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
90 TF/01 Failure to get a lock on a hints database can have serious
91 consequences so log it to the panic log.
93 TF/02 Log LMTP confirmation messages in the same way as SMTP,
94 controlled using the smtp_confirmation log selector.
96 TF/03 Include the error message when we fail to unlink a spool file.
98 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
99 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
100 for maintaining out-of-tree patches for some time.
102 PP/01 Bugzilla 139: Documentation and portability issues.
103 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
104 Handle per-OS dynamic-module compilation flags.
106 PP/02 Let /dev/null have normal permissions.
107 The 4.73 fixes were a little too stringent and complained about the
108 permissions on /dev/null. Exempt it from some checks.
109 Reported by Andreas M. Kirchwitz.
111 PP/03 Report version information for many libraries, including
112 Exim version information for dynamically loaded libraries. Created
113 version.h, now support a version extension string for distributors
114 who patch heavily. Dynamic module ABI change.
116 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
117 privilege escalation vulnerability whereby the Exim run-time user
118 can cause root to append content of the attacker's choosing to
121 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
124 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
125 If dropping privileges for untrusted macros, we disabled normal logging
126 on the basis that it would fail; for the Exim run-time user, this is not
127 the case, and it resulted in successful deliveries going unlogged.
128 Fixed. Reported by Andreas Metzler.
134 PP/01 Date: & Message-Id: revert to normally being appended to a message,
135 only prepend for the Resent-* case. Fixes regression introduced in
136 Exim 4.70 by NM/22 for Bugzilla 607.
138 PP/02 Include check_rfc2047_length in configure.default because we're seeing
139 increasing numbers of administrators be bitten by this.
141 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
143 PP/03 Bugzilla 994: added openssl_options main configuration option.
145 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
147 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
149 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
151 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
152 without a peer certificate, leading to a segfault because of an
153 assumption that peers always have certificates. Be a little more
154 paranoid. Problem reported by Martin Tscholak.
156 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
157 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
158 NB: ClamAV planning to remove STREAM in "middle of 2010".
159 CL also introduces -bmalware, various -d+acl logging additions and
160 more caution in buffer sizes.
162 PP/09 Implemented reverse_ip expansion operator.
164 PP/10 Bugzilla 937: provide a "debug" ACL control.
166 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
168 PP/12 Bugzilla 973: Implement --version.
170 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
172 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
174 PP/15 Bugzilla 816: support multiple condition rules on Routers.
176 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
177 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
178 ignore trailing whitespace.
180 JJ/02 prevent non-panic DKIM error from being sent to paniclog
182 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
185 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
186 Notification from Dr Andrew Aitchison.
188 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
189 ExtendedDetectionInfo response format.
190 Notification from John Horne.
192 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
195 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
196 XSL and documented dependency on system catalogs, with examples of how
199 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
202 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
203 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
204 configuration file which is writeable by the Exim user or group.
206 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
207 of configuration files to cover files specified with the -C option if
208 they are going to be used with root privileges, not just the default
211 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
212 option (effectively making it always true).
214 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
215 files to be used while preserving root privileges.
217 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
218 that rogue child processes cannot use them.
220 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
221 run-time user, instead of root.
223 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
224 Exim run-time user without dropping privileges.
226 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
227 result string, instead of calling string_vformat() twice with the same
230 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
231 for other users. Others should always drop root privileges if they use
232 -C on the command line, even for a whitelisted configure file.
234 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
236 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
242 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
243 $data_path, and $header_path variables; fixed documentation bugs and
246 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
247 exipick to access non-standard spools, including the "frozen" queue
250 NM/01 Bugzilla 965: Support mysql stored procedures.
251 Patch from Alain Williams
253 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
255 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
256 Patch from Andreas Metzler
258 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
259 Patch from Kirill Miazine
261 NM/05 Bugzilla 671: Added umask to procmail example.
263 JJ/03 installed exipick 20100323.0, fixing doc bug
265 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
266 directory. Notification and patch from Dan Rosenberg.
268 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
270 TK/02 Improve log output when DKIM signing operation fails.
272 MH/01 Treat the transport option dkim_domain as a colon separated
273 list, not as a single string, and sign the message with each element,
274 omitting multiple occurences of the same signer.
276 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
277 Bugzilla 985, 986. Patch by Simon Arlott
279 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
280 Patch by Simon Arlott
282 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
283 MBX locking. Notification from Dan Rosenberg.
289 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
291 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
293 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
295 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
297 NM/04 Bugzilla 915: Flag broken perl installation during build.
303 TK/01 Added patch by Johannes Berg that expands the main option
304 "spamd_address" if it starts with a dollar sign.
306 TK/02 Write list of recipients to X-Envelope-Sender header when building
307 the mbox-format spool file for content scanning (suggested by Jakob
310 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
311 (http://www.dcc-servers.net/) support via dccifd. Activated by
312 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
314 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
315 by Mark Daniel Reidel <mr@df.eu>.
317 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
318 When building exim an external PCRE library is now needed -
319 PCRE is a system library on the majority of modern systems.
320 See entry on PCRE_LIBS in EDITME file.
322 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
323 conversation. Added nologin parameter to request.
324 Patch contributed by Kirill Miazine.
326 TF/01 Do not log submission mode rewrites if they do not change the address.
328 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
330 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
331 log files in place. Contributed by Roberto Lima.
333 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
335 TF/03 Bugzilla 615: When checking the local_parts router precondition
336 after a local_part_suffix or local_part_prefix option, Exim now
337 does not use the address's named list lookup cache, since this
338 contains cached lookups for the whole local part.
340 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
341 Robert Millan. Documentation is in experimental-spec.txt.
343 TF/04 Bugzilla 668: Fix parallel build (make -j).
345 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
347 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
348 Patch provided by Jan Srzednicki.
350 TF/05 Leading white space used to be stripped from $spam_report which
351 wrecked the formatting. Now it is preserved.
353 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
354 that they are available at delivery time.
356 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
358 TF/08 TLS error reporting now respects the incoming_interface and
359 incoming_port log selectors.
361 TF/09 Produce a more useful error message if an SMTP transport's hosts
362 setting expands to an empty string.
364 NM/06 Bugzilla 744: EXPN did not work under TLS.
365 Patch provided by Phil Pennock.
367 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
368 Patch provided by Richard Godbee.
370 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
371 acl_smtp_notquit, added index entry.
373 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
374 Patch provided by Eugene Bujak.
376 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
377 accept(). Patch provided by Maxim Dounin.
379 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
380 Patch provided by Phil Pennock.
382 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
384 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
385 Patch provided by Brad "anomie" Jorsch.
387 NM/14 Bugzilla 622: Added timeout setting to transport filter.
388 Patch provided by Dean Brooks.
390 TK/05 Add native DKIM support (does not depend on external libraries).
392 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
393 Patch provided by Graeme Fowler.
395 NM/16 Bugzilla 851: Documentation example syntax fix.
397 NM/17 Changed NOTICE file to remove references to embedded PCRE.
399 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
402 NM/19 Bugzilla 745: TLS version reporting.
403 Patch provided by Phil Pennock.
405 NM/20 Bugzilla 167: bool: condition support.
406 Patch provided by Phil Pennock.
408 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
409 clients. Patch provided by Phil Pennock.
411 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
412 Patch provided by Brad "anomie" Jorsch.
414 NM/23 Bugzilla 687: Fix misparses in eximstats.
415 Patch provided by Heiko Schlittermann.
417 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
418 Patch provided by Heiko Schlittermann.
420 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
421 plus update to original patch.
423 NM/26 Bugzilla 799: Documentation correction for ratelimit.
425 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
426 Patch provided by David Brownlee.
428 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
430 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
432 NM/30 Bugzilla 888: TLS documentation bugfixes.
434 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
436 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
437 Unlike the original bugzilla I have changed all shell scripts in src tree.
439 NM/33 Bugzilla 898: Transport filter timeout fix.
440 Patch by Todd Rinaldo.
442 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
443 Patch by Serge Demonchaux.
445 NM/35 Bugzilla 39: Base64 decode bug fixes.
446 Patch by Jakob Hirsch.
448 NM/36 Bugzilla 909: Correct connect() call in dcc code.
450 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
452 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
454 NM/39 Bugzilla 911: Fixed MakeLinks build script.
460 TK/01 Add preliminary DKIM support. Currently requires a forked version of
461 ALT-N's libdkim that I have put here:
462 http://duncanthrax.net/exim-experimental/
464 Note to Michael Haardt: I had to rename some vars in sieve.c. They
465 were called 'true' and it seems that C99 defines that as a reserved
466 keyword to be used with 'bool' variable types. That means you could
467 not include C99-style headers which use bools without triggering
468 build errors in sieve.c.
470 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
471 as mailq or other aliases. Changed the --help handling significantly
472 to do whats expected. exim_usage() emits usage/help information.
474 SC/01 Added the -bylocaldomain option to eximstats.
476 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
478 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
480 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
486 PH/01 Another patch from the Sieve maintainer.
488 PH/02 When an IPv6 address is converted to a string for single-key lookup
489 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
490 dots are used instead of colons so that keys in lsearch files need not
491 contain colons. This was done some time before quoting was made available
492 in lsearch files. However, iplsearch files do require colons in IPv6 keys
493 (notated using the quote facility) so as to distinguish them from IPv4
494 keys. This meant that lookups for IP addresses in host lists did not work
495 for iplsearch lookups.
497 This has been fixed by arranging for IPv6 addresses to be expressed with
498 colons if the lookup type is iplsearch. This is not incompatible, because
499 previously such lookups could never work.
501 The situation is now rather anomolous, since one *can* have colons in
502 ordinary lsearch keys. However, making the change in all cases is
503 incompatible and would probably break a number of configurations.
505 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
508 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
509 conversion specification without a maximum field width, thereby enabling
510 a rogue spamd server to cause a buffer overflow. While nobody in their
511 right mind would setup Exim to query an untrusted spamd server, an
512 attacker that gains access to a server running spamd could potentially
513 exploit this vulnerability to run arbitrary code as the Exim user.
515 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
516 $primary_hostname instead of what libspf2 thinks the hosts name is.
518 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
519 a directory entry by the name of the lookup key. Previously, if a
520 symlink pointed to a non-existing file or a file in a directory that
521 Exim lacked permissions to read, a lookup for a key matching that
522 symlink would fail. Now it is enough that a matching directory entry
523 exists, symlink or not. (Bugzilla 503.)
525 PH/03 The body_linecount and body_zerocount variables are now exported in the
528 PH/04 Added the $dnslist_matched variable.
530 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
531 This means they are set thereafter only if the connection becomes
534 PH/06 Added the client_condition to authenticators so that some can be skipped
535 by clients under certain conditions.
537 PH/07 The error message for a badly-placed control=no_multiline_responses left
538 "_responses" off the end of the name.
540 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
542 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
543 (without spaces) instead of just copying the configuration text.
545 PH/10 Added the /noupdate option to the ratelimit ACL condition.
547 PH/11 Added $max_received_linelength.
549 PH/12 Added +ignore_defer and +include_defer to host lists.
551 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
552 way in which PCRE > 7.0 is built.
554 PH/14 Implemented queue_only_load_latch.
556 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
557 MAIL command. The effect was to mangle the value on 64-bit systems.
559 PH/16 Another patch from the Sieve maintainer.
561 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
563 PH/18 If a system quota error occurred while trying to create the file for
564 a maildir delivery, the message "Mailbox is full" was not appended to the
565 bounce if the delivery eventually timed out. Change 4.67/27 below applied
566 only to a quota excession during the actual writing of the file.
568 PH/19 It seems that peer DN values may contain newlines (and other non-printing
569 characters?) which causes problems in log lines. The DN values are now
570 passed through string_printing() before being added to log lines.
572 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
573 and InterBase are left for another time.)
575 PH/21 Added message_body_newlines option.
577 PH/22 Guard against possible overflow in moan_check_errorcopy().
579 PH/23 POSIX allows open() to be a macro; guard against that.
581 PH/24 If the recipient of an error message contained an @ in the local part
582 (suitably quoted, of course), incorrect values were put in $domain and
583 $local_part during the evaluation of errors_copy.
589 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
590 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
593 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
594 issue a MAIL command.
596 PH/02 In an ACL statement such as
598 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
600 if a client was not listed at all, or was listed with a value other than
601 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
602 the condition was not true (as it should be), so access was not denied.
603 The bug was that the ! inversion was incorrectly passed on to the second
604 item. This has been fixed.
606 PH/03 Added additional dnslists conditions == and =& which are different from
607 = and & when the dns lookup returns more than one IP address.
609 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
610 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
612 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
613 FSYNC, which compiles an option called disable_fsync that allows for
614 bypassing fsync(). The documentation is heavily laced with warnings.
616 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
618 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
619 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
620 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
621 including adding "make clean"; (3) Added -fPIC when compiling the test
622 dynamically loaded module, to get rid of a warning.
624 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
625 message fails, move_frozen_messages = true and ignore_bounce_errors_after
626 = 0s. The bug is otherwise harmless.
628 PH/07 There was a bug in the dovecot authenticator such that the value of
629 $auth1 could be overwritten, and so not correctly preserved, after a
630 successful authentication. This usually meant that the value preserved by
631 the server_setid option was incorrect.
633 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
635 PH/09 Installed PCRE release 7.0.
637 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
638 run for batched SMTP input. It is now run at the start of every message
639 in the batch. While fixing this I discovered that the process information
640 (output by running exiwhat) was not always getting set for -bs and -bS
641 input. This is fixed, and it now also says "batched" for BSMTP.
643 PH/11 Added control=no_pipelining.
645 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
646 patch, slightly modified), and move the expansion of helo_data till after
647 the connection is made in the smtp transport (so it can use these
650 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
652 PH/14 Added log_selector = +pid.
654 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
656 PH/16 Add ${if forany and ${if forall.
658 PH/17 Added dsn_from option to vary the From: line in DSNs.
660 PH/18 Flush SMTP output before performing a callout, unless control =
661 no_callout_flush is set.
663 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
664 was true (the default) a successful delivery failed to delete the retry
665 item, thus causing premature timeout of the address. The bug is now
668 PH/20 Added hosts_avoid_pipelining to the smtp transport.
670 PH/21 Long custom messages for fakedefer and fakereject are now split up
671 into multiline reponses in the same way that messages for "deny" and
672 other ACL rejections are.
674 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
675 with slight modification.
677 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
678 draft, changing the syntax and factoring some duplicate code".
680 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
681 for deliveries of the second and subsequent messages over the same SMTP
684 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
685 ${reduce, with only minor "tidies".
687 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
689 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
690 expansion side effects.
692 PH/27 When a message times out after an over-quota error from an Exim-imposed
693 quota, the bounce message says "mailbox is full". This message was not
694 being given when it was a system quota that was exceeded. It now should
697 MH/03 Made $recipients available in local_scan(). local_scan() already has
698 better access to the recipient list through recipients_list[], but
699 $recipients can be useful in postmaster-provided expansion strings.
701 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
702 in the case of a MAIL command with additional options following the
703 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
704 were accidentally chopped off.
706 PH/29 SMTP synchronization checks are implemented when a command is read -
707 there is a check that no more input is waiting when there shouldn't be
708 any. However, for some commands, a delay in an ACL can mean that it is
709 some time before the response is written. In this time, more input might
710 arrive, invalidly. So now there are extra checks after an ACL has run for
711 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
712 pipelining has not been advertised.
714 PH/30 MH's patch to allow iscntrl() characters to be list separators.
716 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
717 returned in the SMTP response when smtp_return_error_details was false.
720 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
721 instead of the C I/O that was originally supplied, because problems were
724 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
725 Exim which did not show up earlier: it was assuming that a call to
726 SSL_CTX_set_info_callback() might give an error value. In fact, there is
727 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
728 was a macro that became an assignment, so it seemed to work. This has
729 changed to a proper function call with a void return, hence the compile
730 error. Exim's code has been fixed.
732 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
735 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
737 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
739 PH/37 If a message is not accepted after it has had an id assigned (e.g.
740 because it turns out to be too big or there is a timeout) there is no
741 "Completed" line in the log. When some messages of this type were
742 selected by exigrep, they were listed as "not completed". Others were
743 picked up by some special patterns. I have improved the selection
744 criteria to be more general.
746 PH/38 The host_find_failed option in the manualroute router can now be set
747 to "ignore", to completely ignore a host whose IP address cannot be
748 found. If all hosts are ignored, the behaviour is controlled by the new
749 host_all_ignored option.
751 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
752 homing or because of multiple MX records with /mx) generated more than
753 one IP address, and the following item turned out to be the local host,
754 all the secondary addresses of the first item were incorrectly removed
755 from the list, along with the local host and any following hosts (which
756 is what is supposed to happen).
758 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
759 whoever called Exim into the -H file. In the case of the daemon it was
760 behaving confusingly. When first started, it used values for whoever
761 started the daemon, but after a SIGHUP it used the Exim user (because it
762 calls itself on a restart). I have changed the code so that it now always
765 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
766 message are rejected with the same error (e.g. no authentication or bad
767 sender address), and a DATA command is nevertheless sent (as can happen
768 with PIPELINING or a stupid MUA), the error message that was given to the
769 RCPT commands is included in the rejection of the DATA command. This is
770 intended to be helpful for MUAs that show only the final error to their
773 PH/42 Another patch from the Sieve maintainer.
775 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
776 Eximstats - Fixed some broken HTML links and added missing column headers
778 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
779 columns for Rejects, Temp Rejects, Ham, and Spam rows.
781 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
783 PH/43 Yet another patch from the Sieve maintainer.
785 PH/44 I found a way to check for a TCP/IP connection going away before sending
786 the response to the final '.' that terminates a message, but only in the
787 case where the client has not sent further data following the '.'
788 (unfortunately, this is allowed). However, in many cases there won't be
789 any further data because there won't be any more messages to send. A call
790 to select() can be used: if it shows that the input is "ready", there is
791 either input waiting, or the socket has been closed. An attempt to read
792 the next input character can distinguish the two cases. Previously, Exim
793 would have sent an OK response which the client would never have see.
794 This could lead to message repetition. This fix should cure that, at
795 least in a lot of common cases.
797 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
798 advertised in response to EHLO.
804 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
805 fixed by 4.65/MH/01 (is this a record?) are fixed:
807 (i) An empty string was always treated as zero by the numeric comparison
808 operators. This behaviour has been restored.
810 (ii) It is documented that the numeric comparison operators always treat
811 their arguments as decimal numbers. This was broken in that numbers
812 starting with 0 were being interpreted as octal.
814 While fixing these problems I realized that there was another issue that
815 hadn't been noticed. Values of message_size_limit (both the global option
816 and the transport option) were treated as octal if they started with 0.
817 The documentation was vague. These values are now always treated as
818 decimal, and I will make that clear in the documentation.
824 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
825 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
828 MH/01 Don't check that the operands of numeric comparison operators are
829 integers when their expansion is in "skipping" mode (fixes bug
830 introduced by 4.64-PH/07).
832 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
833 child addresses, Exim now panics and dies. Previously, because the count
834 is held in a short int, deliveries were likely to be lost. As such a
835 large number of recipients for a single message is ridiculous
836 (performance will be very, very poor), I have chosen to impose a limit
837 rather than extend the field.
843 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
844 leftover -K file (the existence of which was triggered by #402).
845 While we were at it, introduced process PID as part of the -K
846 filename. This should rule out race conditions when creating
849 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
850 processing considerably. Previous code took too long for large mails,
851 triggering a timeout which in turn triggers #401.
853 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
854 in the DK code in transports.c. sendfile() is not really portable,
855 hence the _LINUX specificness.
857 TF/01 In the add_headers option to the mail command in an Exim filter,
858 there was a bug that Exim would claim a syntax error in any
859 header after the first one which had an odd number of characters
862 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
863 callout verification, Exim cached a "reject" for the entire domain. This
864 is correct for most verifications, but it is not correct for a recipient
865 verification with use_sender or use_postmaster set, because in that case
866 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
867 case of MAIL FROM:<> rejection from other early rejections (e.g.
868 rejection of HELO). When verifying a recipient using a non-null MAIL
869 address, the cache is ignored if it shows MAIL FROM:<> rejection.
870 Whatever the result of the callout, the value of the domain cache is
871 left unchanged (for any other kind of callout, getting as far as trying
872 RCPT means that the domain itself is ok).
874 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
877 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
878 manifest itself as EPIPE rather than ECONNECT. When tidying away a
879 session, the daemon ignores ECONNECT errors and logs others; it now
880 ignores EPIPE as well.
882 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
883 (quoted-printable decoding).
885 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
886 later the small subsequent patch to fix an introduced bug.
888 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
890 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
892 PH/08 An error is now given if message_size_limit is specified negative.
894 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
895 to be given (somewhat) arbitrary names.
897 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
900 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
901 miscellaneous code fixes
903 PH/10 Added the log_reject_target ACL modifier to specify where to log
906 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
907 hostname. This is wrong, because it relates to the incoming message (and
908 probably the interface on which it is arriving) and not to the outgoing
909 callout (which could be using a different interface). This has been
910 changed to use the value of the helo_data option from the smtp transport
911 instead - this is what is used when a message is actually being sent. If
912 there is no remote transport (possible with a router that sets up host
913 addresses), $smtp_active_hostname is used.
915 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
916 tweaks were necessary in order to get it to work (see also 21 below):
917 (a) The code assumed that strncpy() returns a negative number on buffer
918 overflow, which isn't the case. Replaced with Exim's string_format()
920 (b) There were several signed/unsigned issues. I just did the minimum
921 hacking in of casts. There is scope for a larger refactoring.
922 (c) The code used strcasecmp() which is not a standard C function.
923 Replaced with Exim's strcmpic() function.
924 (d) The code set only $1; it now sets $auth1 as well.
925 (e) A simple test gave the error "authentication client didn't specify
926 service in request". It would seem that Dovecot has changed its
927 interface. Fortunately there's a specification; I followed it and
928 changed what the client sends and it appears to be working now.
930 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
933 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
934 address A is aliased to B and C, where B exists and C does not. Without
935 -v the output is "A verified" because verification stops after a
936 successful redirection if more than one address is generated. However,
937 with -v the child addresses are also verified. Exim was outputting "A
938 failed to verify" and then showing the successful verification for C,
939 with its parentage. It now outputs "B failed to verify", showing B's
940 parentage before showing the successful verification of C.
942 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
943 look up a TXT record in a specific list after matching in a combined
946 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
947 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
948 they consult the DNS. I had assumed they would set it the way they
949 wanted; and indeed my experiments on Linux seem to show that in some
950 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
951 To be on the safe side, however, I have now made the interface to
952 host_find_byname() similar to host_find_bydns(), with an argument
953 containing the DNS resolver options. The host_find_byname() function now
954 sets these options at its start, just as host_find_bydns() does. The smtp
955 transport options dns_qualify_single and dns_search_parents are passed to
956 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
957 of host_find_byname() use the default settings of RES_DEFNAMES
958 (qualify_single) but not RES_DNSRCH (search_parents).
960 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
961 spool_read_header() do less string testing, by means of a preliminary
962 switch on the second character of optional "-foo" lines. (This is
963 overdue, caused by the large number of possibilities that now exist.
964 Originally there were few.) While I was there, I also converted the
965 str(n)cmp tests so they don't re-test the leading "-" and the first
966 character, in the hope this might squeeze out yet more improvement.
968 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
969 flag allowing group syntax was set by the header_syntax check but not
970 turned off, possible causing trouble later; (2) The flag was not being
971 set at all for the header_verify test, causing "group"-style headers to
972 be rejected. I have now set it in this case, and also caused header_
973 verify to ignore an empty address taken from a group. While doing this, I
974 came across some other cases where the code for allowing group syntax
975 while scanning a header line wasn't quite right (mostly, not resetting
976 the flag correctly in the right place). These bugs could have caused
977 trouble for malformed header lines. I hope it is now all correct.
979 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
980 with the "reply" argument non-NULL. The code, however (which originally
981 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
982 but it didn't always do it. This confused somebody who was copying the
983 code for some other use. I have removed all the tests.
985 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
986 feature that was used to support insecure browsers during the U.S. crypto
987 embargo. It requires special client support, and Exim is probably the
988 only MTA that supported it -- and would never use it because real RSA is
989 always available. This code has been removed, because it had the bad
990 effect of slowing Exim down by computing (never used) parameters for the
991 RSA_EXPORT functionality.
993 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
994 authenticator to fail if there's a tab character in the incoming data
995 (there should never be unless someone is messing about, as it's supposed
996 to be base64-encoded). Also added, on Timo's advice, the "secured" option
997 if the connection is using TLS or if the remote IP is the same as the
998 local IP, and the "valid-client-cert option" if a client certificate has
1001 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1002 authenticators. This can be used for authorization after authentication
1003 succeeds. (In the case of plaintext, it servers for both authentication
1006 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1007 if any retry times were supplied.
1009 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1010 connection, where there is no client IP address to check. In this
1011 situation, the verify now always succeeds.
1013 PH/25 Applied John Jetmore's -Mset patch.
1015 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1017 PH/27 In a string expansion for a processed (not raw) header when multiple
1018 headers of the same name were present, leading whitespace was being
1019 removed from all of them, but trailing whitespace was being removed only
1020 from the last one. Now trailing whitespace is removed from each header
1021 before concatenation. Completely empty headers in a concatenation (as
1022 before) are ignored.
1024 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1025 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1027 PH/29 [Removed. This was a change that I later backed out, and forgot to
1028 correct the ChangeLog entry (that I had efficiently created) before
1029 committing the later change.]
1031 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1032 address errors (4xx response to RCPT) over the same connection as other
1033 messages routed to the same hosts. Such deliveries are always "forced",
1034 so retry times are not inspected. This resulted in far too many retries
1035 for the affected addresses. The effect occurred only when there were more
1036 hosts than the hosts_max_try setting in the smtp transport when it had
1037 the 4xx errors. Those hosts that it had tried were not added to the list
1038 of hosts for which the message was waiting, so if all were tried, there
1039 was no problem. Two fixes have been applied:
1041 (i) If there are any address or message errors in an SMTP delivery, none
1042 of the hosts (tried or untried) are now added to the list of hosts
1043 for which the message is waiting, so the message should not be a
1044 candidate for sending over the same connection that was used for a
1045 successful delivery of some other message. This seems entirely
1046 reasonable: after all the message is NOT "waiting for some host".
1047 This is so "obvious" that I'm not sure why it wasn't done
1048 previously. Hope I haven't missed anything, but it can't do any
1049 harm, as the worst effect is to miss an optimization.
1051 (ii) If, despite (i), such a delivery is accidentally attempted, the
1052 routing retry time is respected, so at least it doesn't keep
1053 hammering the server.
1055 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1056 in ${readsocket because some servers need this prod.
1058 PH/32 Added some extra debug output when updating a wait-xxx database.
1060 PH/33 The hint "could be header name not terminated by colon", which has been
1061 given for certain expansion errors for a long time, was not being given
1062 for the ${if def:h_colon_omitted{... case.
1064 PH/34 The spec says: "With one important exception, whenever a domain list is
1065 being scanned, $domain contains the subject domain." There was at least
1066 one case where this was not true.
1068 PH/35 The error "getsockname() failed: connection reset by peer" was being
1069 written to the panic log as well as the main log, but it isn't really
1070 panic-worthy as it just means the connection died rather early on. I have
1071 removed the panic log writing for the ECONNRESET error when getsockname()
1074 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1075 runs only) independently of the message's sender address. This meant
1076 that, if the 4xx error was in fact related to the sender, a different
1077 message to the same recipient with a different sender could confuse
1078 things. In particualar, this can happen when sending to a greylisting
1079 server, but other circumstances could also provoke similar problems.
1080 I have changed the default so that the retry time for these errors is now
1081 based a combination of the sender and recipient addresses. This change
1082 can be overridden by setting address_retry_include_sender=false in the
1085 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1086 remote server are returned as part of bounce messages. This was not
1087 happening for LMTP over a pipe (the lmtp transport), but now it is the
1088 same for both kinds of LMTP.
1090 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1091 in header lines that were in fact CNAMEs. This is no longer the case.
1093 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1094 and queue runs started by the daemon processed all messages. This has
1095 been fixed so that -R and -S can now usefully be given with -q<time>.
1097 PH/40 Import PCRE release 6.7 (fixes some bugs).
1099 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1101 PH/42 Give an error if -q is specified more than once.
1103 PH/43 Renamed the variables $interface_address and $interface_port as
1104 $received_ip_address and $received_port, to make it clear that these
1105 values apply to message reception, and not to the outgoing interface when
1106 a message is delivered. (The old names remain recognized, of course.)
1108 PH/44 There was no timeout on the connect() call when using a Unix domain
1109 socket in the ${readsocket expansion. There now is.
1111 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1112 be meaningful with "accept".
1114 SC/01 Eximstats V1.43
1115 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1117 SC/02 Eximstats V1.44
1118 Use a glob alias rather than an array ref in the generated
1119 parser. This improves both readability and performance.
1121 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1122 Collect SpamAssassin and rejection statistics.
1123 Don't display local sender or destination tables unless
1124 there is data to show.
1125 Added average volumes into the top table text output.
1127 SC/04 Eximstats V1.46
1128 Collect data on the number of addresses (recipients)
1129 as well as the number of messages.
1131 SC/05 Eximstats V1.47
1132 Added 'Message too big' to the list of mail rejection
1133 reasons (thanks to Marco Gaiarin).
1135 SC/06 Eximstats V1.48
1136 Mainlog lines which have GMT offsets and are too short to
1137 have a flag are now skipped.
1139 SC/07 Eximstats V1.49 (Alain Williams)
1140 Added the -emptyok flag.
1142 SC/08 Eximstats V1.50
1143 Fixes for obtaining the IP address from reject messages.
1145 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1146 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1147 whitesspace changes from 4.64-PH/27
1149 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1152 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1153 are found, allow negative numbers in numeric criteria)
1155 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1157 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1158 to match changes made in 4.64-PH/43
1160 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1162 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1163 "domain" is actually an IP address, and force a failure. This locks out
1164 those revolvers/nameservers that support "A-for-A" lookups, in
1165 contravention of the specifications.
1167 PH/48 When a host name was looked up from an IP address, and the subsequent
1168 forward lookup of the name timed out, the host name was left in
1169 $sender_host_name, contrary to the specification.
1171 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1172 restricted to single-key lookups, Exim was not diagnosing an error if
1173 * or *@ was used with a query-style lookup.
1175 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1177 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1178 long ago, but noone interested enough thought of it. Let's just say that
1179 the "1.1" means that there are some new functions that weren't there at
1180 some point in the past.
1182 PH/51 Error processing for expansion failure of helo_data from an smtp
1183 transport during callout processing was broken.
1185 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1186 tested/used via the -bh/-bhc/-bs options.
1188 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1189 bug, fixed in subsequent PCRE releases).
1191 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1192 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1194 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1200 SC/01 Use a glob alias rather than an array ref in eximstats generated
1201 parser. This improves both readability and performance.
1203 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1204 Don't display local sender or destination tables in eximstats unless
1205 there is data to show.
1206 Added average volumes into the eximstats top table text output.
1208 SC/03 Collect data on the number of addresses (recipients) as well
1209 as the number of messages in eximstats.
1211 TF/01 Correct an error in the documentation for the redirect router. Exim
1212 does (usually) call initgroups() when daemonizing.
1214 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1215 with consistent privilege compared to when running as a daemon.
1217 TF/03 Note in the spec that $authenticated_id is not set for local
1218 submissions from trusted users.
1220 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1221 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1223 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1224 by adding some example configuration directives to the default
1225 configuration file. A little bit of work is required to uncomment the
1226 directives and define how usernames and passwords are checked, but
1227 there is now a framework to start from.
1229 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1230 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1231 without this. I don't know how relevant this is to other LDAP libraries.
1233 PH/02 Add the verb name to the "unknown ACL verb" error.
1235 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1237 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1239 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1240 directory not expanded when it should be if an expanded home directory
1241 was set for the address (which is overridden by the transport).
1243 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1246 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1247 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1248 because it is too late at that time, and has no effect.
1250 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1251 security issue with \' (bugzilla #107). I could not use the
1252 PQescapeStringConn() function, because it needs a PGconn value as one of
1255 PH/08 When testing addresses using -bt, indicate those final addresses that
1256 are duplicates that would not cause an additional delivery. At least one
1257 person was confused, thinking that -bt output corresponded to deliveries.
1258 (Suppressing duplicates isn't a good idea as you lose the information
1259 about possibly different redirections that led to the duplicates.)
1261 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1262 systems where poll() doesn't work, in particular OS X.
1264 PH/10 Added more information to debugging output for retry time not reached.
1266 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1267 operations in malware.c.
1269 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1272 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1273 syslog_duplication set false, log lines that would normally be written
1274 both the the main log and to the reject log were not written to syslog at
1277 PH/14 In the default configuration, change the use of "message" in ACL warn
1278 statements to "add_header".
1280 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1281 not followed by a command (e.g. "seen endif").
1283 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1284 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1287 PH/17 Added extra conditions to the default value of delay_warning_condition
1291 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1292 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1293 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1296 The Auto-Submitted: and various List- headers are standardised, whereas I
1297 don't think Precedence: ever was.
1299 PH/18 Refactored debugging code in route_finduser() to show more information,
1300 in particular, the error code if getpwnam() issues one.
1302 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1303 This is apparently needed in addition to the PH/07 change above to avoid
1304 any possible encoding problems.
1306 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1307 but not after initializing Perl.
1309 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1310 output them only if debugging. By default they are written stderr,
1311 apparently, which is not desirable.
1313 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1316 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1319 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1321 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1322 authenticated or an ident call has been made. Suppress the default
1323 values for $authenticated_id and $authenticated_sender (but permit -oMai
1324 and -oMas) when testing with -bh.
1326 PH/24 Re-jigged the order of the tests in the default configuration so that the
1327 tests for valid domains and recipients precede the DNS black list and CSA
1328 tests, on the grounds that those ones are more expensive.
1330 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1331 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1332 This bug exists in every version of Exim that I still have, right back to
1335 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1336 However, an attempt to turn on case-sensitivity in a regex key by
1337 including (?-i) didn't work because the subject string was already
1338 lowercased, and the effects were non-intuitive. It turns out that a
1339 one-line patch can be used to allow (?-i) to work as expected.
1345 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1346 other effects) broke the use of negated acl sub-conditions.
1348 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1351 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1352 "Deny" causes Exim to reject the incoming connection with a 554 error.
1353 Unfortunately, if there is a major crisis, such as a disk failure,
1354 tcp-wrappers gives "deny", whereas what one would like would be some
1355 kind of temporary error. A kludge has been added to help with this.
1356 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1357 554 error is used if errno is still zero or contains ENOENT (which occurs
1358 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1361 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1363 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1364 errors. Otherwise a message that provokes a temporary error (when other
1365 messages do not) can cause a whole host to time out.
1367 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1368 addresses were routed directly to files or pipes from a redirect router.
1369 File deliveries just didn't batch; pipe deliveries might have suffered
1372 PH/06 A failure to get a lock for a hints database would erroneously always say
1373 "Failed to get write lock", even when it was really a read lock.
1375 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1376 of 0600. This has been changed to use the value of the lockfile_mode
1377 option (which defaults to 0600).
1379 PH/08 Applied small patch from the Sieve maintainer.
1381 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1382 folder from quota calculations, a direct delivery into this folder messed
1383 up the contents of the maildirsize file. This was because the regex was
1384 used only to exclude .Trash (or whatever) when the size of the mailbox
1385 was calculated. There was no check that a delivery was happening into an
1386 excluded directory. This bug has been fixed by ignoring all quota
1387 processing for deliveries into excluded directories.
1389 PH/10 Added the maildirfolder_create_regex option to appendfile.
1395 PH/01 The code for finding all the local interface addresses on a FreeBSD
1396 system running IPv6 was broken. This may well have applied to all BSD
1397 systems, as well as to others that have similar system calls. The broken
1398 code found IPv4 interfaces correctly, but gave incorrect values for the
1399 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1400 that it would not match correctly against @[] and not recognize the IPv6
1403 PH/02 The ipliteral router was not recognizing addresses of the form user@
1404 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1406 PH/03 Added disable_ipv6.
1408 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1409 decoded form, because it is most often used to construct To: headers
1410 lines in autoreplies, and the decoded form may well be syntactically
1411 invalid. However, $reply_address has leading white space removed, and all
1412 newlines turned into spaces so that the autoreply transport does not
1415 PH/05 If group was specified without a user on a router, and no group or user
1416 was specified on a transport, the group from the router was ignored.
1418 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1419 for visible compile-time settings that can be used to change these
1420 numbers, for those that want even more. Backwards compatibility with old
1421 spool files has been maintained. However, going back to a previous Exim
1422 release will lost any variables that are in spool files.
1424 PH/07 Two small changes when running in the test harness: increase delay when
1425 passing a TCP/IP connection to a new process, in case the original
1426 process has to generate a bounce, and remove special handling of
1427 127.0.0.2 (sic), which is no longer necessary.
1429 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1430 be the same on different OS.
1432 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1435 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1436 whether --show-vars was specified or not
1438 JJ/02 exipick: Added support for new ACL variable spool format introduced
1441 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1442 syntactically invalid From: or Reply-to: line, and a filter used this to
1443 generate an autoreply, and therefore failed to obtain an address for the
1444 autoreply, Exim could try to deliver to a non-existent relative file
1445 name, causing unrelated and misleading errors. What now happens is that
1446 it logs this as a hard delivery error, but does not attempt to create a
1449 PH/11 The exinext utility has a -C option for testing purposes, but although
1450 the given file was scanned by exinext itself; it wasn't being passed on
1451 when Exim was called.
1453 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1454 an end-of-file indication when reading a command response.
1456 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1457 compiled. In many other places in Exim, IPv6 addresses are always
1458 recognized, so I have changed this. It also means that IPv4 domain
1459 literals of the form [IPV4:n.n.n.n] are now always recognized.
1461 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1462 used if the router is not running as root, for example, when verifying at
1463 ACL time, or when using -bh. The debugging output from this situation was
1464 non-existent - all you got was a failure to exec. I have made two
1467 (a) Failures to set uid/gid, the current directory, or a process leader
1468 in a subprocess such as that created by queryprogram now generate
1469 suitable debugging ouput when -d is set.
1471 (b) The queryprogram router detects when it is not running as root,
1472 outputs suitable debugging information if -d is set, and then runs
1473 the subprocess without attempting to change uid/gid.
1475 PH/15 Minor change to Makefile for building test_host (undocumented testing
1478 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1479 additional section of a DNS packet that returns MX or SRV records.
1480 Instead, it always explicitly searches for A/AAAA records. This avoids
1481 major problems that occur when a DNS server includes only records of one
1482 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1483 fixed another bug: if SRV records were looked up and the corresponding
1484 address records were *not* found in the additional section, the port
1485 values from the SRV records were lost.
1487 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1488 using the correct key (the original address) when searching the retry
1489 rules in order to find which one to use for generating the retry hint.
1491 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1492 adding the default one. Similarly, if it contains a Reply-To: header, the
1493 errors_reply_to option, if set, is not used.
1495 PH/19 When calculating a retry time, Exim used to measure the "time since
1496 failure" by looking at the "first failed" field in the retry record. Now
1497 it does not use this if it is later than than the arrival time of the
1498 message. Instead it uses the arrival time. This makes for better
1499 behaviour in cases where some deliveries succeed, thus re-setting the
1500 "first failed" field. An example is a quota failure for a huge message
1501 when small messages continue to be delivered. Without this change, the
1502 "time since failure" will always be short, possible causing more frequent
1503 delivery attempts for the huge message than are intended.
1504 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1506 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1507 $1, $2, $3) because the numerical variables can be reset during some
1508 expansion items (e.g. "match"), thereby losing the authentication data.
1510 PH/21 Make -bV show the size of off_t variables so that the test suite can
1511 decide whether to run tests for quotas > 2G.
1513 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1514 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1515 filecount value is greater than 2G or if a quota value is greater than 2G
1516 on a system where the size of off_t is not greater than 4, a panic error
1519 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1520 never match. The debug and -bh output now contains an explicit error
1521 message indicating a malformed IPv4 address or mask.
1523 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1524 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1525 PH/23 above applies.
1527 PH/25 Do not write to syslog when running in the test harness. The only
1528 occasion when this arises is a failure to open the main or panic logs
1529 (for which there is an explicit test).
1531 PH/26 Added the /no_tell option to "control=freeze".
1533 PH/27 If a host name lookup failed very early in a connection, for example, if
1534 the IP address matched host_lookup and the reverse lookup yielded a name
1535 that did not have a forward lookup, an error message of the form "no IP
1536 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1537 could be logged. Now it outputs the IP address instead of "NULL".
1539 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1540 allows the sender and the authenticated sender to be set when
1541 submitting a message from within Exim. Since child_open_exim() is
1542 documented for local_scan(), the new function should be too.
1544 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1545 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1546 results in an empty string is now treated as unset.
1548 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1550 PH/31 Added sender_verify_fail logging option.
1552 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1553 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1554 tidied the source and removed it altogether.
1556 PH/33 When a queue run was abandoned because the load average was too high, a
1557 log line was always written; now it is written only if the queue_run log
1558 selector is set. In addition, the log line for abandonment now contains
1559 information about the queue run such as the pid. This is always present
1560 in "start" and "stop" lines but was omitted from the "abandon" line.
1562 PH/34 Omit spaces between a header name and the colon in the error message that
1563 is given when verify = headers_syntax fails (if there are lots of them,
1564 the message gets confusing).
1566 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1567 names, as there are now some PTR records that contain slashes. This check
1568 is only to protect against broken name servers that fall over on strange
1569 characters, so the fact that it applies to all lookups doesn't matter.
1571 PH/36 Now that the new test suite is complete, we can remove some of the
1572 special code in Exim that was needed for the old test suite. For example,
1573 sorting DNS records because real resolvers return them in an arbitrary
1574 order. The new test suite's fake resolver always returns records in the
1577 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1578 bounces) except when queue_only is set, to avoid logging races between
1579 the different processes.
1581 PH/38 Panic-die if .include specifies a non-absolute path.
1583 PH/39 A tweak to the "H" retry rule from its user.
1585 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1586 a label. They prevented compilation on older perls.
1588 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1589 a warning to be raised on newish perls.
1591 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1592 on queue. Changes to match documented behaviour of showing count of
1593 messages matching specified criteria.
1595 PH/40 Changed the default ident timeout from 30s to 5s.
1597 PH/41 Added support for the use of login_cap features, on those BSD systems
1598 that have them, for controlling the resources used by pipe deliveries.
1600 PH/42 The content-scanning code uses fopen() to create files in which to put
1601 message data. Previously it was not paying any attention to the mode of
1602 the files. Exim runs with umask(0) because the rest of the code creates
1603 files with open(), and sets the required mode explicitly. Thus, these
1604 files were ending up world-writeable. This was not a big issue, because,
1605 being within the spool directory, they were not world-accessible. I have
1606 created a function called modefopen, which takes an additional mode
1607 argument. It sets umask(777), creates the file, chmods it to the required
1608 mode, then resets the umask. All the relevant calls to fopen() in the
1609 content scanning code have been changed to use this function.
1611 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1612 to 24 hours. This avoids potential overflow problems when processing G
1613 and H retry rules. I suspect nobody ever tinkers with this value.
1615 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1617 PH/45 When the plaintext authenticator is running as a client, the server's
1618 challenges are checked to ensure they are valid base64 strings. By
1619 default, the authentication attempt is cancelled if an invalid string is
1620 received. Setting client_ignore_invalid_base64 true ignores these errors.
1621 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1622 they are received. Thus, the responses can be made to depend on the
1623 challenges. If an invalid string is ignored, an empty string is placed in
1626 PH/46 Messages that are created by the autoreply transport now contains a
1627 References: header, in accordance with RFCs 2822 and 3834.
1629 PH/47 Added authenticated_sender_force to the smtp transport.
1631 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1633 PH/49 Installed latest patch from the Sieve maintainer.
1635 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1636 was also set, the appendfile transport was unnecessarily scanning a
1637 directory of message files (e.g. for maildir delivery) to find the count
1638 of files (along with the size), even though it did not need this
1639 information. It now does the scan only if it needs to find either the
1640 size of the count of files.
1642 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1644 PH/52 Two bugs concerned with error handling when the smtp transport is
1647 (i) Exim was not creating retry information for temporary errors given
1648 for individual recipients after the DATA command when the smtp transport
1649 was used in LMTP mode. This meant that they could be retried too
1650 frequently, and not timed out correctly.
1652 (ii) Exim was setting the flag that allows error details to be returned
1653 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1654 recipients that were returned after the DATA command.
1656 PH/53 This is related to PH/52, but is more general: for any failing address,
1657 when detailed error information was permitted to be returned to the
1658 sender, but the error was temporary, then after the final timeout, only
1659 "retry timeout exceeded" was returned. Now it returns the full error as
1660 well as "retry timeout exceeded".
1662 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1663 do this, and (what is worse) MTAs that accept it.
1665 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1666 will now be deprecated.
1668 PH/56 New os.c-cygwin from the Cygwin maintainer.
1670 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1671 formats (previously only available in exim formats via -bpr, -bpru,
1672 and -bpra. Now also available in native and exiqgrep formats)
1674 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1675 with very large, slow to parse queues
1677 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1679 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1681 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1682 responses to authentication challenges, though it was showing the
1683 challenges; (ii) I've removed the CR characters from the debug output for
1686 PH/58 Allow for the insertion of a newline as well as a space when a string
1687 is turned into more than one encoded-word during RFC 2047 encoding. The
1688 Sieve code now uses this.
1690 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1691 data_4xx, lost_connection, tls_required.
1693 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1694 message was being sent as an SMTP response.
1696 PH/61 Add -l and -k options to exicyclog.
1698 PH/62 When verifying, if an address was redirected to one new address, so that
1699 verification continued, and the new address failed or deferred after
1700 having set something in $address_data, the value of $address_data was not
1701 passed back to the ACL. This was different to the case when no
1702 redirection occurred. The value is now passed back in both cases.
1704 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1705 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1706 use login_cap.h, so on its own it isn't the distinguishing feature. The
1707 new name refers directly to the setclassresources() function.
1709 PH/65 Added configuration files for NetBSD3.
1711 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1713 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1714 is preferred over IPv4.
1716 PH/68 The bounce_return_message and bounce_return_body options were not being
1717 honoured for bounces generated during the reception of non-SMTP messages.
1718 In particular, this applied to messages rejected by the ACL. This bug has
1719 been fixed. However, if bounce_return_message is true and bounce_return_
1720 body is false, the headers that are returned for a non-SMTP message
1721 include only those that have been read before the error was detected.
1722 (In the case of an ACL rejection, they have all been read.)
1724 PH/69 The HTML version of the specification is now built in a directory called
1725 spec_html instead of spec.html, because the latter looks like a path with
1726 a MIME-type, and this confuses some software.
1728 PH/70 Catch two compiler warnings in sieve.c.
1730 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1731 function verify_get_ident() calls ip_connect() to connect a socket, but
1732 if the "connect()" function timed out, ip_connect() used to close the
1733 socket. However, verify_get_ident() also closes the socket later, and in
1734 between Exim writes to the log, which may get opened at this point. When
1735 the socket was closed in ip_connect(), the log could get the same file
1736 descriptor number as the socket. This naturally causes chaos. The fix is
1737 not to close the socket in ip_connect(); the socket should be closed by
1738 the function that creates it. There was only one place in the code where
1739 this was missing, in the iplookup router, which I don't think anybody now
1740 uses, but I've fixed it anyway.
1742 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1743 well as to direct DNS lookups. Otherwise the handling of names in host
1744 lists is inconsistent and therefore confusing.
1750 PH/01 Two changes to the default runtime configuration:
1752 (1) Move the checks for relay_from_hosts and authenticated clients from
1753 after to before the (commented out) DNS black list checks.
1755 (2) Add control=submission to the relay_from_hosts and authenticated
1756 clients checks, on the grounds that messages accepted by these
1757 statements are most likely to be submissions.
1759 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1761 (1) Generate an error if the third argument for the ${prvs expansion is
1764 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1767 (3) Reset the variables that are obtained from the first argument of
1768 ${prvscheck and used in the second argument before leaving the code,
1769 because their memory is reclaimed, so using them afterwards may do
1772 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1773 one (it's much easier than Tom thought :-).
1775 (5) Because of (4), we can now allow for the use of $prvscheck_result
1776 inside the third argument.
1778 PH/03 For some reason, the default setting of PATH when running a command from
1779 a pipe transport was just "/usr/bin". I have changed it to
1782 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1783 anything to be listed in the output from -bV.
1785 PH/05 When a filter generated an autoreply, the entire To: header line was
1786 quoted in the delivery log line, like this:
1788 => >A.N.Other <ano@some.domain> <original@ddress> ...
1790 This has been changed so that it extracts the operative address. There
1791 may be more than one such address. If so, they are comma-separated, like
1794 => >ano@some.domain,ona@other.domain <original@ddress> ...
1796 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1797 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1798 not being looked up in the rDNS to get a host name, Exim was showing the
1799 IP address twice in Received: lines, even though the IP addresses were
1800 identical. For example:
1802 Received: from [1.2.3.4] (helo=[1.2.3.4])
1804 However, if the real host name was known, it was omitting the HELO data
1805 if it matched the actual IP address. This has been tidied up so that it
1806 doesn't show the same IP address twice.
1808 PH/07 When both +timestamp and +memory debugging was on, the value given by
1809 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1810 called by the debug printing, thereby overwriting the timestamp buffer.
1811 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1813 PH/08 When the original message was included in an autoreply transport, it
1814 always said "this is a copy of the message, including all the headers",
1815 even if body_only or headers_only was set. It now gives an appropriate
1818 PH/09 Applied a patch from the Sieve maintainer which:
1820 o fixes some comments
1821 o adds the (disabled) notify extension core
1822 o adds some debug output for the result of if/elsif tests
1823 o points to the current vacation draft in the documentation
1824 and documents the missing references header update
1828 o fixes a bug in processing the envelope test (when testing
1829 multiple envelope elements, the last element determinted the
1832 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1833 Electronic Mail") by including:
1835 Auto-submitted: auto-generated
1837 in the messages that it generates (bounce messages and others, such as
1838 warnings). In the case of bounce messages for non-SMTP mesages, there was
1839 also a typo: it was using "Auto_submitted" (underscore instead of
1840 hyphen). Since every message generated by Exim is necessarily in response
1841 to another message, thes have all been changed to:
1843 Auto-Submitted: auto-replied
1845 in accordance with these statements in the RFC:
1847 The auto-replied keyword:
1849 - SHOULD be used on messages sent in direct response to another
1850 message by an automatic process,
1852 - MUST NOT be used on manually-generated messages,
1854 - MAY be used on Delivery Status Notifications (DSNs) and Message
1855 Disposition Notifications (MDNs),
1857 - MUST NOT be used on messages generated by automatic or periodic
1858 processes, except for messages which are automatic responses to
1861 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1862 to the default Received: header definition.
1864 PH/12 Added log selector acl_warn_skipped (default on).
1866 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1867 variables because (a) they are in the wrong storage pool and (b) even if
1868 they were copied, it wouldn't work properly because of the caching.
1870 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1871 checking when decoding. Apparently there are clients that generate
1872 overlong encoded strings. Why am I not surprised?
1874 PH/15 If the first argument of "${if match_address" was not empty, but did not
1875 contain an "@" character, Exim crashed. Now it writes a panic log message
1876 and treats the condition as false.
1878 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1880 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1881 extension "envelope-auth". The code is finished and in agreement with
1882 other implementations, but there is no documentation so far and in fact,
1883 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1884 not changing the active code.
1886 Print executed "if" and "elsif" statements when debugging is used. This
1887 helps a great deal to understand what a filter does.
1889 Document more things not specified clearly in RFC3028. I had all this
1890 sorted out, when out of a sudden new issues came to my mind. Oops."
1892 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1895 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1896 canonical form (as documented). However, after a host name lookup from
1897 the IP address, check_host() was doing a simple string comparison with
1898 addresses acquired from the DNS when checking that the found name did
1899 have the original IP as one of its addresses. Since any found IPv6
1900 addresses are likely to be in abbreviated form, the comparison could
1901 fail. Luckily, there already exists a function for doing the comparison
1902 by converting both addresses to binary, so now that is used instead of
1903 the text comparison.
1905 PH/20 There was another similar case to PH/19, when a complete host name was
1906 given in a host list; looking up its IP address could give an abbreviated
1907 form, whereas the current host's name might or might not be abbreviated.
1908 The same fix has been applied.
1914 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1915 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1918 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1919 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1921 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1923 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1924 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1925 Also, for "Auto-Submitted", treat anything other than "no" as
1926 non-personal, in accordance with RFC 3834. (Previously it treated
1927 anything starting "auto-" as non-personal.)
1929 TF/01 The control=submission/name=... option had a problem with syntax
1930 errors if the name included a slash character. The /name= option
1931 now slurps the rest of the string, so it can include any characters
1932 but it must come last in the list of options (after /sender_retain
1935 PH/05 Some modifications to the interface to the fake nameserver for the new
1943 TK/01 Added the "success_on_redirect" address verification option. See
1944 NewStuff for rationale and an example.
1946 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1948 PH/02 Patch to exigrep to allow it to work on syslog lines.
1950 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1951 fread() to skip over the body file's header line, because in Cygwin the
1952 header line is locked and is inaccessible.
1954 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1955 co-exist for some time) to make it clear that it is the Exim ID that is
1956 referenced, not the Message-ID: header line.
1958 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1959 string_format() function, because snprintf() does not exist on all
1962 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1963 ${stat: expansion item.
1965 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1966 protocol synchronization error", to keep the pedants happy.
1968 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1969 well as for IRIX systems, when gcc is being used. See the host.c source
1972 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1974 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1977 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1978 transport and to the smtp transport in LMTP mode.
1980 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1982 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1983 run a filter in a subprocess. This could lead to confusion in subsequent
1984 lookups in the parent process. There should also be a search_tidyup() at
1985 the end of the subprocess.
1987 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1988 only if the host matched helo_try_verify_hosts, which caused the
1989 verification to occur when the EHLO/HELO command was issued. The ACL just
1990 tested the remembered result. Now, if a previous verification attempt has
1991 not happened, "verify = helo" does it there and then.
1993 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1995 TK/03 Fix log output including CR from clamd.
1997 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1998 exist provoked a memory error which could cause a segfault.
2000 PH/15 Installed PCRE 6.2
2002 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2004 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2005 of the problem. Specifically, suggested +O2 rather than +O1 for the
2008 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2010 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2011 record, the port was not correctly logged when the outgoing_port log
2012 selector was set (it logged the transort's default port).
2014 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2015 fallback_hosts, and "hosts" in the smtp transport.
2017 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2018 host errors such as "Connection refused".
2020 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2021 authentication with radiusclient 0.4.9:
2023 - Error returned from rc_read_config was caught wrongly
2024 - Username/password not passed on to radius server due to wrong length.
2026 The presumption is that some radiusclient API changes for 4.51/PH/17
2027 were not taken care of correctly. The code is still untested by me (my
2028 Linux distribution still has 0.3.2 of radiusclient), but it was
2029 contributed by a Radius user.
2031 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2032 expanding the "port" option of the smtp transport.
2034 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2035 while reading a MIME header. Thanks to Tom Hughes for a patch.
2037 PH/24 Include config.h inside local_scan.h so that configuration settings are
2040 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2041 that in an ACL for RCPT (for example), you can examine exactly what was
2044 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2045 commands, but it was not correctly comparing the address with the actual
2046 client host address. Thus, it would show the EHLO address in Received:
2047 header lines when this was not necessary.
2049 PH/27 Added the % operator to ${eval:}.
2051 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2052 it should be ignoring failures (because with -C, for example, it has lost
2053 privilege). It wasn't ignoring creation failures other than "already
2056 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2059 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2060 because an input line was too long, either on its own, or by virtue of
2061 too many continuations, the temporary file was not being removed, and the
2062 return code was incorrect.
2064 PH/31 Missing "BOOL" in function definition in filtertest.c.
2066 PH/32 Applied Sieve patches from the maintainer.
2068 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2070 PH/33 Added "verify = not_blind".
2072 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2073 Local/Makefile (with some defaults set). These are used in built scripts
2074 such as exicyclog, but they have never been used in the exim_install
2075 script (though there are many overriding facilities there). I have
2076 arranged that the exim_install script now takes note of these two
2079 PH/35 Installed configuration files for Dragonfly.
2081 PH/36 When a locally submitted message by a trusted user did not contain a
2082 From: header, and the sender address was obtained from -f or from an SMTP
2083 MAIL command, and the trusted user did not use -F to supply a sender
2084 name, $originator_name was incorrectly used when constructing a From:
2085 header. Furthermore, $originator_name was used for submission mode
2086 messages from external hosts without From: headers in a similar way,
2087 which is clearly wrong.
2089 PH/37 Added control=suppress_local_fixups.
2091 PH/38 When log_selector = +received_sender was set, and the addition of the
2092 sender made the log line's construction buffer exactly full, or one byte
2093 less than full, an overflow happened when the terminating "\n" was
2096 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2097 when the result of a list match is failure because a DNS lookup failed.
2099 PH/40 RM_COMMAND is now used in the building process.
2101 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2102 the "build-* directories that it finds.
2104 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2105 address in a domain literal was a prefix of an interface address.
2107 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2108 when verifying a sender address, unless rewrite_headers is false.
2110 PH/44 Wrote a long comment about why errors_to addresses are verified as
2111 recipients, not senders.
2113 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2114 the ratelimit ACL was added.
2116 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2118 PH/46 Added extra information about PostgreSQL errors to the error string.
2120 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2121 avoiding the need to install special zones in a real server. This is
2122 backwards compatible; if it can't find the fake resolver, it drops back.
2123 Thus, both old and new test suites can be run.
2125 TF/02 Added util/ratelimit.pl
2127 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2128 clock is set back in time.
2130 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2131 Candler <B.Candler@pobox.com>.
2133 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2134 OK for addresses that are the result of redirections.
2136 PH/48 A number of further additions for the benefit of the new test suite,
2137 including a fake gethostbyname() that interfaces to the fake DNS resolver
2140 TF/06 The fix for widen_domains has also been applied to qualify_single and
2141 search_parents which are the other dnslookup options that can cause
2144 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2147 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2149 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2150 service" if it encountered an MX record with an empty target hostname.
2151 The message is now "an MX or SRV record indicated no SMTP service".
2153 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2154 if the DNS of the sending site is misconfigured. This is quite a
2155 common situation. This change restores the behaviour of treating a
2156 helo verification defer as a failure.
2158 PH/51 If self=fail was set on a router, the bounce message did not include the
2159 actual error message.
2165 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2167 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2168 command itself ended in error, the underlying message about the transport
2169 filter timeout was being overwritten with the pipe command error. Now the
2170 underlying error message should be appended to the second error message.
2172 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2174 PH/02 Reduce the amount of output that "make" produces by default. Full output
2175 can still be requested.
2177 PH/03 The warning log line about a condition test deferring for a "warn" verb
2178 was being output only once per connection, rather than after each
2179 occurrence (because it was using the same function as for successful
2180 "warn" verbs). This seems wrong, so I have changed it.
2182 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2183 it should not have, which might have caused a crash in the right
2184 circumstances, but probably never did.
2186 PH/04 Installed a modified version of Tony Finch's patch to make submission
2187 mode fix the return path as well as the Sender: header line, and to
2188 add a /name= option so that you can make the user's friendly name appear
2191 TF/03 Added the control = fakedefer ACL modifier.
2193 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2194 Mark Lowes for thorough testing.
2196 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2198 TK/03 Merged latest SRS patch from Miles Wilton.
2200 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2201 with the definition in sysexits.h (which is #included earlier).
2202 Fortunately, Exim does not actually use EX_OK. The code used to try to
2203 preserve the sysexits.h value, by assumimg that macro definitions were
2204 scanned for macro replacements. I have been disabused of this notion,
2205 so now the code just undefines EX_OK before #including unistd.h.
2207 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2208 in the smtp transport. When a block could not be written in a single
2209 write() function, the timeout was being re-applied to each part-write.
2210 This seems wrong - if the receiver was accepting one byte at a time it
2211 would take for ever. The timeout is now adjusted when this happens. It
2212 doesn't have to be particularly precise.
2214 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2215 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2217 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2218 without a domain if the check to <postmaster@domain> fails.
2220 SC/01 Eximstats: added -xls and the ability to specify output files
2221 (patch written by Frank Heydlauf).
2223 SC/02 Eximstats: use FileHandles for outputing results.
2225 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2227 SC/04 Eximstats: fixed display of large numbers with -nvr option
2229 SC/05 Eximstats: fixed merging of reports with empty tables.
2231 SC/06 Eximstats: added the -include_original_destination flag
2233 SC/07 Eximstats: removed tabs and trailing whitespace.
2235 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2237 TK/06 MBOX spool code: Add real "From " MBOX separator line
2238 so the .eml file is really in mbox format (even though
2239 most programs do not really care). Patch from Alex Miller.
2241 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2242 The latter is generated from $received_to and is only set if the
2243 message has one envelope recipient. SA can use these headers,
2244 obviously out-of-the-box. Patch from Alex Miller.
2246 PH/08 The ${def test on a variable was returning false if the variable's
2247 value was "0", contrary to what the specification has always said!
2248 The result should be true unless the variable is empty.
2250 PH/09 The syntax error of a character other than { following "${if
2251 def:variable_name" (after optional whitespace) was not being diagnosed.
2252 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2253 accidental colon was present, for example, could give incorrect results.
2255 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2256 result is used (not including appendfile, where other changes are about
2259 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2260 This involved changing a lot of size variables from int to off_t. It
2261 should work with maildirs and everything.
2263 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2264 spamd dying while we are connected to it.
2266 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2269 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2270 function for BDB 4.3.
2272 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2274 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2275 the sources for PCRE have been re-arranged and more files are now
2278 PH/15 The code I had for printing potentially long long variables in PH/11
2279 above was not the best (it lost precision). The length of off_t variables
2280 is now inspected at build time, and an appropriate printing format (%ld
2281 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2282 to be "long long int" or "long int". This is needed for the internal
2283 formatting function string_vformat().
2285 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2286 the configuration file to be ":syslog", then the script "guesses" where
2287 the logs files are, rather than using the compiled in default. In our
2288 case the guess is not the same as the compiled default, so the script
2289 suddenly stopped working when I started to use syslog. The patch checks
2290 to see if log_file_path is "". If so, it attempts to read it from exim
2291 with no configuration file to get the compiled in version, before it
2292 falls back to the previous guessing code."
2294 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2295 implementing BATV in an Exim configuration. See NewStuff for the gory
2298 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2299 Makefile that are specific to HP-UX.
2301 PH/18 If the "use_postmaster" option was set for a recipient callout together
2302 with the "random" option, the postmaster address was used as the MAIL
2303 FROM address for the random test, but not for the subsequent recipient
2304 test. It is now used for both.
2306 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2307 patch removes a few documentation additions to RFC 3028, because the
2308 latest draft now contains them. It adds the new en;ascii-case comparator
2309 and a new error check for 8bit text in MIME parts. Comparator and
2310 require names are now matched exactly. I enabled the subaddress
2311 extension, but it is not well tested yet (read: it works for me)."
2313 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2314 rework some of the code of TK/09 above to avoid the hardwired use of
2315 "%lld" and "long long". Replaced the call to snprintf() with a call to
2318 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2319 records point to non-existent hosts", "retry timeout exceeded", and
2320 "retry time not reached for any host after a long failure period".
2322 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2323 experimental DomainKeys support:
2325 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2326 (2) On an error such as an illegally used "control", the wrong name for
2327 the control was given.
2329 These problems did NOT occur unless DomainKeys support was compiled.
2331 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2333 PH/24 Added ${if match_ip condition.
2335 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2336 databases so that it will be absolutely obvious if a crash occurs in the
2337 DB library. This is a regular occurrence (often caused by mis-matched
2340 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2341 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2342 picked up on a user's system that detects such things. There doesn't seem
2343 to be a gcc warning option for this - only an attribute that has to be
2344 put on the function's prototype. It seems that in Fedora Core 4 they have
2345 set this on a number of new functions. No doubt there will be more in due
2348 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2349 specify a transport. However, if an address that was verified by such a
2350 router was the subject of a callout, Exim crashed because it tried to
2351 read the rcpt_include_affixes from the non-existent transport. Now it
2352 just assumes that the setting of that option is false. This bug was
2353 introduced by 4.51/PH/31.
2355 PH/28 Changed -d+all to exclude +memory, because that information is very
2356 rarely of interest, but it makes the output a lot bigger. People tend to
2357 do -d+all out of habit.
2359 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2360 code in os-type was giving problems when libc.so lives in lib64, like on
2363 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2364 aren't the modern standard, and it seems that some systems' include files
2365 don't always have them. Exim was already checking for some of the newer
2366 ones like T_AAAA, and defining it itself. I've added checks for all the
2367 record types that Exim uses.
2369 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2370 not automatically generating a new one, as it is supposed to. This
2371 prevented TLS from working. If the file did exist, but contained invalid
2372 data, a new version was generated, as expected. It was only the case of a
2373 non-existent file that was broken.
2375 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2376 with a change in libdomainkeys > 0.64.
2378 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2379 from DNS. If the selector record carries the flag, it now has
2380 precedence over the domain-wide flag.
2382 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2384 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2385 the use of a port name that isn't defined in /etc/services) occurred, the
2386 message was deferred as in a normal delivery, and thus remained on the
2387 spool, instead of being failed because of the mua_wrapper setting. This
2388 is now fixed, and I tidied up some of the mua_wrapper messages at the
2391 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2392 the messages in a hash of arrays rather than using individual hashes.
2393 This is a bit cleaner and results in dramatic memory savings, albeit
2394 at a slight CPU cost.
2396 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2397 as requested by Marc Sherman.
2399 SC/10 Eximstats: added histograms for user specified patterns as requested
2402 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2404 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2405 fopen() in the content-scanning modules that did not already have it.
2411 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2412 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2414 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2416 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2418 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2419 to test Sieve filters that use "vacation".
2421 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2422 that changes the way the GnuTLS parameters are stored in the cache file.
2423 The new format can be generated externally. For backward compatibility,
2424 if the data in the cache doesn't make sense, Exim assumes it has read an
2425 old-format file, and it generates new data and writes a new file. This
2426 means that you can't go back to an older release without removing the
2429 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2430 work if there are any delivery delays because "one_time" forces the
2431 parent to be marked "delivered", so its unseen clone is never tried
2432 again. For this reason, Exim now forbids the simultaneous setting of
2435 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2436 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2437 entry if you want to know the details. The fix, however, neglected to
2438 consider the case when local delivery batching is involved. The test for
2439 "previously delivered" was not happening when checking to see if an
2440 address could be batched with a previous (undelivered) one; under
2441 certain circumstances this could lead to multiple deliveries to the same
2444 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2445 in its include files, and this causes problems building Exim.
2447 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2448 header_syntax) but Exim was just ignoring anything given after a slash.
2449 In particular, this caused confusion with an attempt to use "verify =
2450 reverse_host_lookup/defer_ok". An error is now given when options are
2451 supplied for verify items that do not have them. (Maybe reverse_host_
2452 lookup should have a defer_ok option, but that's a different point.)
2454 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2455 defined by RFC 821) to 2048, because there were problems with some AUTH
2456 commands, and RFC 1869 says the size should be increased for extended
2457 SMTP commands that take arguments.
2459 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2462 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2463 "unknown" error; now it says that the functionality isn't in the binary.
2465 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2466 an address' error message when a string expansion fails (syntax or
2467 whatever). Otherwise the password may appear in the log. Following change
2468 PH/42 below, there is no longer a chance of it appearing in a bounce
2471 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2473 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2474 of its addresses was ever tried. (Bugzilla bug #2.)
2476 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2477 the result incorrectly in the debug output. (It correctly added a newline
2478 to what was transported.)
2480 TF/01 Added $received_time.
2482 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2483 commented out examples of how to interface to a virus scanner and to
2484 SpamAssassin. Also added commented examples of av_scanner and
2485 spamd_address settings.
2487 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2488 and controls are allowed in which ACLs. There were a couple of minor
2489 errors. Some of the entries in the conditions table (which is a table of
2490 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2491 negation of where the condition IS allowed.
2493 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2495 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2496 header file does not have a version number, so I've had to invent a new
2497 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2498 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2499 radiusclient), but it was contributed by a Radius user.
2501 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2502 files or directories, for OpenSSL.
2504 PH/19 When an Exim process that is running as root has to create an Exim log
2505 file, it does so in a subprocess that runs as exim:exim so as to get the
2506 ownership right at creation (otherwise, other Exim processes might see
2507 the file with the wrong ownership). There was no test for failure of this
2508 fork() call, which would lead to the process getting stuck as it waited
2509 for a non-existent subprocess. Forks do occasionally fail when resources
2510 run out. I reviewed all the other calls to fork(); they all seem to check
2513 PH/20 When checking for unexpected SMTP input at connect time (before writing
2514 the banner), Exim was not dealing correctly with a non-positive return
2515 from the read() function. If the client had disconnected by this time,
2516 the result was a log entry for a synchronization error with an empty
2517 string after "input=" when read() returned zero. If read() returned -1
2518 (an event I could not check), uninitialized data bytes were printed.
2519 There were reports of junk text (parts of files, etc) appearing after
2522 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2524 PH/22 Added support for macro redefinition, and (re)definition in between
2525 driver and ACL definitions.
2527 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2528 forgetting to use the resulting value; it was using the unexpanded value.
2530 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2531 hadn't been configured. The fix is from Juergen Kreileder, who
2532 understands it better than I do:
2534 "Here's what I see happening with three configured cyrus_sasl
2535 authenticators configured (plain, login, cram-md5):
2537 On startup auth_cyrus_sasl_init() gets called for each of these.
2538 This means three calls to sasl_listmech() without a specified mech_list.
2539 => SASL tests which mechs of all available mechs actually work
2540 => three warnings about OTP not working
2541 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2543 With the patch, sasl_listmech() also gets called three times. But now
2544 SASL's mech_list option is set to the server_mech specified in the the
2545 authenticator. Or in other words, the answer from sasl_listmech()
2546 gets limited to just the mech you're testing for (which is different
2548 => the return list contains just 'plain' or 'login', 'cram-md5' or
2549 nothing depending on the value of ob->server_mech.
2551 I've just tested the patch: Authentication still works fine,
2552 unavailable mechs specified in the exim configuration are still
2553 caught, and the auth.log warnings about OTP are gone."
2555 PH/25 When debugging is enabled, the contents of the command line are added
2556 to the debugging output, even when log_selector=+arguments is not
2559 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2560 answer is "GNU", and only if the return is "GNU/something" is the answer
2563 PH/27 $acl_verify_message is now set immediately after the failure of a
2564 verification in an ACL, and so is available in subsequent modifiers. In
2565 particular, the message can be preserved by coding like this:
2567 warn !verify = sender
2568 set acl_m0 = $acl_verify_message
2570 Previously, $acl_verify_message was set only while expanding "message"
2571 and "log_message" when a very denied access.
2573 PH/28 Modified OS/os.c-Linux with
2575 -#ifndef OS_LOAD_AVERAGE
2576 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2578 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2579 nomenclature these days.)
2581 PH/29 Installed patch from the Sieve maintainer that adds the options
2582 sieve_useraddress and sieve_subaddress to the redirect router.
2584 PH/30 In these circumstances:
2585 . Two addresses routed to the same list of hosts;
2586 . First host does not offer TLS;
2587 . First host accepts first address;
2588 . First host gives temporary error to second address;
2589 . Second host offers TLS and a TLS session is established;
2590 . Second host accepts second address.
2591 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2592 and peerdn, if requested) that were in fact used only for the second
2595 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2596 attention to any local part prefix or suffix that was matched by the
2597 router that accepted the address. It now behaves in the same way as it
2598 does for delivery: the affixes are removed from the local part unless
2599 rcpt_include_affixes is set on the transport.
2601 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2602 timeout during the DATA phase of an incoming message.
2604 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2605 applied a patch sanctioned by the Sieve maintainer.
2607 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2608 the uid or gid is negative. A case of a negative gid caused this to be
2609 noticed. The fix allows for either to be negative.
2611 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2612 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2615 PH/36 The change PH/12 above was broken. Fixed it.
2617 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2618 the grounds that routing the same address twice would always produce the
2619 same answer. This might have been true once, but it is certainly no
2620 longer true now. Routing a child address may depend on the previous
2621 routing that produced that child. Some complicated redirection strategies
2622 went wrong when messages had multiple recipients, and made Exim's
2623 behaviour dependent on the order in which the addresses were given.
2625 I have moved the duplicate checking until after the routing is complete.
2626 Exim scans the addresses that are assigned to local and remote
2627 transports, and removes any duplicates. This means that more work will be
2628 done, as duplicates will always all be routed, but duplicates are
2629 presumably rare, so I don't expect this is of any significance.
2631 For deliveries to pipes, files, and autoreplies, the duplicate checking
2632 still happens during the routing process, since they are not going to be
2635 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2636 It corrects a timeout issue with spamd. This is Ian's comment: "The
2637 background is that sometimes spamd either never reads data from a
2638 connection it has accepted, or it never writes response data. The exiscan
2639 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2640 blindly assumes that writes won't block so it may never time out."
2642 PH/39 Allow G after quota size as well as K and M.
2644 PH/40 The value set for $authenticated_id in an authenticator may not contain
2645 binary zeroes or newlines because the value is written to log lines and
2646 to spool files. There was no check on this. Now the value is run through
2647 the string_printing() function so that such characters are converted to
2648 printable escape sequences.
2650 PH/41 $message_linecount is a new variable that contains the total number of
2651 lines in the message. Compare $body_linecount, which is the count for the
2654 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2655 bounce and delay warning messages, except in certain special cases, which
2658 (a) An SMTP error message from a remote host;
2659 (b) A message specified in a :fail: redirection;
2660 (c) A message specified in a "fail" command in a system filter;
2661 (d) A message specified in a FAIL return from the queryprogram router;
2662 (e) A message specified by the cannot_route_message router option.
2664 In these cases only, Exim does include the error details in bounce and
2665 warning messages. There are also a few cases where bland messages such
2666 as "unrouteable address" or "local delivery error" are given.
2668 PH/43 $value is now also set for the "else" part of a ${run expansion.
2670 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2671 being worked on, but at least Exim now implements the latest version to
2674 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2675 process to complete was treated as a delivery failure, a timeout while
2676 writing the message to the pipe was logged, but erroneously treated as a
2677 successful delivery. Such timeouts include transport filter timeouts. For
2678 consistency with the overall process timeout, these timeouts are now
2679 treated as errors, giving rise to delivery failures by default. However,
2680 there is now a new Boolean option for the pipe transport called
2681 timeout_defer, which, if set TRUE, converts the failures into defers for
2682 both kinds of timeout. A transport filter timeout is now identified in
2685 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2686 systems where "make" and "gmake" are different, calling "gmake" at top
2687 level broke things. I've arranged for the value of $(MAKE) to be passed
2688 from the Makefile to this script so that it can call the same version of
2692 A note about Exim versions 4.44 and 4.50
2693 ----------------------------------------
2695 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2696 changes of various kinds. As a consequence, a big documentation update was
2697 needed. This delayed the release for rather longer than seemed good, especially
2698 in the light of a couple of (minor) security issues. Therefore, the changes
2699 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2700 release. So 4.44 and 4.50 are in effect two different branches that both start
2703 I have left the 4.50 change log unchanged; it contains all the changes since
2704 4.43. The change log for 4.44 is below; many of its items are identical to
2705 those for 4.50. This seems to be the most sensible way to preserve the
2706 historical information.
2712 1. Minor wording change to the doc/README.SIEVE file.
2714 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2715 computation of the current number of files was incorrect.
2717 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2718 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2721 4. Give more explanation in the error message when the command for a transport
2722 filter fails to execute.
2724 5. There are several places where Exim runs a non-Exim command in a
2725 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2726 was being done only for the command run by the queryprogram router. It is
2727 now done for all such subprocesses. The other cases are: ${run, transport
2728 filters, and the commands run by the lmtp and pipe transports.
2730 6. Added CONFIGURE_GROUP build-time option.
2732 7. Some older OS have a limit of 256 on the maximum number of file
2733 descriptors. Exim was using setrlimit() to set 1000 as a large value
2734 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2735 systems. I've change it so that if it can't get 1000, it tries for 256.
2737 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2738 was an oversight, and furthermore, ever since the addition of extra
2739 controls (e.g. 4.43/32), the checks on when to allow different forms of
2740 "control" were broken. There should now be diagnostics for all cases when a
2741 control that does not make sense is encountered.
2743 9. Added the /retain_sender option to "control=submission".
2745 10. $recipients is now available in the predata ACL (oversight).
2747 11. Tidy the search cache before the fork to do a delivery from a message
2748 received from the command line. Otherwise the child will trigger a lookup
2749 failure and thereby defer the delivery if it tries to use (for example) a
2750 cached ldap connection that the parent has called unbind on.
2752 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2753 of $address_data from the recipient verification was clobbered by the
2754 sender verification.
2756 13. The value of address_data from a sender verification is now available in
2757 $sender_address_data in subsequent conditions in the ACL statement.
2759 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2761 15. Added a new option "connect=<time>" to callout options, to set a different
2764 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2765 was its contents. (It was OK if the option was not defined at all.)
2767 17. A "Completed" log line is now written for messages that are removed from
2768 the spool by the -Mrm option.
2770 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2771 information about exactly what failed.
2773 19. Added -dd to debug only the daemon process.
2775 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2776 handles timeouts, both on the server side and network timeouts. Renamed the
2777 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2779 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2780 It is now set to "smtps".
2782 22. $host_address is now set to the target address during the checking of
2783 ignore_target_hosts.
2785 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2786 being passed; this would have caused $sender_host_name to have been used if
2787 matching the list had actually called for a host name (not very likely,
2788 since this list is usually IP addresses). A host name is now passed as
2791 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2792 code that specifies a non-restarting handler (typically sigaction() in
2793 modern systems) in an attempt to fix a rare and obscure crash bug.
2795 25. Narrowed the window for a race in the daemon that could cause it to ignore
2796 SIGCHLD signals. This is not a major problem, because they are used only to
2797 wake it up if nothing else does.
2799 26. A malformed maildirsize file could cause Exim to calculate negative values
2800 for the mailbox size or file count. Odd effects could occur as a result.
2801 The maildirsize information is now recalculated if the size or filecount
2804 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2805 support for a long time. Removed HAVE_SYS_VFS_H.
2807 28. Installed the latest version of exipick from John Jetmore.
2809 29. In an address list, if the pattern was not a regular expression, an empty
2810 subject address (from a bounce message) matched only if the pattern was an
2811 empty string. Non-empty patterns were not even tested. This was the wrong
2812 because it is perfectly reasonable to use an empty address as part of a
2813 database query. An empty address is now tested by patterns that are
2814 lookups. However, all the other forms of pattern expect the subject to
2815 contain a local part and a domain, and therefore, for them, an empty
2816 address still always fails if the pattern is not itself empty.
2818 30. Exim went into a mad DNS loop when attempting to do a callout where the
2819 host was specified on an smtp transport, and looking it up yielded more
2820 than one IP address.
2822 31. Re-factored the code for checking spool and log partition space into a
2823 function that finds that data and another that does the check. The former
2824 is then used to implement four new variables: $spool_space, $log_space,
2825 $spool_inodes, and $log_inodes.
2827 32. The RFC2047 encoding function was originally intended for short strings
2828 such as real names; it was not keeping to the 75-character limit for
2829 encoded words that the RFC imposes. It now respects the limit, and
2830 generates multiple encoded words if necessary. To be on the safe side, I
2831 have increased the buffer size for the ${rfc2047: expansion operator from
2834 33. It is now permitted to omit both strings after an "if" condition; if the
2835 condition is true, the result is "true". As before, when the second string
2836 is omitted, a false condition yields an empty string. This makes it less
2837 cumbersome to write custom ACL and router conditions.
2839 34. Failure to deliver a bounce message always caused it to be frozen, even if
2840 there was an errors_to setting on the router. The errors_to setting is now
2843 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2844 canonical form (fully expanded) before being placed in
2845 $sender_host_address.
2847 36. The table in the code that translates DNS record types into text (T_A to
2848 "A" for instance) was missing entries for NS and CNAME. It is just possible
2849 that this could have caused confusion if both these types were looked up
2850 for the same domain, because the text type is used as part of Exim's
2851 per-process caching. But the chance of anyone hitting this buglet seems
2854 37. The dnsdb lookup has been extended in a number of ways.
2856 (1) There is a new type, "zns", which walks up the domain tree until it
2857 finds some nameserver records. It should be used with care.
2859 (2) There is a new type, "mxh", which is like "mx" except that it returns
2860 just the host names, not the priorities.
2862 (3) It is now possible to give a list of domains (or IP addresses) to be
2863 looked up. The behaviour when one of the lookups defers can be
2864 controlled by a keyword.
2866 (4) It is now possible to specify the separator character for use when
2867 multiple records are returned.
2869 38. The dnslists ACL condition has been extended: it is now possible to supply
2870 a list of IP addresses and/or domains to be looked up in a particular DNS
2873 39. Added log_selector=+queue_time_overall.
2875 40. When running the queue in the test harness, wait just a tad after forking a
2876 delivery process, to get repeatability of debugging output.
2878 41. Include certificate and key file names in error message when GnuTLS fails
2879 to set them up, because the GnuTLS error message doesn't include the name
2880 of the failing file when there is a problem reading it.
2882 42. Allow both -bf and -bF in the same test run.
2884 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2886 44. The "Exiscan patch" is now merged into the mainline Exim source.
2888 45. Sometimes the final signoff response after QUIT could fail to get
2889 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2890 < 0 before doing a fflush(). This bug looks as though it goes back to the
2891 introduction of TLS in release 3.20, but "sometimes" must have been rare
2892 because the tests only now provoked it.
2894 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2895 (this can affect the format of dates).
2897 47. exim_tidydb, when checking for the continued existence of a message for
2898 which it has found a message-specific retry record, was not finding
2899 messages that were in split spool directories. Consequently, it was
2900 deleting retry records that should have stayed in existence.
2902 48. Steve fixed some bugs in eximstats.
2904 49. The SPA authentication driver was not abandoning authentication and moving
2905 on to the next authenticator when an expansion was forced to fail,
2906 contradicting the general specification for all authenticators. Instead it
2907 was generating a temporary error. It now behaves as specified.
2909 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2910 (the order specifies the preference for clients). The order is now AES256,
2911 AES128, 3DES, ARCFOUR128.
2913 51. Small patch to Sieve code - explicitly set From: when generating an
2916 52. Exim crashed if a remote delivery caused a very long error message to be
2917 recorded - for instance if somebody sent an entire SpamAssassin report back
2918 as a large number of 550 error lines. This bug was coincidentally fixed by
2919 increasing the size of one of Exim's internal buffers (big_buffer) that
2920 happened as part of the Exiscan merge. However, to be on the safe side, I
2921 have made the code more robust (and fixed the comments that describe what
2924 53. Now that there can be additional text after "Completed" in log lines (if
2925 the queue_time_overall log selector is set), a one-byte patch to exigrep
2926 was needed to allow it to recognize "Completed" as not the last thing in
2929 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2930 patch that reportedly fixes this has been added. I am not expert enough to
2931 create a test for it. This is what the patch creator wrote:
2933 "I found a little strange behaviour of ldap code when working with
2934 Windows 2003 AD Domain, where users was placed in more than one
2935 Organization Units. When I tried to give exim partial DN, the exit code
2936 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2937 But simultaneously result of request was absolutely normal ldap result,
2938 so I produce this patch..."
2940 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2941 REFERENCE, so I have modified the code to exclude the patch when that macro
2944 55. Some experimental protocols are using DNS PTR records for new purposes. The
2945 keys for these records are domain names, not reversed IP addresses. The
2946 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2947 leaves it alone. Component reversal etc. now happens only for IP addresses.
2950 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2952 57. Double the size of the debug message buffer (to 2048) so that more of very
2953 long debug lines gets shown.
2955 58. The exicyclog utility now does better if the number of log files to keep
2956 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2958 59. Two changes related to the smtp_active_hostname option:
2960 (1) $smtp_active_hostname is now available as a variable.
2961 (2) The default for smtp_banner uses $smtp_active_hostname instead
2962 of $primary_hostname.
2964 60. The host_aton() function is supposed to be passed a string that is known
2965 to be a valid IP address. However, in the case of IPv6 addresses, it was
2966 not checking this. This is a hostage to fortune. Exim now panics and dies
2967 if the condition is not met. A case was found where this could be provoked
2968 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2969 components; fortuitously, this particular loophole had already been fixed
2970 by change 4.50/55 above.
2972 If there are any other similar loopholes, the new check in host_aton()
2973 itself should stop them being exploited. The report I received stated that
2974 data on the command line could provoke the exploit when Exim was running as
2975 exim, but did not say which command line option was involved. All I could
2976 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2977 running as the user.
2980 61. There was a buffer overflow vulnerability in the SPA authentication code
2981 (which came originally from the Samba project). I have added a test to the
2982 spa_base64_to_bits() function which I hope fixes it.
2985 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2986 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2988 63. The daemon start-up calls getloadavg() while still root for those OS that
2989 need the first call to be done as root, but it missed one case: when
2990 deliver_queue_load_max is set with deliver_drop_privilege. This is
2991 necessary for the benefit of the queue runner, because there is no re-exec
2992 when deliver_drop_privilege is set.
2994 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2995 This has been fixed.
2997 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2998 was in use, was not putting the data itself into the right store pool;
2999 consequently, it could be overwritten for a subsequent message in the same
3000 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3003 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3005 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3006 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3007 the function were treating the return as a boolean value, which happened to
3008 work because 0=false and not-0=true, but is not correct code.
3010 68. The host_aton() function was not handling scoped IPv6 addresses (those
3011 with, for example, "%eth0" on the end) correctly.
3013 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3014 negated items (that is, ~something) in unsigned ints. Some compilers
3015 apparently mutter when there is no cast.
3017 70. If an address verification called from an ACL failed, and did not produce a
3018 user-specific message (i.e. there was only a "system" message), nothing was
3019 put in $acl_verify_message. In this situation, it now puts the system
3022 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3023 session; change 4.31/43 added the unwanted input to the log line - except
3024 that it did not do this in the start of session case. It now does.
3026 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3027 This is wrong and can cause the other end to generate a synchronization
3028 error if it is another Exim or anything else that does the synchronization
3029 check. A QUIT command is no longer sent after a timeout.
3031 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3032 during host lookups.
3034 74. The defer_ok option of callout verification was not working if it was used
3035 when verifying addresses in header lines, that is, for this case:
3037 verify = header_sender/callout=defer_ok
3039 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3040 those file descriptors could be used for SMTP connections. If anything
3041 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3042 could be sent to the SMTP client, causing chaos. The daemon now opens
3043 stdin, stdout, and stderr to /dev/null when it puts itself into the
3046 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3047 log by default. The user can override this with suitable Perl magic.
3049 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3050 the log message when discard triggers, was not working for the DATA ACL or
3051 for the non-SMTP ACL.
3053 78. Error message wording change in sieve.c.
3055 79. If smtp_accept_max_per_host was set, the number of connections could be
3056 restricted to fewer than expected, because the daemon was trying to set up
3057 a new connection before checking whether the processes handling previous
3058 connections had finished. The check for completed processes is now done
3059 earlier. On busy systems, this bug wouldn't be noticed because something
3060 else would have woken the daemon, and it would have reaped the completed
3063 80. If a message was submitted locally by a user whose login name contained one
3064 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3065 It caused a spool format error. I have fixed the spool reading code. A
3066 related problem was that the "from" clause in the Received: line became
3067 illegal because of the space(s). It is now covered by ${quote_local_part.
3069 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3072 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3074 83. Patch from Sieve maintainer to fix unterminated string problem in
3075 "vacation" handling.
3077 84. Some minor changes to the Linux configuration files to help with other
3078 OS variants using glibc.
3080 85. One more patch for Sieve to update vacation handling to latest spec.
3083 ----------------------------------------------------
3084 See the note above about the 4.44 and 4.50 releases.
3085 ----------------------------------------------------
3091 1. Change 4.43/35 introduced a bug that caused file counts to be
3092 incorrectly computed when quota_filecount was set in an appendfile
3095 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3096 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3099 3. Give more explanation in the error message when the command for a transport
3100 filter fails to execute.
3102 4. There are several places where Exim runs a non-Exim command in a
3103 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3104 was being done only for the command run by the queryprogram router. It is
3105 now done for all such subprocesses. The other cases are: ${run, transport
3106 filters, and the commands run by the lmtp and pipe transports.
3108 5. Some older OS have a limit of 256 on the maximum number of file
3109 descriptors. Exim was using setrlimit() to set 1000 as a large value
3110 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3111 systems. I've change it so that if it can't get 1000, it tries for 256.
3113 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3114 was an oversight, and furthermore, ever since the addition of extra
3115 controls (e.g. 4.43/32), the checks on when to allow different forms of
3116 "control" were broken. There should now be diagnostics for all cases when a
3117 control that does not make sense is encountered.
3119 7. $recipients is now available in the predata ACL (oversight).
3121 8. Tidy the search cache before the fork to do a delivery from a message
3122 received from the command line. Otherwise the child will trigger a lookup
3123 failure and thereby defer the delivery if it tries to use (for example) a
3124 cached ldap connection that the parent has called unbind on.
3126 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3127 of $address_data from the recipient verification was clobbered by the
3128 sender verification.
3130 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3131 was its contents. (It was OK if the option was not defined at all.)
3133 11. A "Completed" log line is now written for messages that are removed from
3134 the spool by the -Mrm option.
3136 12. $host_address is now set to the target address during the checking of
3137 ignore_target_hosts.
3139 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3140 being passed; this would have caused $sender_host_name to have been used if
3141 matching the list had actually called for a host name (not very likely,
3142 since this list is usually IP addresses). A host name is now passed as
3145 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3146 code that specifies a non-restarting handler (typically sigaction() in
3147 modern systems) in an attempt to fix a rare and obscure crash bug.
3149 15. Narrowed the window for a race in the daemon that could cause it to ignore
3150 SIGCHLD signals. This is not a major problem, because they are used only to
3151 wake it up if nothing else does.
3153 16. A malformed maildirsize file could cause Exim to calculate negative values
3154 for the mailbox size or file count. Odd effects could occur as a result.
3155 The maildirsize information is now recalculated if the size or filecount
3158 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3159 support for a long time. Removed HAVE_SYS_VFS_H.
3161 18. Updated exipick to current release from John Jetmore.
3163 19. Allow an empty sender to be matched against a lookup in an address list.
3164 Previously the only cases considered were a regular expression, or an
3167 20. Exim went into a mad DNS lookup loop when doing a callout where the
3168 host was specified on the transport, if the DNS lookup yielded more than
3171 21. The RFC2047 encoding function was originally intended for short strings
3172 such as real names; it was not keeping to the 75-character limit for
3173 encoded words that the RFC imposes. It now respects the limit, and
3174 generates multiple encoded words if necessary. To be on the safe side, I
3175 have increased the buffer size for the ${rfc2047: expansion operator from
3178 22. Failure to deliver a bounce message always caused it to be frozen, even if
3179 there was an errors_to setting on the router. The errors_to setting is now
3182 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3183 canonical form (fully expanded) before being placed in
3184 $sender_host_address.
3186 24. Updated eximstats to version 1.33
3188 25. Include certificate and key file names in error message when GnuTLS fails
3189 to set them up, because the GnuTLS error message doesn't include the name
3190 of the failing file when there is a problem reading it.
3192 26. Expand error message when OpenSSL has problems setting up cert/key files.
3195 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3196 (this can affect the format of dates).
3198 28. exim_tidydb, when checking for the continued existence of a message for
3199 which it has found a message-specific retry record, was not finding
3200 messages that were in split spool directories. Consequently, it was
3201 deleting retry records that should have stayed in existence.
3203 29. eximstats updated to version 1.35
3204 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3205 1.35 - bugfix such that pie charts by volume are generated correctly
3207 30. The SPA authentication driver was not abandoning authentication and moving
3208 on to the next authenticator when an expansion was forced to fail,
3209 contradicting the general specification for all authenticators. Instead it
3210 was generating a temporary error. It now behaves as specified.
3212 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3213 (the order specifies the preference for clients). The order is now AES256,
3214 AES128, 3DES, ARCFOUR128.
3216 31. Small patch to Sieve code - explicitly set From: when generating an
3219 32. Exim crashed if a remote delivery caused a very long error message to be
3220 recorded - for instance if somebody sent an entire SpamAssassin report back
3221 as a large number of 550 error lines. This bug was coincidentally fixed by
3222 increasing the size of one of Exim's internal buffers (big_buffer) that
3223 happened as part of the Exiscan merge. However, to be on the safe side, I
3224 have made the code more robust (and fixed the comments that describe what
3227 33. Some experimental protocols are using DNS PTR records for new purposes. The
3228 keys for these records are domain names, not reversed IP addresses. The
3229 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3230 leaves it alone. Component reversal etc. now happens only for IP addresses.
3233 34. The host_aton() function is supposed to be passed a string that is known
3234 to be a valid IP address. However, in the case of IPv6 addresses, it was
3235 not checking this. This is a hostage to fortune. Exim now panics and dies
3236 if the condition is not met. A case was found where this could be provoked
3237 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3238 components; fortuitously, this particular loophole had already been fixed
3239 by change 4.50/55 or 4.44/33 above.
3241 If there are any other similar loopholes, the new check in host_aton()
3242 itself should stop them being exploited. The report I received stated that
3243 data on the command line could provoke the exploit when Exim was running as
3244 exim, but did not say which command line option was involved. All I could
3245 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3246 running as the user.
3249 35. There was a buffer overflow vulnerability in the SPA authentication code
3250 (which came originally from the Samba project). I have added a test to the
3251 spa_base64_to_bits() function which I hope fixes it.
3254 36. The daemon start-up calls getloadavg() while still root for those OS that
3255 need the first call to be done as root, but it missed one case: when
3256 deliver_queue_load_max is set with deliver_drop_privilege. This is
3257 necessary for the benefit of the queue runner, because there is no re-exec
3258 when deliver_drop_privilege is set.
3260 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3261 was in use, was not putting the data itself into the right store pool;
3262 consequently, it could be overwritten for a subsequent message in the same
3263 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3266 38. Sometimes the final signoff response after QUIT could fail to get
3267 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3268 < 0 before doing a fflush(). This bug looks as though it goes back to the
3269 introduction of TLS in release 3.20, but "sometimes" must have been rare
3270 because the tests only now provoked it.
3276 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3277 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3278 Now it returns an integer. A number of calls were still expecting a T/F
3279 return. Fortuitously, in all cases, the tests worked in OK situations,
3280 which is the norm. However, things would have gone wrong on any write
3281 failures on the smtp file descriptor. This function is used when sending
3282 messages over SMTP and also when doing verify callouts.
3284 2. When Exim is called to do synchronous delivery of a locally submitted
3285 message (the -odf or -odi options), it no longer closes stderr before doing
3288 3. Implemented the mua_wrapper option.
3290 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3292 5. Implemented the functions header_remove(), header_testname(),
3293 header_add_at_position(), and receive_remove_recipient(), and exported them
3296 6. If an ACL "warn" statement specified the addition of headers, Exim already
3297 inserted X-ACL-Warn: at the start if there was no header name. However, it
3298 was not making this test for the second and subsequent header lines if
3299 there were newlines in the string. This meant that an invalid header could
3300 be inserted if Exim was badly configured.
3302 7. Allow an ACL "warn" statement to add header lines at the start or after all
3303 the Received: headers, as well as at the end.
3305 8. Added the rcpt_4xx retry error code.
3307 9. Added postmaster_mailfrom=xxx to callout verification option.
3309 10. Added mailfrom=xxxx to the callout verification option, for verify=
3312 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3313 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3315 12. Inserted some casts to stop certain compilers warning when using pointer
3316 differences as field lengths or precisions in printf-type calls (mostly
3317 affecting debugging statements).
3319 13. Added optional readline() support for -be (dynamically loaded).
3321 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3322 same clock tick as a message's arrival, so that its received time was the
3323 same as the "first fail" time on the retry record, and that message
3324 remained on the queue past the ultimate address timeout, every queue runner
3325 would try a delivery (because it was past the ultimate address timeout) but
3326 after another failure, the ultimate address timeout, which should have then
3327 bounced the address, did not kick in. This was a "< instead of <=" error;
3328 in most cases the first failure would have been in the next clock tick
3329 after the received time, and all would be well.
3331 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3332 being recognized when the domain list was tested by the match_domain
3333 condition in an expansion string.
3335 16. Added the ${str2b64: operator.
3337 17. Exim was always calling setrlimit() to set a large limit for the number of
3338 processes, without checking whether the existing limit was already
3339 adequate. (It did check for the limit on file descriptors.) Furthermore,
3340 errors from getrlimit() and setrlimit() were being ignored. Now they are
3341 logged to the main and panic logs, but Exim does carry on, to try to do its
3342 job under whatever limits there are.
3344 18. Imported PCRE 5.0.
3346 19. Trivial typo in log message " temporarily refused connection" (the leading
3349 20. If the log selector return_path_on_delivery was set and an address was
3350 redirected to /dev/null, the delivery process crashed because it assumed
3351 that a return path would always be set for a "successful" delivery. In this
3352 case, the whole delivery is bypassed as an optimization, and therefore no
3355 21. Internal re-arrangement: the function for sending a challenge and reading
3356 a response while authentication was assuming a zero-terminated challenge
3357 string. It's now changed to take a pointer and a length, to allow for
3358 binary data in such strings.
3360 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3362 23. Exim was not respecting finduser_retries when seeking the login of the
3363 uid under which it was called; it was always trying 10 times. (The default
3364 setting of finduser_retries is zero.) Also, it was sleeping after the final
3365 failure, which is pointless.
3367 24. Implemented tls_on_connect_ports.
3369 25. Implemented acl_smtp_predata.
3371 26. If the domain in control=submission is set empty, Exim assumes that the
3372 authenticated id is a complete email address when it generates From: or
3373 Sender: header lines.
3375 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3376 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3377 chown and chgrp in /bin and hostname in /usr/bin.
3379 28. Exim was keeping the "process log" file open after each use, just as it
3380 does for the main log. This opens the possibility of it remaining open for
3381 long periods when the USR1 signal hits a daemon. Occasional processlog
3382 errors were reported, that could have been caused by this. Anyway, it seems
3383 much more sensible not to leave this file open at all, so that is what now
3386 29. The long-running daemon process does not normally write to the log once it
3387 has entered its main loop, and it closes the log before doing so. This is
3388 so that log files can straightforwardly be renamed and moved. However,
3389 there are a couple of unusual error situations where the daemon does write
3390 log entries, and I had neglected to close the log afterwards.
3392 30. The text of an SMTP error response that was received during a remote
3393 delivery was being truncated at 512 bytes. This is too short for some of
3394 the long messages that one sometimes sees. I've increased the limit to
3397 31. It is now possible to make retry rules that apply only when a message has a
3398 specific sender, in particular, an empty sender.
3400 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3401 it possible to be selective about when SMTP synchronization is enforced.
3403 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3405 32. Implemented hosts_connection_nolog.
3407 33. Added an ACL for QUIT.
3409 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3412 35. Added mailbox_size and mailbox_filecount to appendfile.
3414 36. Added control = no_multiline_responses to ACLs.
3416 37. There was a bug in the logic of the code that waits for the clock to tick
3417 in the case where the clock went backwards by a substantial amount such
3418 that the microsecond fraction of "now" was more than the microsecond
3419 fraction of "then" (but the whole seconds number was less).
3421 38. Added support for the libradius Radius client library this is found on
3422 FreeBSD (previously only the radiusclient library was supported).
3428 1. When certain lookups returned multiple values in the form name=value, the
3429 quoting of the values was not always being done properly. Specifically:
3430 (a) If the value started with a double quote, but contained no whitespace,
3432 (b) If the value contained whitespace other than a space character (i.e.
3433 tabs or newlines or carriage returns) it was not quoted.
3434 This fix has been applied to the mysql and pgsql lookups by writing a
3435 separate quoting function and calling it from the lookup code. The fix
3436 should probably also be applied to nisplus, ibase and oracle lookups, but
3437 since I cannot test any of those, I have not disturbed their existing code.
3439 2. A hit in the callout cache for a specific address caused a log line with no
3440 reason for rejecting RCPT. Now it says "Previous (cached) callout
3441 verification failure".
3443 3. There was an off-by-one bug in the queryprogram router. An over-long
3444 return line was truncated at 256 instead of 255 characters, thereby
3445 overflowing its buffer with the terminating zero. As well as fixing this, I
3446 have increased the buffer size to 1024 (and made a note to document this).
3448 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3449 when Exim is waiting for an SMTP response from a remote server, Exim
3450 restarts its select() call on the socket, thereby resetting its timeout.
3451 This is not a problem when such interrupts are rare. Somebody set up a cron
3452 job to run exiwhat every 2 minutes, which is less than the normal select()
3453 timeout (5 or 10 minutes). This meant that the select() timeout never
3454 kicked in because it was always reset. I have fixed this by comparing the
3455 time when an interrupt arrives with the time at the start of the first call
3456 to select(). If more time than the timeout has elapsed, the interrupt is
3457 treated as a timeout.
3459 5. Some internal re-factoring in preparation for the addition of Sieve
3460 extensions (by MH). In particular, the "personal" test is moved to a
3461 separate function, and given an option for scanning Cc: and Bcc: (which is
3462 not set for Exim filters).
3464 6. When Exim created an email address using the login of the caller as the
3465 local part (e.g. when creating a From: or Sender: header line), it was not
3466 quoting the local part when it contained special characters such as @.
3468 7. Installed new OpenBSD configuration files.
3470 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3471 try to make them clearer.
3473 9. Callout options, other than the timeout value, were being ignored when
3474 verifying sender addresses in header lines. For example, when using
3476 verify = header_sender/callout=no_cache
3478 the cache was (incorrectly) being used.
3480 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3481 only the Cygwin environment.
3483 11. When return_path_on_delivery was set as a log selector, if different remote
3484 addresses in the same message used different return paths and parallel
3485 remote delivery occurred, the wrong values would sometimes be logged.
3486 (Whenever a remote delivery process finished, the return path value from
3487 the most recently started remote delivery process was logged.)
3489 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3490 lines when AUTH and/or TLS are in use. This is the "received protocol"
3491 field. Exim used to use "asmtp" for authenticated SMTP, without any
3492 indication (in the protocol name) for TLS use. Now it follows the RFC and
3493 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3494 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3495 names appear in log lines as well as in Received: header lines.
3497 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3498 extensions, and comparison tests, and to fix some bugs.
3500 14. Changes to the "personal" filter test:
3502 (1) The test was buggy in that it was just doing the equivalent of
3503 "contains" tests on header lines. For example, if a user's address was
3504 anne@some.where, the "personal" test would incorrectly be true for
3506 To: susanne@some.where
3508 This test is now done by extracting each address from the header in turn,
3509 and checking the entire address. Other tests that are part of "personal"
3510 are now done using regular expressions (for example, to check local parts
3511 of addresses in From: header lines).
3513 (2) The list of non-personal local parts in From: addresses has been
3514 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3515 taken from the Sieve specification recommendations.
3517 (3) If the message contains any header line starting with "List-" it is
3518 treated as non-personal.
3520 (4) The test for "circular" in the Subject: header line has been removed
3521 because it now seems ill-conceived.
3523 15. Minor typos in src/EDITME comments corrected.
3525 16. Installed latest exipick from John Jetmore.
3527 17. If headers_add on a router specified a text string that was too long for
3528 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3529 of string_sprintf() is now avoided.
3531 18. $message_body_size was not set (it was always zero) when running the DATA
3532 ACL and the local_scan() function.
3534 19. For the "mail" command in an Exim filter, no default was being set for
3535 the once_repeat time, causing a random time value to be used if "once" was
3536 specified. (If the value happened to be <= 0, no repeat happened.) The
3537 default is now 0s, meaning "never repeat". The "vacation" command was OK
3538 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3539 (I found it when inspecting the code).
3541 20. There is now an overall timeout for performing a callout verification. It
3542 defaults to 4 times the callout timeout, which applies to individual SMTP
3543 commands during the callout. The overall timeout applies when there is more
3544 than one host that can be tried. The timeout is checked before trying the
3545 next host. This prevents very long delays if there are a large number of
3546 hosts and all are timing out (e.g. when the network connections are timing
3547 out). The value of the overall timeout can be changed by specifying an
3548 additional sub-option for "callout", called "maxwait". For example:
3550 verify = sender/callout=5s,maxwait=20s
3552 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3553 to the end before writing, but this should make it even safer).
3555 22. Exim was forgetting that it had advertised PIPELINING for the second and
3556 subsequent messages on an SMTP connection. It was also not resetting its
3557 memory on STARTTLS and an internal HELO.
3559 23. When Exim logs an SMTP synchronization error within a session, it now
3560 records whether PIPELINING has been advertised or not.
3562 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3563 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3564 rather than long int.
3566 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3568 26. Added the never_mail option to autoreply.
3574 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3575 crash if the getsockname() call failed; this can happen if a connection is
3576 closed very soon after it is established. The problem was simply in the
3577 order in which certain operations were done, causing Exim to try to write
3578 to the SMTP stream before it had set up the file descriptor. The bug has
3579 been fixed by making things happen in the correct order.
3585 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3586 before the connection was closed, thus losing the rejection response.
3588 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3589 some early Solaris releases, but causes trouble in current releases where
3590 socklen_t is defined.
3592 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3595 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3598 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3599 that releasing the top of it at the end releases what was used for sub-
3600 expansions (unless the block got too big). However, discard this block if
3601 the first thing is a variable or header, so that we can use its block when
3602 it is dynamic (useful for very large $message_headers, for example).
3604 6. Lookups now cache *every* query, not just the most recent. A new, separate
3605 store pool is used for this. It can be recovered when all lookup caches are
3606 flushed. Lookups now release memory at the end of their result strings.
3607 This has involved some general refactoring of the lookup sources.
3609 7. Some code has been added to the store_xxx() functions to reduce the amount
3610 of flapping under certain conditions.
3612 8. log_incoming_interface used to affect only the <= reception log lines. Now
3613 it causes the local interface and port to be added to several more SMTP log
3614 lines, for example "SMTP connection from", and rejection lines.
3616 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3618 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3620 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3621 could be overwritten at the end of the current message (or the start of a
3622 new message if it was set in a HELO ACL). The value is now preserved for
3623 the duration of the SMTP connection.
3625 12. If a transport had a headers_rewrite setting, and a matching header line
3626 contained an unqualified address, that address was qualified, even if it
3627 did not match any rewriting rules. The underlying bug was that the values
3628 of the flags that permit the existence of unqualified sender and recipient
3629 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3630 non-local messages, and by -bnq for local messages) were not being
3631 preserved with the message after it was received.
3633 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3634 "next input=" as part of the text comprising the host identity instead of
3635 the correct text. The code was using the same buffer for two different
3636 strings. However, depending on which order the printing function evaluated
3637 its arguments, the bug did not always show up. Under Linux, for example, my
3638 test suite worked just fine.
3640 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3641 doesn't work with some older versions of Perl. It has been changed to "my",
3642 which in any case is probably the better facility to use.
3644 15. A really picky compiler found some instances of statements for creating
3645 error messages that either had too many or two few arguments for the format
3648 16. The size of the buffer for calls to the DNS resolver has been increased
3649 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3650 for addresses that have a lot of PTR records. This alleviates a problem; it
3651 does not fully solve it.
3653 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3654 buffer now truncates the list and logs the incident, which is the same
3655 action as happens when Exim is looking up a host name and its aliases.
3656 Previously in this situation something unpredictable would happen;
3657 sometimes it was "internal error: store_reset failed".
3659 18. If a server dropped the connection unexpectedly when an Exim client was
3660 using GnuTLS and trying to read a response, the client delivery process
3661 crashed while trying to generate an error log message.
3663 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3664 string, for example:
3666 warn message = H1: something\nH2: something
3668 the text was added as a single header line from Exim's point of view
3669 though it ended up OK in the delivered message. However, searching for the
3670 second and subsequent header lines using $h_h2: did not work. This has been
3671 fixed. Similarly, if a system filter added multiple headers in this way,
3672 the routers could not see them.
3674 20. Expanded the error message when iplsearch is called with an invalid key to
3675 suggest using net-iplsearch in a host list.
3677 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3678 ACLs are no longer actually imposed (and a message to that effect is
3681 22. If a "gecos" field in a passwd entry contained escaped characters, in
3682 particular, if it contained a \" sequence, Exim got it wrong when building
3683 a From: or a Sender: header from that name. A second bug also caused
3684 incorrect handling when an unquoted " was present following a character
3685 that needed quoting.
3687 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3688 was not being matched caselessly.
3690 24. Arranged for all hyphens in the exim.8 source to be escaped with
3693 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3694 properly. Recipient callout cache records were still being keyed to include
3695 the sender, even when use_sender was set false. This led to far more
3696 callouts that were necessary. The sender is no longer included in the key
3697 when use_sender is false.
3699 26. Added "control = submission" modifier to ACLs.
3701 27. Added the ${base62d: operator to decode base 62 numbers.
3703 28. dnsdb lookups can now access SRV records.
3705 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3706 the configuration file.
3708 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3709 (-v) mode. This makes the output for a verbose queue run more intelligible.
3711 31. Added a use_postmaster feature to recipient callouts.
3713 32. Added the $body_zerocount variable, containing the number of binary zero
3714 bytes in the message body.
3716 33. The time of last modification of the "new" subdirectory is now used as the
3717 "mailbox time last read" when there is a quota error for a maildir
3720 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3722 35. Added +ignore_unknown as a special item in host lists.
3724 36. Code for decoding IPv6 addresses in host lists is now included, even if
3725 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3726 address was recognized as an IP address, but was then not correctly decoded
3727 into binary, causing unexpected and incorrect effects when compared with
3734 1. Very minor rewording of debugging text in manualroute to say "list of
3735 hosts" instead of "hostlist".
3737 2. If verify=header_syntax was set, and a header line with an unqualified
3738 address (no domain) and a large number of spaces between the end of the
3739 name and the colon was received, the reception process suffered a buffer
3740 overflow, and (when I tested it) crashed. This was caused by some obsolete
3741 code that should have been removed. The fix is to remove it!
3743 3. When running in the test harness, delay a bit after writing a bounce
3744 message to get a bit more predictability in the log output.
3746 4. Added a call to search_tidyup() just before forking a reception process. In
3747 theory, someone could use a lookup in the expansion of smtp_accept_max_
3748 per_host which, without the tidyup, could leave open a database connection.
3750 5. Added the variables $recipient_data and $sender_data which get set from a
3751 lookup success in an ACL "recipients" or "senders" condition, or a router
3752 "senders" option, similar to $domain_data and $local_part_data.
3754 6. Moved the writing of debug_print from before to after the "senders" test
3757 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3758 problems for message scanning, either using a data ACL, or using
3759 local_scan() because the Received: header was not generated till after they
3760 were called (in order to set the time as the time of reception completion).
3761 I have revised the way this works. The header is now generated after the
3762 body is received, but before the ACL or local_scan() are called. After they
3763 are run, the timestamp in the header is updated.
3769 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3770 before starting a queue runner without re-exec. This happened only when
3771 deliver_drop_privilege was set or when the Exim user was set to root. The
3772 effect of the bug was that timeouts during subsequent deliveries caused
3773 crashes instead of being properly handled. The handler is now left at its
3774 default (and expected) setting.
3776 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3777 message, again when deliver_drop_privilege is set or Exim is run as root.
3778 The bug described in (1) was not present in this case, but the tidying up
3779 of the other signals was missing. I have made the two cases consistent.
3781 3. The ignore_target_hosts setting on a manualroute router was being ignored
3782 for hosts that were looked up using the /MX notation.
3784 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3787 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3788 operated on the sender address. After changing the $sender_address to <>
3789 for the sender address verify, Exim was re-instated it as the original
3790 (before rewriting) address, but remembering that it had rewritten it, so it
3791 wasn't rewriting it again. This bug also had the effect of breaking the
3792 sender address verification caching when the sender address was rewritten.
3794 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3795 This has been changed so that if the ip literal address matches
3796 ignore_target_hosts, the router declines.
3798 7. Added expansion conditions match_domain, match_address, and match_local_
3799 part (NOT match_host).
3801 8. The placeholder for the Received: header didn't have a length field set.
3803 9. Added code to Exim itself and to exim_lock to test for a specific race
3804 condition that could lead to file corruption when using MBX delivery. The
3805 issue is with the lockfile that is created in /tmp. If this file is removed
3806 after a process has opened it but before that process has acquired a lock,
3807 there is the potential for a second process to recreate the file and also
3808 acquire a lock. This could lead to two Exim processes writing to the file
3809 at the same time. The added code performs the same test as UW imapd; it
3810 checks after acquiring the lock that its file descriptor still refers to
3811 the same named file.
3813 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3814 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3817 11. Added the smtp_active_hostname option. If used, this will typically be made
3818 to depend on the incoming interface address. Because $interface_address is
3819 not set up until the daemon has forked a reception process, error responses
3820 that can happen earlier (such as "too many connections") no longer contain
3823 12. If an expansion in a condition on a "warn" statement fails because a lookup
3824 defers, the "warn" statement is abandoned, and the next ACL statement is
3825 processed. Previously this caused the whole ACL to be aborted.
3827 13. Added the iplsearch lookup type.
3829 14. Added ident_timeout as a log selector.
3831 15. Added tls_certificate_verified as a log selector.
3833 16. Added a global option tls_require_ciphers (compare the smtp transport
3834 option of the same name). This controls incoming TLS connections.
3836 17. I finally figured out how to make tls_require_ciphers do a similar thing
3837 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3838 before starting the TLS session.
3840 18. Tabs are now shown as \t in -bP output.
3842 19. If the log selector return_path_on_delivery was set, Exim crashed when
3843 bouncing a message because it had too many Received: header lines.
3845 20. If two routers both had headers_remove settings, and the first one included
3846 a superfluous trailing colon, the final name in the first list and the
3847 first name in the second list were incorrectly joined into one item (with a
3848 colon in the middle).
3854 1. Added -C and -D options to the exinext utility, mainly to make it easier
3855 to include in the automated testing, but these could be helpful when
3856 multiple configurations are in use.
3858 2. The exinext utility was not formatting the output nicely when there was
3859 an alternate port involved in the retry record key, nor when there was a
3860 message id as well (for retries that were specific to a specific message
3861 and a specific host). It was also confused by IPv6 addresses, because of
3862 the additional colons they contain. I have fixed the IPv4 problem, and
3863 patched it up to do a reasonable job for IPv6.
3865 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3866 delivery, the log line now contains "pipelined" if PIPELINING was used.
3868 4. An SMTP transport process used to panic and die if the bind() call to set
3869 an explicit outgoing interface failed. This has been changed; it is now
3870 treated in the same way as a connect() failure.
3872 5. A reference to $sender_host_name in the part of a conditional expansion
3873 that was being skipped was still causing a DNS lookup. This no longer
3876 6. The def: expansion condition was not recognizing references to header lines
3877 that used bh_ and bheader_.
3879 7. Added the _cache feature to named lists.
3881 8. The code for checking quota_filecount in the appendfile transport was
3882 allowing one more file than it should have been.
3884 9. For compatibility with Sendmail, the command line option
3892 and sets the incoming protocol and host name (for trusted callers). The
3893 host name and its colon can be omitted when only the protocol is to be set.
3894 Note the Exim already has two private options, -pd and -ps, that refer to
3895 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3896 "s", but I don't think that's a major issue.
3898 10. A number of refactoring changes to the code, none of which should affect
3901 (a) The number of logging options was getting close to filling up the
3902 32-bit word that was used as a bit map. I have split them into two classes:
3903 those that are passed in the argument to log_write(), and those that are
3904 only ever tested independently outside of that function. These are now in
3905 separate 32-bit words, so there is plenty of room for expansion again.
3906 There is no change in the user interface or the logging behaviour.
3908 (b) When building, for example, log lines, the code previously used a
3909 macro that called string_cat() twice, in order to add two strings. This is
3910 not really sufficiently general. Furthermore, there was one instance where
3911 it was actually wrong because one of the argument was used twice, and in
3912 one call a function was used. (As it happened, calling the function twice
3913 did not affect the overall behaviour.) The macro has been replaced by a
3914 function that can join an arbitrary number of extra strings onto a growing
3917 (c) The code for expansion conditions now uses a table and a binary chop
3918 instead of a serial search (which was left over from when there were very
3919 few conditions). Also, it now recognizes conditions like "pam" even when
3920 the relevant support is not compiled in: a suitably worded error message is
3921 given if an attempt is made to use such a condition.
3923 11. Added ${time_interval:xxxxx}.
3925 12. A bug was causing one of the ddress fields not to be passed back correctly
3926 from remote delivery subprocesses. The field in question was not being
3927 subsequently used, so this caused to problems in practice.
3929 13. Added new log selectors queue_time and deliver_time.
3931 14. Might have fixed a bug in maildirsizefile handling that threw up
3932 "unexpected character" debug warnings, and recalculated the data
3933 unnecessarily. In any case, I expanded the warning message to give more
3936 15. Added the message "Restricted characters in address" to the statements in
3937 the default ACL that block characters like @ and % in local parts.
3939 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3940 Three changes have been made:
3942 (a) There was a serious bug; a negative response to MAIL caused the whole
3943 recipient domain to be cached as invalid, thereby blocking all messages
3944 to all local parts at the same domain, from all senders. This bug has
3945 been fixed. The domain is no longer cached after a negative response to
3946 MAIL if the sender used is not empty.
3948 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3951 (c) A new callout option, "use_sender" has been added for people who want
3952 the modified behaviour.
3958 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3961 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3962 indeed breaks things for older releases.
3964 3. Added additional logging to the case where there is a problem reading data
3965 from a filter that is running in a subprocess using a pipe, in order to
3966 try to track down a specific problem.
3968 4. Testing facility fudge: when running in the test harness and attempting
3969 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3970 getting "No route to host". Convert this to a timeout.
3972 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3975 6. Some OS don't have socklen_t but use size_t instead. This affects the
3976 fifth argument of getsockopt() amongst other things. This is now
3977 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3978 can be set for individual OS. I have set it for SunOS5, OSF1, and
3979 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3980 some earlier ones do not.
3982 7. Change 4.30/15 was not doing the test caselessly.
3984 8. The standard form for an IPv6 address literal was being rejected by address
3985 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3986 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3987 this, as well as the form without the "IPv6" on the front (but only when
3988 address literals are enabled, of course).
3990 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3992 10. Exim crashed if a message with an empty sender address specified by -f
3993 encountered a router with an errors_to setting. This could be provoked only
3994 by a command such as
3998 where an empty string was supplied; "<>" did not hit this bug.
4000 11. Installed PCRE release 4.5.
4002 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4003 remained set. It is now erased.
4005 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4006 times from message ids (which are base 36 rather than the normal 62).
4008 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4009 were being counted as actual protocol errors, and logged if the log
4010 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4011 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4012 and DATA following a set of rejected RCPTs do not count as protocol errors.
4013 In other words, Exim assumes they were pipelined, though this may not
4014 actually be the case. Of course, in all cases the client gets an
4015 appropriate error code.
4017 15. If a lookup fails in an ACL condition, a message about the failure may
4018 be available; it is used if testing the ACL cannot continue, because most
4019 such messages specify what the cause of the deferral is. However, some
4020 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4021 that caused an old message to be retained and used if a later statement
4022 caused a defer, replacing the real cause of the deferral.
4024 16. If an IP address had so many PTR records that the DNS lookup buffer
4025 was not large enough to hold them, Exim could crash while trying to process
4026 the truncated data. It now detects and logs this case.
4028 17. Further to 4.21/58, another change has been made: if (and only if) the
4029 first line of a message (the first header line) ends with CRLF, a bare LF
4030 in a subsequent header line has a space inserted after it, so as not to
4031 terminate the header.
4033 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4034 unnecessarily, used atoi() instead of strtol(), and didn't check the
4035 termination when getting file sizes from file names by regex.
4037 19. Completely re-implemented the support for maildirsize files, in the light
4038 of a number of problems with the previous contributed implementation
4039 (4.30/29). In particular:
4041 . If the quota is zero, the maildirsize file is maintained, but no quota is
4044 . If the maildir directory does not exist, it is created before any attempt
4045 to write a maildirsize file.
4047 . The quota value in the file is just a cache; if the quota is changed in
4048 the transport, the new value overrides.
4050 . A regular expression is available for excluding directories from the
4053 20. The autoreply transport checks the characters in options that define the
4054 message's headers; it allows continued headers, but it was checking with
4055 isspace() after an embedded newline instead of explicitly looking for a
4058 21. If all the "regular" hosts to which an address was routed had passed their
4059 expiry times, and had not reached their retry times, the address was
4060 bounced, even if fallback hosts were defined. Now Exim should go on to try
4063 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4064 equivalent code in the SMTP transport. Some hosts send humungous responses
4065 to HELO/EHLO, more than 1024 it seems.
4067 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4068 gives compiler warnings in some environments. I've now done it "properly",
4071 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4072 (because of problems with the built-in one) was declared to return uschar *
4073 instead of char *, causing compiler failure.
4075 25. Fixed a file descriptor leak when processing alias/forward files.
4077 26. Fixed a minor format string issue in dbfn.c.
4079 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4081 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4082 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4085 29. When a sender address is verified, it is cached, to save repeating the test
4086 when there is more than one recipient in a message. However, when the
4087 verification involves a callout, it is possible for different callout
4088 options to be set for different recipients. It is too complicated to keep
4089 track of this in the cache, so now Exim always runs a verification when a
4090 callout is required, relying on the callout cache for the optimization.
4091 The overhead is duplication of the address routing, but this should not be
4094 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4095 to be verified with callout=postmaster, and the main callout worked but the
4096 postmaster check failed, the verification correctly failed. However, if a
4097 subsequent RCPT command asked for sender verification *without* the
4098 postmaster check, incorrect caching caused this verification also to fail,
4101 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4102 it was not caching the DNS options (qualify_single, search_parents) that
4103 were used when the lookup failed. A subsequent lookup with different
4104 options therefore always gave the same answer, though there were cases
4105 where it should not have. (Example: a "domains = !$mx_any" option on a
4106 dnslookup router: the "domains" option is always processed without any
4107 widening, but the router might have qualify_single set.) Now Exim uses the
4108 cached value only when the same options are set.
4110 32. Added John Jetmore's "exipick" utility to the distribution.
4112 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4113 than a timeout (e.g. a certificate is required, and is not provided), an
4114 Exim server now closes the connection immediately. Previously it waited for
4115 the client to close - but if the client is SSL, it seems that they each
4116 wait for each other, leading to a delay before one of them times out.
4118 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4119 maintained 0.8.x compatibility because I don't think many are using it, and
4120 it is clearly obsolete.
4122 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4125 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4126 client certificate was expired. A simple patch fixes this, though I don't
4127 understand the full logic of why the verify callback is called multiple
4130 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4131 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4132 which causes problems with some clients (such as the Certicom SSL Plus
4133 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4134 disables the coutermeasure allowing Eudora to connect."
4136 38. Exim was not checking that a write() to a log file succeeded. This could
4137 lead to Bad Things if a log got too big, in particular if it hit a file
4138 size limit. Exim now panics and dies if it cannot write to a log file, just
4139 as it does if it cannot open a log file.
4141 39. Modified OS/Makefile-Linux so that it now contains
4143 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4145 The two -D definitions ensure that Exim is compiled with large file
4146 support, which makes it possible to handle log files that are bigger than
4149 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4150 instance) a domain was checked against a named list that involved a lookup,
4151 causing $domain_data to be set, then another domain was checked against the
4152 same list, then the first domain was re-checked, the value of $domain_data
4153 after the final check could be wrong. In particular, if the second check
4154 failed, it could be set empty. This bug probably also applied to
4157 41. The strip_trailing_dot option was not being applied to the address given
4158 with the -f command-line option.
4160 42. The code for reading a message's header from the spool was incrementing
4161 $received_count, but never initializing it. This meant that the value was
4162 incorrect (doubled) while delivering a message in the same process in which
4163 it was received. In the most common configuration of Exim, this never
4164 happens - a fresh exec is done - but it can happen when
4165 deliver_drop_privilege is set.
4167 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4168 it now includes up to 150 characters of the unexpected data in the log
4171 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4172 and building data strings. The size of both of these buffers was 10 000
4173 bytes - far larger than anybody would *ever* want, thought I. Needless to
4174 say, somebody hit the limit. I have increased the maximum line length to
4175 20 000 and the maximum data length of concatenated lines to 100 000. I have
4176 also fixed two bugs, because there was no checking on these buffers. Tsk,
4177 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4178 buffer is too small.
4180 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4181 lsearch lookups. Now it does.
4183 46. When parsing a route_list item in a manualroute router, a fixed-length
4184 buffer was used for the list of hosts. I made this 1024 bytes long,
4185 thinking that nobody would ever have a list of hosts that long. Wrong.
4186 Somebody had a whole pile of complicated expansion conditions, and the
4187 string was silently truncated, leading to an expansion error. It turns out
4188 that it is easier to change to an unlimited length (owing to other changes
4189 that have happened since this code was originally written) than to build
4190 structure for giving a limitation error. The length of the item that
4191 expands into the list of hosts is now unlimited.
4193 47. The lsearch lookup could not handle data where the length of text line was
4194 more than 4095 characters. Such lines were truncated, leading to shortened
4195 data being returned. It should now handle lines of any length.
4197 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4198 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4201 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4202 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4203 no longer happens in comment lines. A list of replacements is now placed
4204 at the head of all of the source files, except those whose only change is
4205 to replace PERL_COMMAND in the very first #! line.
4207 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4208 messages on the queue, with a bottom-up merge sort, using code contributed
4209 by Michael Haardt. This should make operations like -bp somewhat faster on
4210 large queues. It won't affect queue runners, except when queue_run_in_order
4213 51. Installed eximstats 1.31 in the distribution.
4215 52. Added support for SRV lookups to the dnslookup router.
4217 53. If an ACL referred to $message_body or $message_body_end, the value was not
4218 reset for any messages that followed in the same SMTP session.
4220 54. The store-handling optimization for building very long strings was not
4221 differentiating between the different store pools. I don't think this
4222 actually made any difference in practice, but I've tidied it.
4224 55. While running the routers to verify a sender address, $sender_address
4225 was still set to the sender address. This is wrong, because when routing to
4226 send a bounce to the sender, it would be empty. Therefore, I have changed
4227 it so that, while verifying a sender address, $sender_address is set to <>.
4228 (There is no change to what happens when verifying a recipient address.)
4230 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4231 target A or AAAA records (if not already returned) without resetting the
4232 qualify_single or search_parents options of the DNS resolver. These are
4233 inappropriate in this case because the targets of MX and SRV records must
4234 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4235 target that, when qualified, matched something in the local domain. These
4236 two options are now turned off when doing these lookups.
4238 57. It seems that at least some releases of Reiserfs (which does not have the
4239 concept of a fixed number of inodes) returns zero and not -1 for the
4240 number of available inodes. This interacted badly with check_spool_inodes,
4241 which assumed that -1 was the "no such thing" setting. What I have done is
4242 to check that the total number of inodes is greater than zero before doing
4243 the test of how many are available.
4245 58. When a "warn" ACL statement has a log_message modifier, the message is
4246 remembered, and not repeated. This is to avoid a lot of repetition when a
4247 message has many recipients that cause the same warning to be written.
4248 Howewer, Exim was preserving the list of already written lines for an
4249 entire SMTP session, which doesn't seem right. The memory is now reset if a
4250 new message is started.
4252 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4253 debugging output unless log_rewrite was also set.
4255 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4256 of (int)(handle) when we know that handle contains (void *)(-1).
4258 61. The Exim daemon panic-logs an error return when it closes the incoming
4259 connection. However "connection reset by peer" seems to be common, and
4260 isn't really an error worthy of noting specially, so that particular error
4263 62. When Exim is trying to find all the local interfaces, it used to panic and
4264 die if the ioctl to get the interface flags failed. However, it seems that
4265 on at least one OS (Solaris 9) it is possible to have an interface that is
4266 included in the list of interfaces, but for which you get a failure error
4267 for this call. This happens when the interface is not "plumbed" into a
4268 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4269 failure of the "get flags" call assumes that the interface is down.
4271 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4272 makes life easier for people who are doing arithmetic on fields extracted
4273 from dates, where you often get leading zeros that should not be
4274 interpreted as octal.
4276 64. Added qualify_domain to the redirect router, to override the global
4279 65. If a pathologically long header line contained very many addresses (the
4280 report of this problem mentioned 10 000) and each of them was rewritten,
4281 Exim could use up a very large amount of memory. (It kept on making new
4282 copies of the header line as it rewrote, and never released the old ones.)
4283 At the expense of a bit more processing, the header rewriting function has
4284 been changed so that it no longer eats memory in this way.
4286 66. The generation of the Received: header has been moved from the time that a
4287 message starts to be received, to the time that it finishes. The timestamp
4288 in the Received: header should now be very close to that of the <= log
4289 line. There are two side-effects of this change:
4291 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4292 logged header lines no longer include the local Received: line, because
4293 it has not yet been created. The same applies to a copy of the message
4294 that is returned to a non-SMTP sender when a message is rejected.
4296 (b) When a filter file is tested using -bf, no additional Received: header
4297 is added to the test message. After some thought, I decided that this
4300 This change does not affect the value of $received_for. It is still set
4301 after address rewriting, but before local_scan() is called.
4303 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4305 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4306 gave an unhelpful panic error message, and a defer error. I have managed to
4307 change this behaviour so that it now rejects any supplied certificate,
4308 which seems right, as the list of acceptable certificates is empty.
4310 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4311 gave an unhelpful defer error. I have not managed to make this reject any
4312 supplied certificates, but the error message it gives is "no certificate
4313 supplied", which is not helpful.
4315 70. exigrep's output now also includes lines that are not associated with any
4316 message, but which match the given pattern. Implemented by a patch from
4317 Martin Sluka, which also tidied up the Perl a bit.
4319 71. Recipient callout verification, like sender verification, was using <> in
4320 the MAIL FROM command. This isn't really the right thing, since the actual
4321 sender may affect whether the remote host accepts the recipient or not. I
4322 have changed it to use the actual sender in the callout; this means that
4323 the cache record is now keyed on a recipient/sender pair, not just the
4324 recipient address. There doesn't seem to be a real danger of callout loops,
4325 since a callout by the remote host to check the sender would use <>.
4326 [SEE ABOVE: changed after hitting problems.]
4328 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4329 temporary errors. However, in the case of such a code being given after
4330 the end of a data transmission (i.e. after ".") Exim was failing to write
4331 a retry record for the message. (Yes, there was some broken host that was
4332 actually sending 8xx at this point.)
4334 73. An unknown lookup type in a host list could cause Exim to panic-die when
4335 the list was checked. (An example that provoked this was putting <; in the
4336 middle of a list instead of at the start.) If this happened during a DATA
4337 ACL check, a -D file could be left lying around. This kind of configuration
4338 error no longer causes Exim to die; instead it causes a defer errror. The
4339 incident is still logged to the main and panic logs.
4341 74. Buglet left over from Exim 3 conversion. The message "too many messages
4342 in one connection" was written to the rejectlog but not the mainlog, except
4343 when address rewriting (yes!) was being logged.
4345 75. Added write_rejectlog option.
4347 76. When a system filter was run not as root (that is, when system_filter_user
4348 was set), the values of the $n variables were not being returned to the
4349 main process; thus, they were not subsequently available in the $sn
4352 77. Added +return_path_on_delivery log selector.
4354 78. A connection timeout was being treated differently from recipients deferred
4355 when testing hosts_max_try with a message that was older than the host's
4356 retry timeout. (The host should not be counted, thus allowing all hosts to
4357 be tried at least once before bouncing.) This may have been the cause of an
4358 occasionally reported bug whereby a message would remain on the queue
4359 longer than the retry timeout, but would be bounced if a delivery was
4360 forced. I say "may" because I never totally pinned down the problem;
4361 setting up timeout/retry tests is difficult. See also the next item.
4363 79. The ultimate address timeout was not being applied to errors that involved
4364 a combination of host plus message (for example, a timeout on a MAIL
4365 command). When an address resolved to a number of possible hosts, and they
4366 were not all tried for each delivery (e.g. because of hosts_max_try), a
4367 message could remain on the queue longer than the retry timeout.
4369 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4372 81. Fixed an obscure SMTP outgoing bug which required at least the following
4373 conditions: (a) there was another message waiting for the same server;
4374 (b) the server returned 5xx to all RCPT commands in the first message so
4375 that the message was not completed; (c) the server dropped the connection
4376 or gave a negative response to the RSET that Exim sends to abort the
4377 transaction. The observed case was a dropped connection after DATA that had
4378 been sent in pipelining mode. That is, the server had advertised PIPELINING
4379 but was not implementing it correctly. The effect of the bug was incorrect
4380 behaviour, such as trying another host, and this could lead to a crash.
4386 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4387 and daemon.c were passed as pointers to ints; they should have been
4388 pointers to socklen_t variables (which are typically unsigned ints).
4390 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4393 3. Fixed a really odd bug that affected only the testing scheme; patching a
4394 certain fixed string in the binary changed the value of another string that
4395 happened to be identical to the end of the original first string.
4397 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4398 name", it returns that address as the IP address. On some operating
4399 systems (e.g. Solaris), it also passes back the IP address string as the
4400 "host name". However, on others (e.g. Linux), it passes back an empty
4401 string. Exim wasn't checking for this, and was changing the host name to an
4402 empty string, assuming it had been canonicized.
4404 5. Although rare, it is permitted to have more than one PTR record for a given
4405 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4406 all the names associated with an address, because they do in Solaris.
4407 However, it seems that they do not in Linux for data that comes from the
4408 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4409 I found this out when I moved to a new Linux workstation and tried to run
4410 the Exim test suite.
4412 To get round this problem I have changed the code so that it now does its
4413 own call to the DNS to look up PTR records when searching for a host name.
4414 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4415 addresses that are only in /etc/hosts are still found.
4417 This behaviour is, however, controlled by an option called host_lookup_
4418 order, which defaults to "bydns:byaddr". If people want to use the other
4419 order, or indeed, just use one or the other means of lookup, they can
4420 specify it in this variable.
4422 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4423 some operating systems, this comes back from gethostbyaddr() as an empty
4424 string, and this is what Exim used to test for. However, it seems that in
4425 other systems, "." is yielded. Exim now tests for this case too.
4427 7. The values of check_spool_space and check_log_space are now held internally
4428 as a number of kilobytes instead of an absolute number of bytes. If a
4429 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4430 kilobyte. This means that much larger values can be stored.
4432 8. Exim monitor: an attempt to get the action menu when not actually pointing
4433 at a message produces an empty menu entitled "No message selected". This
4434 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4435 no entries in it ("Shell widget menu has zero width and/or height"). So I
4436 have added a single, blank menu entry in this case.
4438 9. Added ${quote_local_part.
4440 10. MIME decoding is now applied to the contents of Subject: header lines when
4443 11. Now that a reference to $sender_host_address automatically causes a reverse
4444 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4445 host lookup before query-style lookups in lists that might use this
4446 variable. This has therefore been abolished, and the "net-" prefix is no
4447 longer necessary for query-style lookups.
4449 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4450 have been a typo for LFLAGS, so it has been changed.
4452 13. The install script calls Exim with "-C /dev/null" in order to find the
4453 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4454 to be output. Howeve, since Exim outputs its version number before the
4455 error, it didn't break the script. It just looked ugly. I fixed this by
4456 always allowing "-C /dev/null" if the caller is root.
4458 14. Ignore overlarge ACL variable number when reading spool file - insurance
4459 against a later release with more variables having written the file.
4461 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4462 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4463 this, as well as the form without the "IPv6" on the front.
4465 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4466 OS/Makefile-Darwin file.
4468 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4469 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4471 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4472 variable. However, some OS use ints for inodes. I've added cast to long int
4473 to get rid of the compiler warning.
4475 19. I had forgotten to lock out "/../" in configuration file names when
4476 ALT_CONFIG_PREFIX was set.
4478 20. Routers used for verification do not need to specify transports. However,
4479 if such a router generated a host list, and callout was configured, Exim
4480 crashed, because it could not find a port number from the (non-existent)
4481 transport. It now assumes port 25 in this circumstance.
4483 21. Added the -t option to exigrep.
4485 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4486 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4487 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4488 them is set, LOOKUP_LSEARCH is forced.
4490 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4492 24. Added sender and host information to the "rejected by local_scan()" log
4493 line; previously there was no indication of these.
4495 25. Added .include_if_exists.
4497 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4498 Linux. It turns out that not all file systems support this. Apparently some
4499 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4500 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4501 not supported on the file descriptor, is now ignored when Exim is trying to
4502 sync a directory. This applies only to Linux.
4504 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4506 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4507 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4510 29. Added support for maildirsize files from supplied patch (modified a bit).
4512 30. The use of :fail: followed by an empty string could lead Exim to respond to
4513 sender verification failures with (e.g.):
4515 550 Verification failed for <xxx>
4516 550 Sender verify failed
4518 where the first response line was missing the '-' that indicates it is not
4519 the final line of the response.
4521 31. The loop for finding the name of the user that called Exim had a hardwired
4522 limit of 10; it now uses the value of finduser_retries, which is used for
4523 all other user lookups.
4525 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4528 33. Exim was neglecting to zero errno before one call of strtol() when
4529 expanding a string and expecting an integer value. On some systems this
4530 resulted in spurious "integer overflow" errors. Also, it was casting the
4531 result into an int without checking.
4533 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4534 did not work. The code looks as if it has *never* worked, though it appears
4535 to have been documented since at least releast 1.62. I have made it work.
4537 35. The "timeout_DNS" error in retry rules, also documented since at least
4538 1.62, also never worked. As it isn't clear exactly what this means, and
4539 clearly it isn't a major issue, I have abolished the feature by treating it
4540 as "timeout", and writing a warning to the main and panic logs.
4542 36. The display of retry rules for -brt wasn't always showing the error code
4545 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4546 timeout_connect_A, timeout_connect_MX.
4548 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4549 to the empty sender.
4551 39. The daemon was not analysing the content of -oX till after it had closed
4552 stderr and disconnected from the controlling terminal. This meant that any
4553 syntax errors were only noted on the panic log, and the return code from
4554 the command was 0. By re-arranging the code a little, I've made the
4555 decoding happen first, so such errors now appear on stderr, and the return
4556 code is 1. However, the actual setting up of the sockets still happens in
4557 the disconnected process, so errors there are still only recorded on the
4560 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4561 connections (as happens on some IP stacks) was logged at start up time as
4562 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4563 it from "IPv6 and IPv4", which means that two separate sockets are being
4566 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4567 says whether AF_INET or AF_INET6 was passed as an argument.
4569 42. Exiwhat output was messed up when time zones were included in log
4572 43. Exiwhat now gives more information about the daemon's listening ports,
4573 and whether -tls-on-connect was used.
4575 44. The "port" option of the smtp transport is now expanded.
4577 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4578 silently ignored. Now an error message is written to the main and panic
4581 46. There's a new ACL modifier called "logwrite" which writes to a log file
4582 as soon as it is encountered.
4584 47. Added $local_user_uid and $local_user_gid at routing time.
4586 48. Exim crashed when trying to verify a sender address that was being
4589 49. Exim was recognizing only a space character after ".include". It now also
4590 recognizes a tab character.
4592 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4593 extracting the relevant information from the specification. The man page no
4594 longer contains scrambled data for the -d option, and I've added a section
4595 at the front about calling Exim under different names.
4597 51. Added "extra_headers" argument to the "mail" command in filter files.
4599 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4602 53. Installed eximstats 1.29.
4604 54. Added transport_filter_timeout as a generic transport option.
4606 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4607 Cc: header lines. This was required by RFC 822, but it not required by RFC
4610 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4611 incoming messages that did not have them. Now it does so only if the
4612 message originates locally, that is, if there is no associated remote host
4613 address. When Resent- header lines are present, this applies to the Resent-
4614 lines rather than the non-Resent- lines.
4616 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4617 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4619 58. Messages for configuration errors now include the name of the main
4620 configuration file - useful now that there may be more than one file in a
4621 list (.included file names were always shown).
4623 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4624 for those rare installations that do not start the daemon as root or run it
4625 setuid root. I've cut out the call to initgroups() if the daemon is not
4628 60. The Exim user and group can now be bound into the binary as text strings
4629 that are looked up at the start of Exim's processing.
4631 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4633 62. Added $mailstore_basename variable.
4635 63. Installed patch to sieve.c from Michael Haardt.
4637 64. When Exim failed to open the panic log after failing to open the main log,
4638 the original message it was trying to log was written to stderr and debug
4639 output, but if they were not available (the usual case in production), it
4640 was lost. Now it is written to syslog before the two lines that record the
4641 failures to open the logs.
4643 65. Users' Exim filters run in subprocesses under the user's uid. It is
4644 possible for a "deliver" command or an alias in a "personal" command to
4645 provoke an address rewrite. If logging of address rewriting is configured,
4646 this fails because the process is not running as root or exim. There may be
4647 a better way of dealing with this, but for the moment (because 4.30 needs
4648 to be released), I have disabled address rewrite logging when running a
4649 filter in a non-root, non-exim process.
4655 1. The buildconfig auxiliary program wasn't quoting the value set for
4656 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4657 not defined. This bug was masked in 4.22 by the effect that was fixed in
4660 2. Some messages that were rejected after a message id was allocated were
4661 shown as "incomplete" by exigrep. It no longer does this for messages that
4662 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4664 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4665 have allow_domain_literals set, the ID did not get logged in the <= line.
4666 Domain literals are now always recognized in Message-ID: header lines.
4668 4. The first argument for a ${extract expansion item is the key name or field
4669 number. Leading and trailing spaces in this item were not being ignored,
4670 causing some misleading effects.
4672 5. When deliver_drop_privilege was set, single queue runner processes started
4673 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4674 same command in the process it spins off) were not dropping privilege.
4676 6. When the daemon running as "exim" started a queue runner, it always
4677 re-executed Exim in the spun-off process. This is a waste of effort when
4678 deliver_drop_privilege is set. The new process now just calls the
4679 queue-runner function directly.
4685 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4688 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4689 ignored. Though the use of -oP was forcing the writing of a pid file, it
4690 was always written to the default place.
4692 3. If the message "no IP address found for host xxxx" is generated during
4693 incoming verification, it is now followed by identification of the incoming
4694 connection (so you can more easily find what provoked it).
4696 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4698 5. Added some features to "harden" Exim a bit more against certain attacks:
4700 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4701 be put in Local/Makefile. This is like the never_users runtime option,
4702 but it cannot be overridden. The default setting is "root".
4704 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4705 prefix string with which any file named in a -C command line option
4708 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4709 is retained for -C and -D only if the caller of Exim is root. Without
4710 it, the exim user may also use -C and -D and retain privilege.
4712 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4713 command line option is disabled.
4715 6. Macro names set by the -D option must start with an upper case letter, just
4716 like macro names defined in the configuration file.
4718 7. Added "dereference=" facility to LDAP.
4720 8. Two instances of the typo "uknown" in the source files are fixed.
4722 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4723 the Configure-Makefile script screwed up while processing it.
4725 10. Incorporated PCRE 4.4.
4727 11. The SMTP synchronization check was not operating right at the start of an
4728 SMTP session. For example, it could not catch a HELO sent before the client
4729 waited for the greeting. There is now a check for outstanding input at the
4730 point when the greeting is written. Because of the duplex, asynchronous
4731 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4732 way, but not yet received, when the check is performed.
4734 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4735 on TCP/IP sockets, because this apparently causes some broken clients to
4738 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4739 unchanged) from the Cygwin maintainer.
4741 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4742 was supported instead of testing for mbx. Effectively a typo.
4744 15. The spa authenticator server code was not checking that the input it
4745 received was valid base64.
4747 16. The debug output line for the "set" modifier in ACLs was not showing the
4748 name of the variable that was being set.
4750 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4752 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4753 lookup on its own. Something else was needed to trigger the lookup. For
4754 example, a match in host_lookup or the need for a host name in a host list.
4755 Now, if $sender_host_name is referenced and the host name has not yet been
4756 looked up, a lookup is performed. If the lookup fails, the variable remains
4757 empty, and $host_lookup_failed is set to "1".
4759 19. Added "eqi" as a case-independent comparison operator.
4761 20. The saslauthd authentication condition could segfault if neither service
4762 nor realm was specified.
4764 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4765 error message that was logged was misleading, and incoming SMTP
4766 connections were dropped. The message is now more accurate, and temporary
4767 errors are given to SMTP connections.
4769 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4770 (except RSET) with a 421 error, until QUIT is received. However, it was
4771 failing to send a response to QUIT.
4773 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4774 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4775 ACL is now run after the helo_try_verify_hosts code.
4777 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4778 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4779 case-independent, so other case variants are also recognized). Apparently
4780 some systems use these upper case variants.
4782 25. If more than two messages were waiting for the same host, and a transport
4783 filter was specified for the transport, Exim sent two messages over the
4784 same TCP/IP connection, and then failed with "socket operation on non-
4785 socket" when it tried to send the third.
4787 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4789 27. The extern definition of crypt16() in expand.c was not being excluded when
4790 the OS had its own crypt16() function.
4792 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4793 as a preferred synonym for return_size_limit, both as an option and as an
4796 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4798 30. Changed the default configuration ACL to relax the local part checking rule
4799 for addresses that are not in any local domains. For these addresses,
4800 slashes and pipe symbols are allowed within local parts, but the sequence
4801 /../ is explicitly forbidden.
4803 31. SPA server authentication was not clearing the challenge buffer before
4806 32. log_message in a "warn" ACL statement was writing to the reject log as
4807 well as to the main log, which contradicts the documentation and doesn't
4808 seem right (because no rejection is happening). So I have stopped it.
4810 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4811 However, I am unable to do any testing of this.
4813 34. Fixed an infelicity in the appendfile transport. When checking directories
4814 for a mailbox, to see if any needed to be created, it was accidentally
4815 using path names with one or more superfluous leading slashes; tracing
4816 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4818 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4819 contents are added to the log line that is written for every discarded
4820 recipient. (Previously a log_message setting was ignored.)
4822 36. The ${quote: operator now quotes the string if it is empty.
4824 37. The install script runs exim in order to find its version number. If for
4825 some reason other than non-existence or emptiness, which it checks, it
4826 could not run './exim', it was installing it with an empty version number,
4827 i.e. as "exim-". This error state is now caught, and the installation is
4830 38. An argument was missing from the function that creates an error message
4831 when Exim fails to connect to the socket for saslauthd authentication.
4832 This could cause Exim to crash, or give a corrupted message.
4834 39. Added isip, isip4, and isip6 to ${if conditions.
4836 40. The ACL variables $acl_xx are now saved with the message, and can be
4837 accessed later in routers, transports, and filters.
4839 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4840 strings in the file are not string-expanded.
4842 42. If a MAIL command specified a SIZE value that was too large to fit into an
4843 int variable, the check against message_size_limit failed. Such values are
4844 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4845 day this will have to be increased, but I don't think I want to be around
4846 when emails are that large.
4853 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4854 iconv() is not standard in FreeBSD.
4856 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4857 IPv6 enabled. The observed symptom was a segmentation fault on return from
4858 the function os_common_find_running_interfaces() in src/os.c.
4860 3. In the check_special_case() function in daemon.c I had used "errno" as an
4861 argument name, which causes warnings on some systems. This was basically a
4862 typo, since it was named "eno" in the comments!
4864 4. The code that waits for the clock to tick (at a resolution of some fraction
4865 of a second) so as to ensure message-id uniqueness was always waiting for
4866 at least one whole tick, when it could have waited for less. [This is
4867 almost certainly not relevant at current processor speeds, where it is
4868 unlikely to ever wait at all. But we try to future-proof.]
4870 5. The function that sleeps for a time interval that includes fractions of a
4871 second contained a race. It did not block SIGALRM between setting the
4872 timer, and suspending (a couple of lines later). If the interval was short
4873 and the sigsuspend() was delayed until after it had expired, the suspension
4874 never ended. On busy systems this could lead to processes getting stuck for
4877 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4878 process, before it forks any delivery processes. The open lookup caching
4879 mechanism meant that the open file or database connection was passed into
4880 the delivery process. The problem was that delivery processes always tidy
4881 up cached lookup data. This could cause a problem for the next delivery
4882 process started by the queue runner, because the external queue runner
4883 process does not know about the closure. So the next delivery process
4884 still has data in the lookup cache. In the case of a file lookup, there was
4885 no problem because closing a file descriptor in a subprocess doesn't affect
4886 the parent. However, if the lookup was caching a connection to a database,
4887 the connection was closed, and the second delivery process was likely to
4888 see errors such as "PGSQL: query failed: server closed the connection
4889 unexpectedly". The problem has been fixed by closing all cached lookups
4890 in a queue runner before running a delivery process.
4892 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4893 seem to have the "const" qualifier which it has on other OS. I've
4896 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4897 *of the same type* (client or server) with the same public name that an
4898 error should be diagnosed.
4900 9. When Exim looked up a host name for an IP address, but failed to find the
4901 original IP address when looking up the host name (a safety check), it
4902 output the message "<ip address> does not match any IP for NULL", which was
4903 confusing, to say the least. The bug was that the host name should have
4904 appeared instead of "NULL".
4906 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4907 user that is built into the binary, and the -C or -D options is used, root
4908 privilege is dropped before the configuration file is read. In addition,
4909 logging is switched to stderr instead of the normal log files. If the
4910 configuration then re-defines the Exim user, the unprivileged environment
4911 is probably not what is expected, so Exim logs a panic warning message (but
4914 However, if deliver_drop_privilege is set, the unprivileged state may well
4915 be exactly what is intended, so the warning has been cut out in that case,
4916 and Exim is allowed to try to write to its normal log files.
4922 1. smtp_return_error_details was not giving details for temporary sender
4923 or receiver verification errors.
4925 2. Diagnose a configuration error if two authenticators have the same public
4928 3. Exim used not to create the message log file for a message until the first
4929 delivery attempt. This could be confusing when incoming messages were held
4930 for policy or load reasons. The message log file is now created at the time
4931 the message is received, and an initial "Received" line is written to it.
4933 4. The automatically generated man page for command line options had a minor
4934 bug that caused no ill effects; however, a more serious problem was that
4935 the procedure for building the man page automatically didn't always
4936 operate. Consequently, release 4.20 contains an out-of-date version. This
4937 shouldn't happen again.
4939 5. When building Exim with embedded Perl support, the script that builds the
4940 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4941 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4943 6. The freeze_tell option was not being used for messages that were frozen on
4944 arrival, either by an ACL or by local_scan().
4946 7. Added the smtp_incomplete_transaction log selector.
4948 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4949 was accepting AUTH without a new EHLO.
4951 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4952 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4955 10. Logging of TCP/IP connections (when configured) now happens in the main
4956 daemon process instead of the child process, so that the TCP/IP connection
4957 count is more accurate (but it can never be perfect).
4959 11. The use of "drop" in a nested ACL was not being handled correctly in the
4960 outer ACL. Now, if condition failure induced by the nested "drop" causes
4961 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4962 or "require"), the connection is dropped.
4964 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4965 that yield "discard" can now be used with an "accept" or a "discard" verb,
4966 but an error is generated for any others (because I can't see a useful way
4967 to define what should happen).
4969 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4970 are now processed in the same way as lines in the Exim configuration file.
4971 In particular, continuation lines are supported.
4973 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4975 15. Added -ti meaning -t -i.
4977 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4978 domains, and warn by logging if others are found.
4980 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4981 returned from GIFCONF when Exim is trying to find the list of interfaces on
4982 a host. The code in os.c has been modified to copy each ifreq to an aligned
4983 structure in all cases.
4985 Also, in some cases, the returned ifreq's were being copied to a 'struct
4986 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4987 means the last couple of bytes of an IPv6 address could be chopped if the
4988 ifreq contained only a normal sockaddr (14 bytes storage).
4990 18. Named domain lists were not supported in the hosts_treat_as_local option.
4991 An entry such as +xxxx was not recognized, and was treated as a literal
4994 19. Ensure that header lines added by a DATA ACL are included in the reject log
4995 if the ACL subsequently rejects the message.
4997 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4998 MD5 (which is deprecated).
5000 21. When testing a filter file using -bf, Exim was writing a message when it
5001 took the sender from a "From " line in the message, but it was not doing so
5002 when it took $return_path from a Return-Path: header line. It now does.
5004 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5005 with a valid header line field name (a series of printing characters
5006 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5008 23. Changed "disc" in the source to "disk" to conform to the documentation and
5009 the book and for uniformity.
5011 24. Ignore Sendmail's -Ooption=value command line item.
5013 25. When execve() failed while trying to run a command in a pipe transport,
5014 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5015 could be confused with a return value of 69 from the command itself. This
5016 has been changed to 127, the value the shell returns if it is asked to run
5017 a non-existent command. The wording for the related log line suggests a
5018 non-existent command as the problem.
5020 26. If received_header_text expands to an empty string, do not add a Received:
5021 header line to the message. (Well, it adds a token one on the spool, but
5022 marks it "old" so that it doesn't get used or transmitted.)
5024 27. Installed eximstats 1.28 (addition of -nt option).
5026 28. There was no check for failure on the call to getsockname() in the daemon
5027 code. This can fail if there is a shortage of resources on the system, with
5028 ENOMEM, for example. A temporary error is now given on failure.
5030 29. Contrary to the C standard, it seems that in some environments, the
5031 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5032 program. Exim now does this explicitly; it affects the formatting of
5033 timestamps using strftime().
5035 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5036 complaints. I've now initialized all the variables, to avoid this.
5038 32. Header lines added by a system filter were not being "seen" during
5039 transport-time rewrites.
5041 33. The info_callback() function passed to OpenSSL is set up with type void
5042 (*)(SSL *, int, int), as described somewhere. However, when calling the
5043 function (actually a macro) that sets it up, the type void(*)() is
5044 expected. I've put in a cast to prevent warnings from picky compilers.
5046 34. If a DNS black list lookup found a CNAME record, but there were no A
5047 records associated with the domain it pointed at, Exim crashed.
5049 35. If a DNS black list lookup returned more than one A record, Exim ignored
5050 all but the first. It now scans all returned addresses if a particular IP
5051 value is being sought. In this situation, the contents of the
5052 $dnslist_value variable are a list of all the addresses, separated by a
5055 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5056 to accept a host name and all its aliases if the forward lookup for any of
5057 them yielded the IP address of the incoming connection. Now it accepts only
5058 those names whose forward lookup yields the correct IP address. Any other
5059 names are discarded. This closes a loophole whereby a rogue DNS
5060 administrator could create reverse DNS records to break through a
5061 wildcarded host restriction in an ACL.
5063 37. If a user filter or a system filter that ran in a subprocess used any of
5064 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5065 the wrong values were passed to the pipe command ($thisaddress had the
5066 value of $0, $0 had the value of $1, etc). This bug was introduced by
5067 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5069 38. Improved the line breaking for long SMTP error messages from ACLs.
5070 Previously, if there was no break point between 40 and 75 characters, Exim
5071 left the rest of the message alone. Two changes have been made: (a) I've
5072 reduced the minimum length to 35 characters; (b) if it can't find a break
5073 point between 35 and 75 characters, it looks ahead and uses the first one
5074 that it finds. This may give the occasional overlong line, but at least the
5075 remaining text gets split now.
5077 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5078 file descriptors might be low, and that setting 1000 would always raise it.
5079 It turns out that in some environments, the limit is already over 1000 and
5080 that lowering it causes trouble. So now Exim takes care not to decrease it.
5082 40. When delivering a message, the value of $return_path is set to $sender_
5083 address at the start of routing (routers may change the value). By an
5084 oversight, this default was not being set up when an address was tested by
5085 -bt or -bv, which affected the outcome if any router or filter referred to
5088 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5089 log only when "message" or "log_message" are set. However, if one of the
5090 conditions was an address verification, or a call to a nested ACL, the
5091 messages generated by the underlying test were being passed through. This
5092 no longer happens. The underlying message is available in $acl_verify_
5093 message for both "message" and "log_message" expansions, so it can be
5094 passed through if needed.
5096 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5097 new expansion $bh_ to give the encoded byte string without charset
5098 translation. Translation happens only if iconv() is available; HAVE_ICONV
5099 indicates this at build time. HEADERS_CHARSET gives the charset to
5100 translate to; headers_charset can change it in the configuration, and
5101 "headers charset" can change it in an individual filter file.
5103 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5104 that creates RFC 2047 encoded "words" labels them as that charset instead
5105 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5106 expansion operator; (ii) when Exim creates a From: line for a local
5107 message; (iii) when a header line is rewritten to include a "phrase" part.
5109 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5110 buggy, causing it to skip the first lines of messages whose message ID
5111 ended in 'D'. This would not have bitten before Exim release 4.14, because
5112 message IDs were unlikely to end in 'D' before then. The effect was to have
5113 incorrect size information for certain domains.
5115 45. #include "config.h" was missing at the start of the crypt16.c module. This
5116 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5119 46. If there was a timeout during a "random" callout check, Exim treated it as
5120 a failure of the random address, and carried on sending RSET and the real
5121 address. If the delay was just some slowness somewhere, the response to the
5122 original RCPT would be taken as a response to RSET and so on, causing
5123 mayhem of various kinds.
5125 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5126 when I implemented it. It didn't allow for the fact that some option values
5127 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5128 the right test for positive values.
5130 48. Domain names in DNS records are case-independent. Exim always looks them up
5131 in lower case. Some resolvers return domain names in exactly the case they
5132 appear in the zone file, that is, they may contain uppercase letters. Not
5133 all resolvers do this - some return always lower case. Exim was treating a
5134 change of case by a resolver as a change of domain, similar to a widening
5135 of a domain abbreviation. This triggered its re-routing code and so it was
5136 trying to route what was effectively the same domain again. This normally
5137 caused routing to fail (because the router wouldn't handle the domain
5138 twice). Now Exim checks for this case specially, and just changes the
5139 casing of the domain that it ultimately uses when it transmits the message
5142 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5145 50. If a filter generated a file delivery with a non-absolute name (possible if
5146 no home directory exists for the router), the forbid_file option was not
5149 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5150 the existing equality matching.
5152 52. Exim was using ints instead of ino_t variables in some places where it was
5153 dealing with inode numbers.
5155 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5156 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5157 called TMPDIR, and if it finds it is different, it changes its value.
5159 54. The smtp_printf() function is now made available to local_scan() so
5160 additional output lines can be written before returning. There is also an
5161 smtp_fflush() function to enable the detection of a dropped connection.
5162 The variables smtp_input and smtp_batched_input are exported to
5165 55. Changed the default runtime configuration: the message "Unknown user"
5166 has been removed from the ACL, and instead placed on the localuser router,
5167 using the cannot_route_message feature. This means that any verification
5168 failures that generate their own messages won't get overridden. Similarly,
5169 the "Unrouteable address" message that was in the ACL for unverifiable
5170 relay addresses has also been removed.
5172 56. Added hosts_avoid_esmtp to the smtp transport.
5174 57. The exicyclog script was not checking for the esoteric option
5175 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5176 will work only if exicyclog is run under the appropriate euid.
5178 58. Following a discussion on the list, the rules by which Exim recognises line
5179 endings on incoming messages have been changed. The -dropcr and drop_cr
5180 options are now no-ops, retained only for backwards compatibility. The
5181 following line terminators are recognized: LF CRLF CR. However, special
5182 processing applies to CR:
5184 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5185 nor a local message in the state where . is a terminator.
5187 (ii) If a bare CR is encountered in a header line, an extra space is added
5188 after the line terminator so as not to end the header. The reasoning
5189 behind this is that bare CRs in header lines are most likely either
5190 to be mistakes, or people trying to play silly games.
5192 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5193 was being incorrectly given as 18 bytes larger than it should have been.
5194 This is a VOB (very old bug).
5196 60. This may never have affected anything current, but just in case it has:
5197 When the local host is found other than at the start of a list of hosts,
5198 the local host, those with the same MX, and any that follow, are discarded.
5199 When the list in question was part of a longer list of hosts, the following
5200 hosts (not currently being processed) were also being discarded. This no
5201 longer happens. I'm not sure if this situation could ever has previously
5204 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5207 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5208 header. This is something that is recommended in a new Internet Draft, and
5209 is something that is documented as being done by Sendmail. There are two
5210 possible values. For messages generated by the autoreply transport, Exim
5213 Auto-Submitted: auto-replied
5215 whereas for all other generated messages (e.g. bounces) it adds
5217 Auto-Submitted: auto-generated
5219 63. The "personal" condition in filters now includes a test for the
5220 Auto-Submitted: header. If it contains the string "auto-" the message it
5221 not considered personal.
5223 64. Added rcpt_include_affixes as a generic transport option.
5225 65. Added queue_only_override (default true).
5227 66. Added the syslog_duplication option.
5229 67. If what should have been the first header line of a message consisted of
5230 a space followed by a colon, Exim was mis-interpreting it as a header line.
5231 It isn't of course - it is syntactically invalid and should therefore be
5232 treated as the start of the message body. The misbehaviour could have
5233 caused a number of strange effects, including loss of data in subsequent
5234 header lines, and spool format errors.
5236 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5237 client host had authenticated. This control can now be exercised by an ACL
5238 for more flexibility.
5240 69. By default, callouts do not happen when testing with -bh. There is now a
5241 variant, -bhc, which does actually run the callout code, including
5242 consulting and updating the callout cache.
5244 70. Added support for saslauthd authentication, courtesy of Alexander
5247 71. If statvfs() failed on the spool or log directories while checking their
5248 size for availability, Exim confusingly gave the error "space shortage".
5249 Furthermore, in debugging mode it crashed with a floating point exception.
5250 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5251 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5252 serious problem, Exim now writes to the main and panic logs when this
5253 happens, with details of the failure. It then refuses to accept the
5254 incoming message, giving the message "spool directory problem" or "log
5255 directory problem" with a 421 code for SMTP messages.
5257 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5258 0, 1, and 2 exist, because some OS complain for execs without them (see
5259 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5260 descriptors. However, the code omitted to check that the open succeeded,
5261 causing mysterious errors if for some reason the permissions on /dev/null
5262 got screwed. Now Exim writes a message to the main and panic logs, and
5263 bombs out if it can't open /dev/null.
5265 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5266 interact so that it is all more flexible. It is supposed to remain
5267 backwards compatible. Also added extra_local_interfaces.
5269 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5270 to bomb out with an assertion failure - to the client this appears as a
5271 connection drop. This problem occurs in the part of the code that was taken
5272 from the Samba project. Fortunately, the assertion is in a very simple
5273 function, so I have fixed this by reproducing the function inline in the
5274 one place where it is called, and arranging for authentication to fail
5275 instead of killing the process with assert().
5277 75. The SPA client code was not working when the server requested OEM rather
5278 than Unicode encoding.
5280 76. Added code to make require_files with a specific uid setting more usable in
5281 the case where statting the file as root fails - usually a non-root-mounted
5282 NFS file system. When this happens and the failure is EACCES, Exim now
5283 forks a subprocess and does the per-uid checking as the relevant uid.
5285 77. Added process_log_path.
5287 78. If log_file_path was not explicitly set, a setting of check_log_space or
5288 check_log_inodes was ignored.
5290 79. If a space check for the spool or log partitions fails, the incident is now
5291 logged. Of course, in the latter case the data may get lost...
5293 80. Added the %p formatting code to string_format() so that it can be used to
5294 print addresses in debug_print(). Adjusted all the address printing in the
5295 debugging in store.c to use %p rather than %d.
5297 81. There was a concern that a line of code in smtp_in.c could overflow a
5298 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5299 initially expressed, the concern was not well-founded, because trailing
5300 spaces are removed early. However, if the trailing spaces were followed by
5301 a NULL, they did not get removed, so the overflow was possible. Two fixes
5304 (a) I re-wrote the offending code in a cleaner fashion.
5305 (b) If an incoming SMTP command contains a NULL character, it is rejected
5308 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5309 runs initgroups(), so that if the Exim user is in any additional groups,
5310 they will be used during message reception.
5316 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob