" setting. What I have done is
4245 to check that the total number of inodes is greater than zero before doing
4246 the test of how many are available.
4248 58. When a "warn" ACL statement has a log_message modifier, the message is
4249 remembered, and not repeated. This is to avoid a lot of repetition when a
4250 message has many recipients that cause the same warning to be written.
4251 Howewer, Exim was preserving the list of already written lines for an
4252 entire SMTP session, which doesn't seem right. The memory is now reset if a
4253 new message is started.
4255 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4256 debugging output unless log_rewrite was also set.
4258 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4259 of (int)(handle) when we know that handle contains (void *)(-1).
4261 61. The Exim daemon panic-logs an error return when it closes the incoming
4262 connection. However "connection reset by peer" seems to be common, and
4263 isn't really an error worthy of noting specially, so that particular error
4266 62. When Exim is trying to find all the local interfaces, it used to panic and
4267 die if the ioctl to get the interface flags failed. However, it seems that
4268 on at least one OS (Solaris 9) it is possible to have an interface that is
4269 included in the list of interfaces, but for which you get a failure error
4270 for this call. This happens when the interface is not "plumbed" into a
4271 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4272 failure of the "get flags" call assumes that the interface is down.
4274 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4275 makes life easier for people who are doing arithmetic on fields extracted
4276 from dates, where you often get leading zeros that should not be
4277 interpreted as octal.
4279 64. Added qualify_domain to the redirect router, to override the global
4282 65. If a pathologically long header line contained very many addresses (the
4283 report of this problem mentioned 10 000) and each of them was rewritten,
4284 Exim could use up a very large amount of memory. (It kept on making new
4285 copies of the header line as it rewrote, and never released the old ones.)
4286 At the expense of a bit more processing, the header rewriting function has
4287 been changed so that it no longer eats memory in this way.
4289 66. The generation of the Received: header has been moved from the time that a
4290 message starts to be received, to the time that it finishes. The timestamp
4291 in the Received: header should now be very close to that of the <= log
4292 line. There are two side-effects of this change:
4294 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4295 logged header lines no longer include the local Received: line, because
4296 it has not yet been created. The same applies to a copy of the message
4297 that is returned to a non-SMTP sender when a message is rejected.
4299 (b) When a filter file is tested using -bf, no additional Received: header
4300 is added to the test message. After some thought, I decided that this
4303 This change does not affect the value of $received_for. It is still set
4304 after address rewriting, but before local_scan() is called.
4306 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4308 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4309 gave an unhelpful panic error message, and a defer error. I have managed to
4310 change this behaviour so that it now rejects any supplied certificate,
4311 which seems right, as the list of acceptable certificates is empty.
4313 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4314 gave an unhelpful defer error. I have not managed to make this reject any
4315 supplied certificates, but the error message it gives is "no certificate
4316 supplied", which is not helpful.
4318 70. exigrep's output now also includes lines that are not associated with any
4319 message, but which match the given pattern. Implemented by a patch from
4320 Martin Sluka, which also tidied up the Perl a bit.
4322 71. Recipient callout verification, like sender verification, was using <> in
4323 the MAIL FROM command. This isn't really the right thing, since the actual
4324 sender may affect whether the remote host accepts the recipient or not. I
4325 have changed it to use the actual sender in the callout; this means that
4326 the cache record is now keyed on a recipient/sender pair, not just the
4327 recipient address. There doesn't seem to be a real danger of callout loops,
4328 since a callout by the remote host to check the sender would use <>.
4329 [SEE ABOVE: changed after hitting problems.]
4331 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4332 temporary errors. However, in the case of such a code being given after
4333 the end of a data transmission (i.e. after ".") Exim was failing to write
4334 a retry record for the message. (Yes, there was some broken host that was
4335 actually sending 8xx at this point.)
4337 73. An unknown lookup type in a host list could cause Exim to panic-die when
4338 the list was checked. (An example that provoked this was putting <; in the
4339 middle of a list instead of at the start.) If this happened during a DATA
4340 ACL check, a -D file could be left lying around. This kind of configuration
4341 error no longer causes Exim to die; instead it causes a defer errror. The
4342 incident is still logged to the main and panic logs.
4344 74. Buglet left over from Exim 3 conversion. The message "too many messages
4345 in one connection" was written to the rejectlog but not the mainlog, except
4346 when address rewriting (yes!) was being logged.
4348 75. Added write_rejectlog option.
4350 76. When a system filter was run not as root (that is, when system_filter_user
4351 was set), the values of the $n variables were not being returned to the
4352 main process; thus, they were not subsequently available in the $sn
4355 77. Added +return_path_on_delivery log selector.
