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
83 TF/01 Failure to get a lock on a hints database can have serious
84 consequences so log it to the panic log.
86 TF/02 Log LMTP confirmation messages in the same way as SMTP,
87 controlled using the smtp_confirmation log selector.
89 TF/03 Include the error message when we fail to unlink a spool file.
91 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
92 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
93 for maintaining out-of-tree patches for some time.
95 PP/01 Bugzilla 139: Documentation and portability issues.
96 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
97 Handle per-OS dynamic-module compilation flags.
99 PP/02 Let /dev/null have normal permissions.
100 The 4.73 fixes were a little too stringent and complained about the
101 permissions on /dev/null. Exempt it from some checks.
102 Reported by Andreas M. Kirchwitz.
104 PP/03 Report version information for many libraries, including
105 Exim version information for dynamically loaded libraries. Created
106 version.h, now support a version extension string for distributors
107 who patch heavily. Dynamic module ABI change.
109 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
110 privilege escalation vulnerability whereby the Exim run-time user
111 can cause root to append content of the attacker's choosing to
114 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
117 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
118 If dropping privileges for untrusted macros, we disabled normal logging
119 on the basis that it would fail; for the Exim run-time user, this is not
120 the case, and it resulted in successful deliveries going unlogged.
121 Fixed. Reported by Andreas Metzler.
127 PP/01 Date: & Message-Id: revert to normally being appended to a message,
128 only prepend for the Resent-* case. Fixes regression introduced in
129 Exim 4.70 by NM/22 for Bugzilla 607.
131 PP/02 Include check_rfc2047_length in configure.default because we're seeing
132 increasing numbers of administrators be bitten by this.
134 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
136 PP/03 Bugzilla 994: added openssl_options main configuration option.
138 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
140 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
142 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
144 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
145 without a peer certificate, leading to a segfault because of an
146 assumption that peers always have certificates. Be a little more
147 paranoid. Problem reported by Martin Tscholak.
149 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
150 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
151 NB: ClamAV planning to remove STREAM in "middle of 2010".
152 CL also introduces -bmalware, various -d+acl logging additions and
153 more caution in buffer sizes.
155 PP/09 Implemented reverse_ip expansion operator.
157 PP/10 Bugzilla 937: provide a "debug" ACL control.
159 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
161 PP/12 Bugzilla 973: Implement --version.
163 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
165 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
167 PP/15 Bugzilla 816: support multiple condition rules on Routers.
169 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
170 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
171 ignore trailing whitespace.
173 JJ/02 prevent non-panic DKIM error from being sent to paniclog
175 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
178 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
179 Notification from Dr Andrew Aitchison.
181 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
182 ExtendedDetectionInfo response format.
183 Notification from John Horne.
185 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
188 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
189 XSL and documented dependency on system catalogs, with examples of how
192 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
195 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
196 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
197 configuration file which is writeable by the Exim user or group.
199 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
200 of configuration files to cover files specified with the -C option if
201 they are going to be used with root privileges, not just the default
204 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
205 option (effectively making it always true).
207 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
208 files to be used while preserving root privileges.
210 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
211 that rogue child processes cannot use them.
213 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
214 run-time user, instead of root.
216 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
217 Exim run-time user without dropping privileges.
219 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
220 result string, instead of calling string_vformat() twice with the same
223 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
224 for other users. Others should always drop root privileges if they use
225 -C on the command line, even for a whitelisted configure file.
227 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
229 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
235 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
236 $data_path, and $header_path variables; fixed documentation bugs and
239 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
240 exipick to access non-standard spools, including the "frozen" queue
243 NM/01 Bugzilla 965: Support mysql stored procedures.
244 Patch from Alain Williams
246 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
248 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
249 Patch from Andreas Metzler
251 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
252 Patch from Kirill Miazine
254 NM/05 Bugzilla 671: Added umask to procmail example.
256 JJ/03 installed exipick 20100323.0, fixing doc bug
258 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
259 directory. Notification and patch from Dan Rosenberg.
261 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
263 TK/02 Improve log output when DKIM signing operation fails.
265 MH/01 Treat the transport option dkim_domain as a colon separated
266 list, not as a single string, and sign the message with each element,
267 omitting multiple occurences of the same signer.
269 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
270 Bugzilla 985, 986. Patch by Simon Arlott
272 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
273 Patch by Simon Arlott
275 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
276 MBX locking. Notification from Dan Rosenberg.
282 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
284 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
286 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
288 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
290 NM/04 Bugzilla 915: Flag broken perl installation during build.
296 TK/01 Added patch by Johannes Berg that expands the main option
297 "spamd_address" if it starts with a dollar sign.
299 TK/02 Write list of recipients to X-Envelope-Sender header when building
300 the mbox-format spool file for content scanning (suggested by Jakob
303 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
304 (http://www.dcc-servers.net/) support via dccifd. Activated by
305 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
307 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
308 by Mark Daniel Reidel <mr@df.eu>.
310 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
311 When building exim an external PCRE library is now needed -
312 PCRE is a system library on the majority of modern systems.
313 See entry on PCRE_LIBS in EDITME file.
315 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
316 conversation. Added nologin parameter to request.
317 Patch contributed by Kirill Miazine.
319 TF/01 Do not log submission mode rewrites if they do not change the address.
321 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
323 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
324 log files in place. Contributed by Roberto Lima.
326 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
328 TF/03 Bugzilla 615: When checking the local_parts router precondition
329 after a local_part_suffix or local_part_prefix option, Exim now
330 does not use the address's named list lookup cache, since this
331 contains cached lookups for the whole local part.
333 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
334 Robert Millan. Documentation is in experimental-spec.txt.
336 TF/04 Bugzilla 668: Fix parallel build (make -j).
338 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
340 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
341 Patch provided by Jan Srzednicki.
343 TF/05 Leading white space used to be stripped from $spam_report which
344 wrecked the formatting. Now it is preserved.
346 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
347 that they are available at delivery time.
349 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
351 TF/08 TLS error reporting now respects the incoming_interface and
352 incoming_port log selectors.
354 TF/09 Produce a more useful error message if an SMTP transport's hosts
355 setting expands to an empty string.
357 NM/06 Bugzilla 744: EXPN did not work under TLS.
358 Patch provided by Phil Pennock.
360 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
361 Patch provided by Richard Godbee.
363 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
364 acl_smtp_notquit, added index entry.
366 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
367 Patch provided by Eugene Bujak.
369 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
370 accept(). Patch provided by Maxim Dounin.
372 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
373 Patch provided by Phil Pennock.
375 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
377 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
378 Patch provided by Brad "anomie" Jorsch.
380 NM/14 Bugzilla 622: Added timeout setting to transport filter.
381 Patch provided by Dean Brooks.
383 TK/05 Add native DKIM support (does not depend on external libraries).
385 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
386 Patch provided by Graeme Fowler.
388 NM/16 Bugzilla 851: Documentation example syntax fix.
390 NM/17 Changed NOTICE file to remove references to embedded PCRE.
392 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
395 NM/19 Bugzilla 745: TLS version reporting.
396 Patch provided by Phil Pennock.
398 NM/20 Bugzilla 167: bool: condition support.
399 Patch provided by Phil Pennock.
401 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
402 clients. Patch provided by Phil Pennock.
404 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
405 Patch provided by Brad "anomie" Jorsch.
407 NM/23 Bugzilla 687: Fix misparses in eximstats.
408 Patch provided by Heiko Schlittermann.
410 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
411 Patch provided by Heiko Schlittermann.
413 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
414 plus update to original patch.
416 NM/26 Bugzilla 799: Documentation correction for ratelimit.
418 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
419 Patch provided by David Brownlee.
421 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
423 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
425 NM/30 Bugzilla 888: TLS documentation bugfixes.
427 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
429 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
430 Unlike the original bugzilla I have changed all shell scripts in src tree.
432 NM/33 Bugzilla 898: Transport filter timeout fix.
433 Patch by Todd Rinaldo.
435 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
436 Patch by Serge Demonchaux.
438 NM/35 Bugzilla 39: Base64 decode bug fixes.
439 Patch by Jakob Hirsch.
441 NM/36 Bugzilla 909: Correct connect() call in dcc code.
443 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
445 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
447 NM/39 Bugzilla 911: Fixed MakeLinks build script.
453 TK/01 Add preliminary DKIM support. Currently requires a forked version of
454 ALT-N's libdkim that I have put here:
455 http://duncanthrax.net/exim-experimental/
457 Note to Michael Haardt: I had to rename some vars in sieve.c. They
458 were called 'true' and it seems that C99 defines that as a reserved
459 keyword to be used with 'bool' variable types. That means you could
460 not include C99-style headers which use bools without triggering
461 build errors in sieve.c.
463 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
464 as mailq or other aliases. Changed the --help handling significantly
465 to do whats expected. exim_usage() emits usage/help information.
467 SC/01 Added the -bylocaldomain option to eximstats.
469 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
471 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
473 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
479 PH/01 Another patch from the Sieve maintainer.
481 PH/02 When an IPv6 address is converted to a string for single-key lookup
482 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
483 dots are used instead of colons so that keys in lsearch files need not
484 contain colons. This was done some time before quoting was made available
485 in lsearch files. However, iplsearch files do require colons in IPv6 keys
486 (notated using the quote facility) so as to distinguish them from IPv4
487 keys. This meant that lookups for IP addresses in host lists did not work
488 for iplsearch lookups.
490 This has been fixed by arranging for IPv6 addresses to be expressed with
491 colons if the lookup type is iplsearch. This is not incompatible, because
492 previously such lookups could never work.
494 The situation is now rather anomolous, since one *can* have colons in
495 ordinary lsearch keys. However, making the change in all cases is
496 incompatible and would probably break a number of configurations.
498 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
501 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
502 conversion specification without a maximum field width, thereby enabling
503 a rogue spamd server to cause a buffer overflow. While nobody in their
504 right mind would setup Exim to query an untrusted spamd server, an
505 attacker that gains access to a server running spamd could potentially
506 exploit this vulnerability to run arbitrary code as the Exim user.
508 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
509 $primary_hostname instead of what libspf2 thinks the hosts name is.
511 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
512 a directory entry by the name of the lookup key. Previously, if a
513 symlink pointed to a non-existing file or a file in a directory that
514 Exim lacked permissions to read, a lookup for a key matching that
515 symlink would fail. Now it is enough that a matching directory entry
516 exists, symlink or not. (Bugzilla 503.)
518 PH/03 The body_linecount and body_zerocount variables are now exported in the
521 PH/04 Added the $dnslist_matched variable.
523 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
524 This means they are set thereafter only if the connection becomes
527 PH/06 Added the client_condition to authenticators so that some can be skipped
528 by clients under certain conditions.
530 PH/07 The error message for a badly-placed control=no_multiline_responses left
531 "_responses" off the end of the name.
533 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
535 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
536 (without spaces) instead of just copying the configuration text.
538 PH/10 Added the /noupdate option to the ratelimit ACL condition.
540 PH/11 Added $max_received_linelength.
542 PH/12 Added +ignore_defer and +include_defer to host lists.
544 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
545 way in which PCRE > 7.0 is built.
547 PH/14 Implemented queue_only_load_latch.
549 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
550 MAIL command. The effect was to mangle the value on 64-bit systems.
552 PH/16 Another patch from the Sieve maintainer.
554 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
556 PH/18 If a system quota error occurred while trying to create the file for
557 a maildir delivery, the message "Mailbox is full" was not appended to the
558 bounce if the delivery eventually timed out. Change 4.67/27 below applied
559 only to a quota excession during the actual writing of the file.
561 PH/19 It seems that peer DN values may contain newlines (and other non-printing
562 characters?) which causes problems in log lines. The DN values are now
563 passed through string_printing() before being added to log lines.
565 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
566 and InterBase are left for another time.)
568 PH/21 Added message_body_newlines option.
570 PH/22 Guard against possible overflow in moan_check_errorcopy().
572 PH/23 POSIX allows open() to be a macro; guard against that.
574 PH/24 If the recipient of an error message contained an @ in the local part
575 (suitably quoted, of course), incorrect values were put in $domain and
576 $local_part during the evaluation of errors_copy.
582 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
583 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
586 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
587 issue a MAIL command.
589 PH/02 In an ACL statement such as
591 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
593 if a client was not listed at all, or was listed with a value other than
594 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
595 the condition was not true (as it should be), so access was not denied.
596 The bug was that the ! inversion was incorrectly passed on to the second
597 item. This has been fixed.
599 PH/03 Added additional dnslists conditions == and =& which are different from
600 = and & when the dns lookup returns more than one IP address.
602 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
603 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
605 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
606 FSYNC, which compiles an option called disable_fsync that allows for
607 bypassing fsync(). The documentation is heavily laced with warnings.
609 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
611 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
612 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
613 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
614 including adding "make clean"; (3) Added -fPIC when compiling the test
615 dynamically loaded module, to get rid of a warning.
617 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
618 message fails, move_frozen_messages = true and ignore_bounce_errors_after
619 = 0s. The bug is otherwise harmless.
621 PH/07 There was a bug in the dovecot authenticator such that the value of
622 $auth1 could be overwritten, and so not correctly preserved, after a
623 successful authentication. This usually meant that the value preserved by
624 the server_setid option was incorrect.
626 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
628 PH/09 Installed PCRE release 7.0.
630 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
631 run for batched SMTP input. It is now run at the start of every message
632 in the batch. While fixing this I discovered that the process information
633 (output by running exiwhat) was not always getting set for -bs and -bS
634 input. This is fixed, and it now also says "batched" for BSMTP.
636 PH/11 Added control=no_pipelining.
638 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
639 patch, slightly modified), and move the expansion of helo_data till after
640 the connection is made in the smtp transport (so it can use these
643 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
645 PH/14 Added log_selector = +pid.
647 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
649 PH/16 Add ${if forany and ${if forall.
651 PH/17 Added dsn_from option to vary the From: line in DSNs.
653 PH/18 Flush SMTP output before performing a callout, unless control =
654 no_callout_flush is set.
656 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
657 was true (the default) a successful delivery failed to delete the retry
658 item, thus causing premature timeout of the address. The bug is now
661 PH/20 Added hosts_avoid_pipelining to the smtp transport.
663 PH/21 Long custom messages for fakedefer and fakereject are now split up
664 into multiline reponses in the same way that messages for "deny" and
665 other ACL rejections are.
667 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
668 with slight modification.
670 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
671 draft, changing the syntax and factoring some duplicate code".
673 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
674 for deliveries of the second and subsequent messages over the same SMTP
677 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
678 ${reduce, with only minor "tidies".
680 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
682 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
683 expansion side effects.
685 PH/27 When a message times out after an over-quota error from an Exim-imposed
686 quota, the bounce message says "mailbox is full". This message was not
687 being given when it was a system quota that was exceeded. It now should
690 MH/03 Made $recipients available in local_scan(). local_scan() already has
691 better access to the recipient list through recipients_list[], but
692 $recipients can be useful in postmaster-provided expansion strings.
694 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
695 in the case of a MAIL command with additional options following the
696 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
697 were accidentally chopped off.
699 PH/29 SMTP synchronization checks are implemented when a command is read -
700 there is a check that no more input is waiting when there shouldn't be
701 any. However, for some commands, a delay in an ACL can mean that it is
702 some time before the response is written. In this time, more input might
703 arrive, invalidly. So now there are extra checks after an ACL has run for
704 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
705 pipelining has not been advertised.
707 PH/30 MH's patch to allow iscntrl() characters to be list separators.
709 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
710 returned in the SMTP response when smtp_return_error_details was false.
713 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
714 instead of the C I/O that was originally supplied, because problems were
717 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
718 Exim which did not show up earlier: it was assuming that a call to
719 SSL_CTX_set_info_callback() might give an error value. In fact, there is
720 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
721 was a macro that became an assignment, so it seemed to work. This has
722 changed to a proper function call with a void return, hence the compile
723 error. Exim's code has been fixed.
725 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
728 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
730 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
732 PH/37 If a message is not accepted after it has had an id assigned (e.g.
733 because it turns out to be too big or there is a timeout) there is no
734 "Completed" line in the log. When some messages of this type were
735 selected by exigrep, they were listed as "not completed". Others were
736 picked up by some special patterns. I have improved the selection
737 criteria to be more general.
739 PH/38 The host_find_failed option in the manualroute router can now be set
740 to "ignore", to completely ignore a host whose IP address cannot be
741 found. If all hosts are ignored, the behaviour is controlled by the new
742 host_all_ignored option.
744 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
745 homing or because of multiple MX records with /mx) generated more than
746 one IP address, and the following item turned out to be the local host,
747 all the secondary addresses of the first item were incorrectly removed
748 from the list, along with the local host and any following hosts (which
749 is what is supposed to happen).
751 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
752 whoever called Exim into the -H file. In the case of the daemon it was
753 behaving confusingly. When first started, it used values for whoever
754 started the daemon, but after a SIGHUP it used the Exim user (because it
755 calls itself on a restart). I have changed the code so that it now always
758 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
759 message are rejected with the same error (e.g. no authentication or bad
760 sender address), and a DATA command is nevertheless sent (as can happen
761 with PIPELINING or a stupid MUA), the error message that was given to the
762 RCPT commands is included in the rejection of the DATA command. This is
763 intended to be helpful for MUAs that show only the final error to their
766 PH/42 Another patch from the Sieve maintainer.
768 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
769 Eximstats - Fixed some broken HTML links and added missing column headers
771 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
772 columns for Rejects, Temp Rejects, Ham, and Spam rows.
774 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
776 PH/43 Yet another patch from the Sieve maintainer.
778 PH/44 I found a way to check for a TCP/IP connection going away before sending
779 the response to the final '.' that terminates a message, but only in the
780 case where the client has not sent further data following the '.'
781 (unfortunately, this is allowed). However, in many cases there won't be
782 any further data because there won't be any more messages to send. A call
783 to select() can be used: if it shows that the input is "ready", there is
784 either input waiting, or the socket has been closed. An attempt to read
785 the next input character can distinguish the two cases. Previously, Exim
786 would have sent an OK response which the client would never have see.
787 This could lead to message repetition. This fix should cure that, at
788 least in a lot of common cases.
790 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
791 advertised in response to EHLO.
797 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
798 fixed by 4.65/MH/01 (is this a record?) are fixed:
800 (i) An empty string was always treated as zero by the numeric comparison
801 operators. This behaviour has been restored.
803 (ii) It is documented that the numeric comparison operators always treat
804 their arguments as decimal numbers. This was broken in that numbers
805 starting with 0 were being interpreted as octal.
807 While fixing these problems I realized that there was another issue that
808 hadn't been noticed. Values of message_size_limit (both the global option
809 and the transport option) were treated as octal if they started with 0.
810 The documentation was vague. These values are now always treated as
811 decimal, and I will make that clear in the documentation.
817 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
818 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
821 MH/01 Don't check that the operands of numeric comparison operators are
822 integers when their expansion is in "skipping" mode (fixes bug
823 introduced by 4.64-PH/07).
825 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
826 child addresses, Exim now panics and dies. Previously, because the count
827 is held in a short int, deliveries were likely to be lost. As such a
828 large number of recipients for a single message is ridiculous
829 (performance will be very, very poor), I have chosen to impose a limit
830 rather than extend the field.
836 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
837 leftover -K file (the existence of which was triggered by #402).
838 While we were at it, introduced process PID as part of the -K
839 filename. This should rule out race conditions when creating
842 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
843 processing considerably. Previous code took too long for large mails,
844 triggering a timeout which in turn triggers #401.
846 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
847 in the DK code in transports.c. sendfile() is not really portable,
848 hence the _LINUX specificness.
850 TF/01 In the add_headers option to the mail command in an Exim filter,
851 there was a bug that Exim would claim a syntax error in any
852 header after the first one which had an odd number of characters
855 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
856 callout verification, Exim cached a "reject" for the entire domain. This
857 is correct for most verifications, but it is not correct for a recipient
858 verification with use_sender or use_postmaster set, because in that case
859 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
860 case of MAIL FROM:<> rejection from other early rejections (e.g.
861 rejection of HELO). When verifying a recipient using a non-null MAIL
862 address, the cache is ignored if it shows MAIL FROM:<> rejection.
863 Whatever the result of the callout, the value of the domain cache is
864 left unchanged (for any other kind of callout, getting as far as trying
865 RCPT means that the domain itself is ok).
867 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
870 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
871 manifest itself as EPIPE rather than ECONNECT. When tidying away a
872 session, the daemon ignores ECONNECT errors and logs others; it now
873 ignores EPIPE as well.
875 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
876 (quoted-printable decoding).
878 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
879 later the small subsequent patch to fix an introduced bug.
881 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
883 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
885 PH/08 An error is now given if message_size_limit is specified negative.
887 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
888 to be given (somewhat) arbitrary names.
890 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
893 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
894 miscellaneous code fixes
896 PH/10 Added the log_reject_target ACL modifier to specify where to log
899 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
900 hostname. This is wrong, because it relates to the incoming message (and
901 probably the interface on which it is arriving) and not to the outgoing
902 callout (which could be using a different interface). This has been
903 changed to use the value of the helo_data option from the smtp transport
904 instead - this is what is used when a message is actually being sent. If
905 there is no remote transport (possible with a router that sets up host
906 addresses), $smtp_active_hostname is used.
908 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
909 tweaks were necessary in order to get it to work (see also 21 below):
910 (a) The code assumed that strncpy() returns a negative number on buffer
911 overflow, which isn't the case. Replaced with Exim's string_format()
913 (b) There were several signed/unsigned issues. I just did the minimum
914 hacking in of casts. There is scope for a larger refactoring.
915 (c) The code used strcasecmp() which is not a standard C function.
916 Replaced with Exim's strcmpic() function.
917 (d) The code set only $1; it now sets $auth1 as well.
918 (e) A simple test gave the error "authentication client didn't specify
919 service in request". It would seem that Dovecot has changed its
920 interface. Fortunately there's a specification; I followed it and
921 changed what the client sends and it appears to be working now.
923 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
926 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
927 address A is aliased to B and C, where B exists and C does not. Without
928 -v the output is "A verified" because verification stops after a
929 successful redirection if more than one address is generated. However,
930 with -v the child addresses are also verified. Exim was outputting "A
931 failed to verify" and then showing the successful verification for C,
932 with its parentage. It now outputs "B failed to verify", showing B's
933 parentage before showing the successful verification of C.
935 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
936 look up a TXT record in a specific list after matching in a combined
939 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
940 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
941 they consult the DNS. I had assumed they would set it the way they
942 wanted; and indeed my experiments on Linux seem to show that in some
943 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
944 To be on the safe side, however, I have now made the interface to
945 host_find_byname() similar to host_find_bydns(), with an argument
946 containing the DNS resolver options. The host_find_byname() function now
947 sets these options at its start, just as host_find_bydns() does. The smtp
948 transport options dns_qualify_single and dns_search_parents are passed to
949 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
950 of host_find_byname() use the default settings of RES_DEFNAMES
951 (qualify_single) but not RES_DNSRCH (search_parents).
953 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
954 spool_read_header() do less string testing, by means of a preliminary
955 switch on the second character of optional "-foo" lines. (This is
956 overdue, caused by the large number of possibilities that now exist.
957 Originally there were few.) While I was there, I also converted the
958 str(n)cmp tests so they don't re-test the leading "-" and the first
959 character, in the hope this might squeeze out yet more improvement.
961 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
962 flag allowing group syntax was set by the header_syntax check but not
963 turned off, possible causing trouble later; (2) The flag was not being
964 set at all for the header_verify test, causing "group"-style headers to
965 be rejected. I have now set it in this case, and also caused header_
966 verify to ignore an empty address taken from a group. While doing this, I
967 came across some other cases where the code for allowing group syntax
968 while scanning a header line wasn't quite right (mostly, not resetting
969 the flag correctly in the right place). These bugs could have caused
970 trouble for malformed header lines. I hope it is now all correct.
972 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
973 with the "reply" argument non-NULL. The code, however (which originally
974 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
975 but it didn't always do it. This confused somebody who was copying the
976 code for some other use. I have removed all the tests.
978 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
979 feature that was used to support insecure browsers during the U.S. crypto
980 embargo. It requires special client support, and Exim is probably the
981 only MTA that supported it -- and would never use it because real RSA is
982 always available. This code has been removed, because it had the bad
983 effect of slowing Exim down by computing (never used) parameters for the
984 RSA_EXPORT functionality.
986 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
987 authenticator to fail if there's a tab character in the incoming data
988 (there should never be unless someone is messing about, as it's supposed
989 to be base64-encoded). Also added, on Timo's advice, the "secured" option
990 if the connection is using TLS or if the remote IP is the same as the
991 local IP, and the "valid-client-cert option" if a client certificate has
994 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
995 authenticators. This can be used for authorization after authentication
996 succeeds. (In the case of plaintext, it servers for both authentication
999 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1000 if any retry times were supplied.
1002 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1003 connection, where there is no client IP address to check. In this
1004 situation, the verify now always succeeds.
1006 PH/25 Applied John Jetmore's -Mset patch.
1008 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1010 PH/27 In a string expansion for a processed (not raw) header when multiple
1011 headers of the same name were present, leading whitespace was being
1012 removed from all of them, but trailing whitespace was being removed only
1013 from the last one. Now trailing whitespace is removed from each header
1014 before concatenation. Completely empty headers in a concatenation (as
1015 before) are ignored.
1017 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1018 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1020 PH/29 [