4357 78. A connection timeout was being treated differently from recipients deferred
4358 when testing hosts_max_try with a message that was older than the host's
4359 retry timeout. (The host should not be counted, thus allowing all hosts to
4360 be tried at least once before bouncing.) This may have been the cause of an
4361 occasionally reported bug whereby a message would remain on the queue
4362 longer than the retry timeout, but would be bounced if a delivery was
4363 forced. I say "may" because I never totally pinned down the problem;
4364 setting up timeout/retry tests is difficult. See also the next item.
4366 79. The ultimate address timeout was not being applied to errors that involved
4367 a combination of host plus message (for example, a timeout on a MAIL
4368 command). When an address resolved to a number of possible hosts, and they
4369 were not all tried for each delivery (e.g. because of hosts_max_try), a
4370 message could remain on the queue longer than the retry timeout.
4372 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4375 81. Fixed an obscure SMTP outgoing bug which required at least the following
4376 conditions: (a) there was another message waiting for the same server;
4377 (b) the server returned 5xx to all RCPT commands in the first message so
4378 that the message was not completed; (c) the server dropped the connection
4379 or gave a negative response to the RSET that Exim sends to abort the
4380 transaction. The observed case was a dropped connection after DATA that had
4381 been sent in pipelining mode. That is, the server had advertised PIPELINING
4382 but was not implementing it correctly. The effect of the bug was incorrect
4383 behaviour, such as trying another host, and this could lead to a crash.
4389 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4390 and daemon.c were passed as pointers to ints; they should have been
4391 pointers to socklen_t variables (which are typically unsigned ints).
4393 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4396 3. Fixed a really odd bug that affected only the testing scheme; patching a
4397 certain fixed string in the binary changed the value of another string that
4398 happened to be identical to the end of the original first string.
4400 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4401 name", it returns that address as the IP address. On some operating
4402 systems (e.g. Solaris), it also passes back the IP address string as the
4403 "host name". However, on others (e.g. Linux), it passes back an empty
4404 string. Exim wasn't checking for this, and was changing the host name to an
4405 empty string, assuming it had been canonicized.
4407 5. Although rare, it is permitted to have more than one PTR record for a given
4408 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4409 all the names associated with an address, because they do in Solaris.
4410 However, it seems that they do not in Linux for data that comes from the
4411 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4412 I found this out when I moved to a new Linux workstation and tried to run
4413 the Exim test suite.
4415 To get round this problem I have changed the code so that it now does its
4416 own call to the DNS to look up PTR records when searching for a host name.
4417 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4418 addresses that are only in /etc/hosts are still found.
4420 This behaviour is, however, controlled by an option called host_lookup_
4421 order, which defaults to "bydns:byaddr". If people want to use the other
4422 order, or indeed, just use one or the other means of lookup, they can
4423 specify it in this variable.
4425 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4426 some operating systems, this comes back from gethostbyaddr() as an empty
4427 string, and this is what Exim used to test for. However, it seems that in
4428 other systems, "." is yielded. Exim now tests for this case too.
4430 7. The values of check_spool_space and check_log_space are now held internally
4431 as a number of kilobytes instead of an absolute number of bytes. If a
4432 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4433 kilobyte. This means that much larger values can be stored.
4435 8. Exim monitor: an attempt to get the action menu when not actually pointing
4436 at a message produces an empty menu entitled "No message selected". This
4437 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4438 no entries in it ("Shell widget menu has zero width and/or height"). So I
4439 have added a single, blank menu entry in this case.
4441 9. Added ${quote_local_part.
4443 10. MIME decoding is now applied to the contents of Subject: header lines when
4446 11. Now that a reference to $sender_host_address automatically causes a reverse
4447 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4448 host lookup before query-style lookups in lists that might use this
4449 variable. This has therefore been abolished, and the "net-" prefix is no
4450 longer necessary for query-style lookups.
4452 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4453 have been a typo for LFLAGS, so it has been changed.
4455 13. The install script calls Exim with "-C /dev/null" in order to find the
4456 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4457 to be output. Howeve, since Exim outputs its version number before the
4458 error, it didn't break the script. It just looked ugly. I fixed this by
4459 always allowing "-C /dev/null" if the caller is root.
4461 14. Ignore overlarge ACL variable number when reading spool file - insurance
4462 against a later release with more variables having written the file.
4464 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4465 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4466 this, as well as the form without the "IPv6" on the front.
4468 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4469 OS/Makefile-Darwin file.
4471 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4472 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4474 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4475 variable. However, some OS use ints for inodes. I've added cast to long int
4476 to get rid of the compiler warning.
4478 19. I had forgotten to lock out "/../" in configuration file names when
4479 ALT_CONFIG_PREFIX was set.
4481 20. Routers used for verification do not need to specify transports. However,
4482 if such a router generated a host list, and callout was configured, Exim
4483 crashed, because it could not find a port number from the (non-existent)
4484 transport. It now assumes port 25 in this circumstance.
4486 21. Added the -t option to exigrep.
4488 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4489 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4490 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4491 them is set, LOOKUP_LSEARCH is forced.
4493 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4495 24. Added sender and host information to the "rejected by local_scan()" log
4496 line; previously there was no indication of these.
4498 25. Added .include_if_exists.
4500 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4501 Linux. It turns out that not all file systems support this. Apparently some
4502 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4503 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4504 not supported on the file descriptor, is now ignored when Exim is trying to
4505 sync a directory. This applies only to Linux.
4507 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4509 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4510 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4513 29. Added support for maildirsize files from supplied patch (modified a bit).
4515 30. The use of :fail: followed by an empty string could lead Exim to respond to
4516 sender verification failures with (e.g.):
4518 550 Verification failed for <xxx>
4519 550 Sender verify failed
4521 where the first response line was missing the '-' that indicates it is not
4522 the final line of the response.
4524 31. The loop for finding the name of the user that called Exim had a hardwired
4525 limit of 10; it now uses the value of finduser_retries, which is used for
4526 all other user lookups.
4528 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4531 33. Exim was neglecting to zero errno before one call of strtol() when
4532 expanding a string and expecting an integer value. On some systems this
4533 resulted in spurious "integer overflow" errors. Also, it was casting the
4534 result into an int without checking.
4536 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4537 did not work. The code looks as if it has *never* worked, though it appears
4538 to have been documented since at least releast 1.62. I have made it work.
4540 35. The "timeout_DNS" error in retry rules, also documented since at least
4541 1.62, also never worked. As it isn't clear exactly what this means, and
4542 clearly it isn't a major issue, I have abolished the feature by treating it
4543 as "timeout", and writing a warning to the main and panic logs.
4545 36. The display of retry rules for -brt wasn't always showing the error code
4548 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4549 timeout_connect_A, timeout_connect_MX.
4551 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4552 to the empty sender.
4554 39. The daemon was not analysing the content of -oX till after it had closed
4555 stderr and disconnected from the controlling terminal. This meant that any
4556 syntax errors were only noted on the panic log, and the return code from
4557 the command was 0. By re-arranging the code a little, I've made the
4558 decoding happen first, so such errors now appear on stderr, and the return
4559 code is 1. However, the actual setting up of the sockets still happens in
4560 the disconnected process, so errors there are still only recorded on the
4563 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4564 connections (as happens on some IP stacks) was logged at start up time as
4565 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4566 it from "IPv6 and IPv4", which means that two separate sockets are being
4569 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4570 says whether AF_INET or AF_INET6 was passed as an argument.
4572 42. Exiwhat output was messed up when time zones were included in log
4575 43. Exiwhat now gives more information about the daemon's listening ports,
4576 and whether -tls-on-connect was used.
4578 44. The "port" option of the smtp transport is now expanded.
4580 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4581 silently ignored. Now an error message is written to the main and panic
4584 46. There's a new ACL modifier called "logwrite" which writes to a log file
4585 as soon as it is encountered.
4587 47. Added $local_user_uid and $local_user_gid at routing time.
4589 48. Exim crashed when trying to verify a sender address that was being
4592 49. Exim was recognizing only a space character after ".include". It now also
4593 recognizes a tab character.
4595 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4596 extracting the relevant information from the specification. The man page no
4597 longer contains scrambled data for the -d option, and I've added a section
4598 at the front about calling Exim under different names.
4600 51. Added "extra_headers" argument to the "mail" command in filter files.
4602 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4605 53. Installed eximstats 1.29.
4607 54. Added transport_filter_timeout as a generic transport option.
4609 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4610 Cc: header lines. This was required by RFC 822, but it not required by RFC
4613 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4614 incoming messages that did not have them. Now it does so only if the
4615 message originates locally, that is, if there is no associated remote host
4616 address. When Resent- header lines are present, this applies to the Resent-
4617 lines rather than the non-Resent- lines.
4619 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4620 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4622 58. Messages for configuration errors now include the name of the main
4623 configuration file - useful now that there may be more than one file in a
4624 list (.included file names were always shown).
4626 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4627 for those rare installations that do not start the daemon as root or run it
4628 setuid root. I've cut out the call to initgroups() if the daemon is not
4631 60. The Exim user and group can now be bound into the binary as text strings
4632 that are looked up at the start of Exim's processing.
4634 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4636 62. Added $mailstore_basename variable.
4638 63. Installed patch to sieve.c from Michael Haardt.
4640 64. When Exim failed to open the panic log after failing to open the main log,
4641 the original message it was trying to log was written to stderr and debug
4642 output, but if they were not available (the usual case in production), it
4643 was lost. Now it is written to syslog before the two lines that record the
4644 failures to open the logs.
4646 65. Users' Exim filters run in subprocesses under the user's uid. It is
4647 possible for a "deliver" command or an alias in a "personal" command to
4648 provoke an address rewrite. If logging of address rewriting is configured,
4649 this fails because the process is not running as root or exim. There may be
4650 a better way of dealing with this, but for the moment (because 4.30 needs
4651 to be released), I have disabled address rewrite logging when running a
4652 filter in a non-root, non-exim process.
4658 1. The buildconfig auxiliary program wasn't quoting the value set for
4659 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4660 not defined. This bug was masked in 4.22 by the effect that was fixed in
4663 2. Some messages that were rejected after a message id was allocated were
4664 shown as "incomplete" by exigrep. It no longer does this for messages that
4665 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4667 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4668 have allow_domain_literals set, the ID did not get logged in the <= line.
4669 Domain literals are now always recognized in Message-ID: header lines.
4671 4. The first argument for a ${extract expansion item is the key name or field
4672 number. Leading and trailing spaces in this item were not being ignored,
4673 causing some misleading effects.
4675 5. When deliver_drop_privilege was set, single queue runner processes started
4676 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4677 same command in the process it spins off) were not dropping privilege.
4679 6. When the daemon running as "exim" started a queue runner, it always
4680 re-executed Exim in the spun-off process. This is a waste of effort when
4681 deliver_drop_privilege is set. The new process now just calls the
4682 queue-runner function directly.
4688 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4691 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4692 ignored. Though the use of -oP was forcing the writing of a pid file, it
4693 was always written to the default place.
4695 3. If the message "no IP address found for host xxxx" is generated during
4696 incoming verification, it is now followed by identification of the incoming
4697 connection (so you can more easily find what provoked it).
4699 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4701 5. Added some features to "harden" Exim a bit more against certain attacks:
4703 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4704 be put in Local/Makefile. This is like the never_users runtime option,
4705 but it cannot be overridden. The default setting is "root".
4707 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4708 prefix string with which any file named in a -C command line option
4711 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4712 is retained for -C and -D only if the caller of Exim is root. Without
4713 it, the exim user may also use -C and -D and retain privilege.
4715 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4716 command line option is disabled.
4718 6. Macro names set by the -D option must start with an upper case letter, just
4719 like macro names defined in the configuration file.
4721 7. Added "dereference=" facility to LDAP.
4723 8. Two instances of the typo "uknown" in the source files are fixed.
4725 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4726 the Configure-Makefile script screwed up while processing it.
4728 10. Incorporated PCRE 4.4.
4730 11. The SMTP synchronization check was not operating right at the start of an
4731 SMTP session. For example, it could not catch a HELO sent before the client
4732 waited for the greeting. There is now a check for outstanding input at the
4733 point when the greeting is written. Because of the duplex, asynchronous
4734 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4735 way, but not yet received, when the check is performed.
4737 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4738 on TCP/IP sockets, because this apparently causes some broken clients to
4741 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4742 unchanged) from the Cygwin maintainer.
4744 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4745 was supported instead of testing for mbx. Effectively a typo.
4747 15. The spa authenticator server code was not checking that the input it
4748 received was valid base64.
4750 16. The debug output line for the "set" modifier in ACLs was not showing the
4751 name of the variable that was being set.
4753 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4755 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4756 lookup on its own. Something else was needed to trigger the lookup. For
4757 example, a match in host_lookup or the need for a host name in a host list.
4758 Now, if $sender_host_name is referenced and the host name has not yet been
4759 looked up, a lookup is performed. If the lookup fails, the variable remains
4760 empty, and $host_lookup_failed is set to "1".
4762 19. Added "eqi" as a case-independent comparison operator.
4764 20. The saslauthd authentication condition could segfault if neither service
4765 nor realm was specified.
4767 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4768 error message that was logged was misleading, and incoming SMTP
4769 connections were dropped. The message is now more accurate, and temporary
4770 errors are given to SMTP connections.
4772 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4773 (except RSET) with a 421 error, until QUIT is received. However, it was
4774 failing to send a response to QUIT.
4776 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4777 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4778 ACL is now run after the helo_try_verify_hosts code.
4780 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4781 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4782 case-independent, so other case variants are also recognized). Apparently
4783 some systems use these upper case variants.
4785 25. If more than two messages were waiting for the same host, and a transport
4786 filter was specified for the transport, Exim sent two messages over the
4787 same TCP/IP connection, and then failed with "socket operation on non-
4788 socket" when it tried to send the third.
4790 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4792 27. The extern definition of crypt16() in expand.c was not being excluded when
4793 the OS had its own crypt16() function.
4795 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4796 as a preferred synonym for return_size_limit, both as an option and as an
4799 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4801 30. Changed the default configuration ACL to relax the local part checking rule
4802 for addresses that are not in any local domains. For these addresses,
4803 slashes and pipe symbols are allowed within local parts, but the sequence
4804 /../ is explicitly forbidden.
4806 31. SPA server authentication was not clearing the challenge buffer before
4809 32. log_message in a "warn" ACL statement was writing to the reject log as
4810 well as to the main log, which contradicts the documentation and doesn't
4811 seem right (because no rejection is happening). So I have stopped it.
4813 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4814 However, I am unable to do any testing of this.
4816 34. Fixed an infelicity in the appendfile transport. When checking directories
4817 for a mailbox, to see if any needed to be created, it was accidentally
4818 using path names with one or more superfluous leading slashes; tracing
4819 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4821 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4822 contents are added to the log line that is written for every discarded
4823 recipient. (Previously a log_message setting was ignored.)
4825 36. The ${quote: operator now quotes the string if it is empty.
4827 37. The install script runs exim in order to find its version number. If for
4828 some reason other than non-existence or emptiness, which it checks, it
4829 could not run './exim', it was installing it with an empty version number,
4830 i.e. as "exim-". This error state is now caught, and the installation is
4833 38. An argument was missing from the function that creates an error message
4834 when Exim fails to connect to the socket for saslauthd authentication.
4835 This could cause Exim to crash, or give a corrupted message.
4837 39. Added isip, isip4, and isip6 to ${if conditions.
4839 40. The ACL variables $acl_xx are now saved with the message, and can be
4840 accessed later in routers, transports, and filters.
4842 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4843 strings in the file are not string-expanded.
4845 42. If a MAIL command specified a SIZE value that was too large to fit into an
4846 int variable, the check against message_size_limit failed. Such values are
4847 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4848 day this will have to be increased, but I don't think I want to be around
4849 when emails are that large.
4856 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4857 iconv() is not standard in FreeBSD.
4859 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4860 IPv6 enabled. The observed symptom was a segmentation fault on return from
4861 the function os_common_find_running_interfaces() in src/os.c.
4863 3. In the check_special_case() function in daemon.c I had used "errno" as an
4864 argument name, which causes warnings on some systems. This was basically a
4865 typo, since it was named "eno" in the comments!
4867 4. The code that waits for the clock to tick (at a resolution of some fraction
4868 of a second) so as to ensure message-id uniqueness was always waiting for
4869 at least one whole tick, when it could have waited for less. [This is
4870 almost certainly not relevant at current processor speeds, where it is
4871 unlikely to ever wait at all. But we try to future-proof.]
4873 5. The function that sleeps for a time interval that includes fractions of a
4874 second contained a race. It did not block SIGALRM between setting the
4875 timer, and suspending (a couple of lines later). If the interval was short
4876 and the sigsuspend() was delayed until after it had expired, the suspension
4877 never ended. On busy systems this could lead to processes getting stuck for
4880 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4881 process, before it forks any delivery processes. The open lookup caching
4882 mechanism meant that the open file or database connection was passed into
4883 the delivery process. The problem was that delivery processes always tidy
4884 up cached lookup data. This could cause a problem for the next delivery
4885 process started by the queue runner, because the external queue runner
4886 process does not know about the closure. So the next delivery process
4887 still has data in the lookup cache. In the case of a file lookup, there was
4888 no problem because closing a file descriptor in a subprocess doesn't affect
4889 the parent. However, if the lookup was caching a connection to a database,
4890 the connection was closed, and the second delivery process was likely to
4891 see errors such as "PGSQL: query failed: server closed the connection
4892 unexpectedly". The problem has been fixed by closing all cached lookups
4893 in a queue runner before running a delivery process.
4895 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4896 seem to have the "const" qualifier which it has on other OS. I've
4899 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4900 *of the same type* (client or server) with the same public name that an
4901 error should be diagnosed.
4903 9. When Exim looked up a host name for an IP address, but failed to find the
4904 original IP address when looking up the host name (a safety check), it
4905 output the message "<ip address> does not match any IP for NULL", which was
4906 confusing, to say the least. The bug was that the host name should have
4907 appeared instead of "NULL".
4909 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4910 user that is built into the binary, and the -C or -D options is used, root
4911 privilege is dropped before the configuration file is read. In addition,
4912 logging is switched to stderr instead of the normal log files. If the
4913 configuration then re-defines the Exim user, the unprivileged environment
4914 is probably not what is expected, so Exim logs a panic warning message (but
4917 However, if deliver_drop_privilege is set, the unprivileged state may well
4918 be exactly what is intended, so the warning has been cut out in that case,
4919 and Exim is allowed to try to write to its normal log files.
4925 1. smtp_return_error_details was not giving details for temporary sender
4926 or receiver verification errors.
4928 2. Diagnose a configuration error if two authenticators have the same public
4931 3. Exim used not to create the message log file for a message until the first
4932 delivery attempt. This could be confusing when incoming messages were held
4933 for policy or load reasons. The message log file is now created at the time
4934 the message is received, and an initial "Received" line is written to it.
4936 4. The automatically generated man page for command line options had a minor
4937 bug that caused no ill effects; however, a more serious problem was that
4938 the procedure for building the man page automatically didn't always
4939 operate. Consequently, release 4.20 contains an out-of-date version. This
4940 shouldn't happen again.
4942 5. When building Exim with embedded Perl support, the script that builds the
4943 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4944 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4946 6. The freeze_tell option was not being used for messages that were frozen on
4947 arrival, either by an ACL or by local_scan().
4949 7. Added the smtp_incomplete_transaction log selector.
4951 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4952 was accepting AUTH without a new EHLO.
4954 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4955 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4958 10. Logging of TCP/IP connections (when configured) now happens in the main
4959 daemon process instead of the child process, so that the TCP/IP connection
4960 count is more accurate (but it can never be perfect).
4962 11. The use of "drop" in a nested ACL was not being handled correctly in the
4963 outer ACL. Now, if condition failure induced by the nested "drop" causes
4964 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4965 or "require"), the connection is dropped.
4967 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4968 that yield "discard" can now be used with an "accept" or a "discard" verb,
4969 but an error is generated for any others (because I can't see a useful way
4970 to define what should happen).
4972 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4973 are now processed in the same way as lines in the Exim configuration file.
4974 In particular, continuation lines are supported.
4976 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4978 15. Added -ti meaning -t -i.
4980 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4981 domains, and warn by logging if others are found.
4983 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4984 returned from GIFCONF when Exim is trying to find the list of interfaces on
4985 a host. The code in os.c has been modified to copy each ifreq to an aligned
4986 structure in all cases.
4988 Also, in some cases, the returned ifreq's were being copied to a 'struct
4989 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4990 means the last couple of bytes of an IPv6 address could be chopped if the
4991 ifreq contained only a normal sockaddr (14 bytes storage).
4993 18. Named domain lists were not supported in the hosts_treat_as_local option.
4994 An entry such as +xxxx was not recognized, and was treated as a literal
4997 19. Ensure that header lines added by a DATA ACL are included in the reject log
4998 if the ACL subsequently rejects the message.
5000 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5001 MD5 (which is deprecated).
5003 21. When testing a filter file using -bf, Exim was writing a message when it
5004 took the sender from a "From " line in the message, but it was not doing so
5005 when it took $return_path from a Return-Path: header line. It now does.
5007 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5008 with a valid header line field name (a series of printing characters
5009 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5011 23. Changed "disc" in the source to "disk" to conform to the documentation and
5012 the book and for uniformity.
5014 24. Ignore Sendmail's -Ooption=value command line item.
5016 25. When execve() failed while trying to run a command in a pipe transport,
5017 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5018 could be confused with a return value of 69 from the command itself. This
5019 has been changed to 127, the value the shell returns if it is asked to run
5020 a non-existent command. The wording for the related log line suggests a
5021 non-existent command as the problem.
5023 26. If received_header_text expands to an empty string, do not add a Received:
5024 header line to the message. (Well, it adds a token one on the spool, but
5025 marks it "old" so that it doesn't get used or transmitted.)
5027 27. Installed eximstats 1.28 (addition of -nt option).
5029 28. There was no check for failure on the call to getsockname() in the daemon
5030 code. This can fail if there is a shortage of resources on the system, with
5031 ENOMEM, for example. A temporary error is now given on failure.
5033 29. Contrary to the C standard, it seems that in some environments, the
5034 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5035 program. Exim now does this explicitly; it affects the formatting of
5036 timestamps using strftime().
5038 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5039 complaints. I've now initialized all the variables, to avoid this.
5041 32. Header lines added by a system filter were not being "seen" during
5042 transport-time rewrites.
5044 33. The info_callback() function passed to OpenSSL is set up with type void
5045 (*)(SSL *, int, int), as described somewhere. However, when calling the
5046 function (actually a macro) that sets it up, the type void(*)() is
5047 expected. I've put in a cast to prevent warnings from picky compilers.
5049 34. If a DNS black list lookup found a CNAME record, but there were no A
5050 records associated with the domain it pointed at, Exim crashed.
5052 35. If a DNS black list lookup returned more than one A record, Exim ignored
5053 all but the first. It now scans all returned addresses if a particular IP
5054 value is being sought. In this situation, the contents of the
5055 $dnslist_value variable are a list of all the addresses, separated by a
5058 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5059 to accept a host name and all its aliases if the forward lookup for any of
5060 them yielded the IP address of the incoming connection. Now it accepts only
5061 those names whose forward lookup yields the correct IP address. Any other
5062 names are discarded. This closes a loophole whereby a rogue DNS
5063 administrator could create reverse DNS records to break through a
5064 wildcarded host restriction in an ACL.
5066 37. If a user filter or a system filter that ran in a subprocess used any of
5067 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5068 the wrong values were passed to the pipe command ($thisaddress had the
5069 value of $0, $0 had the value of $1, etc). This bug was introduced by
5070 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5072 38. Improved the line breaking for long SMTP error messages from ACLs.
5073 Previously, if there was no break point between 40 and 75 characters, Exim
5074 left the rest of the message alone. Two changes have been made: (a) I've
5075 reduced the minimum length to 35 characters; (b) if it can't find a break
5076 point between 35 and 75 characters, it looks ahead and uses the first one
5077 that it finds. This may give the occasional overlong line, but at least the
5078 remaining text gets split now.
5080 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5081 file descriptors might be low, and that setting 1000 would always raise it.
5082 It turns out that in some environments, the limit is already over 1000 and
5083 that lowering it causes trouble. So now Exim takes care not to decrease it.
5085 40. When delivering a message, the value of $return_path is set to $sender_
5086 address at the start of routing (routers may change the value). By an
5087 oversight, this default was not being set up when an address was tested by
5088 -bt or -bv, which affected the outcome if any router or filter referred to
5091 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5092 log only when "message" or "log_message" are set. However, if one of the
5093 conditions was an address verification, or a call to a nested ACL, the
5094 messages generated by the underlying test were being passed through. This
5095 no longer happens. The underlying message is available in $acl_verify_
5096 message for both "message" and "log_message" expansions, so it can be
5097 passed through if needed.
5099 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5100 new expansion $bh_ to give the encoded byte string without charset
5101 translation. Translation happens only if iconv() is available; HAVE_ICONV
5102 indicates this at build time. HEADERS_CHARSET gives the charset to
5103 translate to; headers_charset can change it in the configuration, and
5104 "headers charset" can change it in an individual filter file.
5106 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5107 that creates RFC 2047 encoded "words" labels them as that charset instead
5108 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5109 expansion operator; (ii) when Exim creates a From: line for a local
5110 message; (iii) when a header line is rewritten to include a "phrase" part.
5112 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5113 buggy, causing it to skip the first lines of messages whose message ID
5114 ended in 'D'. This would not have bitten before Exim release 4.14, because
5115 message IDs were unlikely to end in 'D' before then. The effect was to have
5116 incorrect size information for certain domains.
5118 45. #include "config.h" was missing at the start of the crypt16.c module. This
5119 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5122 46. If there was a timeout during a "random" callout check, Exim treated it as
5123 a failure of the random address, and carried on sending RSET and the real
5124 address. If the delay was just some slowness somewhere, the response to the
5125 original RCPT would be taken as a response to RSET and so on, causing
5126 mayhem of various kinds.
5128 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5129 when I implemented it. It didn't allow for the fact that some option values
5130 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5131 the right test for positive values.
5133 48. Domain names in DNS records are case-independent. Exim always looks them up
5134 in lower case. Some resolvers return domain names in exactly the case they
5135 appear in the zone file, that is, they may contain uppercase letters. Not
5136 all resolvers do this - some return always lower case. Exim was treating a
5137 change of case by a resolver as a change of domain, similar to a widening
5138 of a domain abbreviation. This triggered its re-routing code and so it was
5139 trying to route what was effectively the same domain again. This normally
5140 caused routing to fail (because the router wouldn't handle the domain
5141 twice). Now Exim checks for this case specially, and just changes the
5142 casing of the domain that it ultimately uses when it transmits the message
5145 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5148 50. If a filter generated a file delivery with a non-absolute name (possible if
5149 no home directory exists for the router), the forbid_file option was not
5152 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5153 the existing equality matching.
5155 52. Exim was using ints instead of ino_t variables in some places where it was
5156 dealing with inode numbers.
5158 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5159 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5160 called TMPDIR, and if it finds it is different, it changes its value.
5162 54. The smtp_printf() function is now made available to local_scan() so
5163 additional output lines can be written before returning. There is also an
5164 smtp_fflush() function to enable the detection of a dropped connection.
5165 The variables smtp_input and smtp_batched_input are exported to
5168 55. Changed the default runtime configuration: the message "Unknown user"
5169 has been removed from the ACL, and instead placed on the localuser router,
5170 using the cannot_route_message feature. This means that any verification
5171 failures that generate their own messages won't get overridden. Similarly,
5172 the "Unrouteable address" message that was in the ACL for unverifiable
5173 relay addresses has also been removed.
5175 56. Added hosts_avoid_esmtp to the smtp transport.
5177 57. The exicyclog script was not checking for the esoteric option
5178 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5179 will work only if exicyclog is run under the appropriate euid.
5181 58. Following a discussion on the list, the rules by which Exim recognises line
5182 endings on incoming messages have been changed. The -dropcr and drop_cr
5183 options are now no-ops, retained only for backwards compatibility. The
5184 following line terminators are recognized: LF CRLF CR. However, special
5185 processing applies to CR:
5187 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5188 nor a local message in the state where . is a terminator.
5190 (ii) If a bare CR is encountered in a header line, an extra space is added
5191 after the line terminator so as not to end the header. The reasoning
5192 behind this is that bare CRs in header lines are most likely either
5193 to be mistakes, or people trying to play silly games.
5195 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5196 was being incorrectly given as 18 bytes larger than it should have been.
5197 This is a VOB (very old bug).
5199 60. This may never have affected anything current, but just in case it has:
5200 When the local host is found other than at the start of a list of hosts,
5201 the local host, those with the same MX, and any that follow, are discarded.
5202 When the list in question was part of a longer list of hosts, the following
5203 hosts (not currently being processed) were also being discarded. This no
5204 longer happens. I'm not sure if this situation could ever has previously
5207 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5210 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5211 header. This is something that is recommended in a new Internet Draft, and
5212 is something that is documented as being done by Sendmail. There are two
5213 possible values. For messages generated by the autoreply transport, Exim
5216 Auto-Submitted: auto-replied
5218 whereas for all other generated messages (e.g. bounces) it adds
5220 Auto-Submitted: auto-generated
5222 63. The "personal" condition in filters now includes a test for the
5223 Auto-Submitted: header. If it contains the string "auto-" the message it
5224 not considered personal.
5226 64. Added rcpt_include_affixes as a generic transport option.
5228 65. Added queue_only_override (default true).
5230 66. Added the syslog_duplication option.
5232 67. If what should have been the first header line of a message consisted of
5233 a space followed by a colon, Exim was mis-interpreting it as a header line.
5234 It isn't of course - it is syntactically invalid and should therefore be
5235 treated as the start of the message body. The misbehaviour could have
5236 caused a number of strange effects, including loss of data in subsequent
5237 header lines, and spool format errors.
5239 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5240 client host had authenticated. This control can now be exercised by an ACL
5241 for more flexibility.
5243 69. By default, callouts do not happen when testing with -bh. There is now a
5244 variant, -bhc, which does actually run the callout code, including
5245 consulting and updating the callout cache.
5247 70. Added support for saslauthd authentication, courtesy of Alexander
5250 71. If statvfs() failed on the spool or log directories while checking their
5251 size for availability, Exim confusingly gave the error "space shortage".
5252 Furthermore, in debugging mode it crashed with a floating point exception.
5253 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5254 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5255 serious problem, Exim now writes to the main and panic logs when this
5256 happens, with details of the failure. It then refuses to accept the
5257 incoming message, giving the message "spool directory problem" or "log
5258 directory problem" with a 421 code for SMTP messages.
5260 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5261 0, 1, and 2 exist, because some OS complain for execs without them (see
5262 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5263 descriptors. However, the code omitted to check that the open succeeded,
5264 causing mysterious errors if for some reason the permissions on /dev/null
5265 got screwed. Now Exim writes a message to the main and panic logs, and
5266 bombs out if it can't open /dev/null.
5268 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5269 interact so that it is all more flexible. It is supposed to remain
5270 backwards compatible. Also added extra_local_interfaces.
5272 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5273 to bomb out with an assertion failure - to the client this appears as a
5274 connection drop. This problem occurs in the part of the code that was taken
5275 from the Samba project. Fortunately, the assertion is in a very simple
5276 function, so I have fixed this by reproducing the function inline in the
5277 one place where it is called, and arranging for authentication to fail
5278 instead of killing the process with assert().
5280 75. The SPA client code was not working when the server requested OEM rather
5281 than Unicode encoding.
5283 76. Added code to make require_files with a specific uid setting more usable in
5284 the case where statting the file as root fails - usually a non-root-mounted
5285 NFS file system. When this happens and the failure is EACCES, Exim now
5286 forks a subprocess and does the per-uid checking as the relevant uid.
5288 77. Added process_log_path.
5290 78. If log_file_path was not explicitly set, a setting of check_log_space or
5291 check_log_inodes was ignored.
5293 79. If a space check for the spool or log partitions fails, the incident is now
5294 logged. Of course, in the latter case the data may get lost...
5296 80. Added the %p formatting code to string_format() so that it can be used to
5297 print addresses in debug_print(). Adjusted all the address printing in the
5298 debugging in store.c to use %p rather than %d.
5300 81. There was a concern that a line of code in smtp_in.c could overflow a
5301 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5302 initially expressed, the concern was not well-founded, because trailing
5303 spaces are removed early. However, if the trailing spaces were followed by
5304 a NULL, they did not get removed, so the overflow was possible. Two fixes
5307 (a) I re-wrote the offending code in a cleaner fashion.
5308 (b) If an incoming SMTP command contains a NULL character, it is rejected
5311 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5312 runs initgroups(), so that if the Exim user is in any additional groups,
5313 they will be used during message reception.
5319 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob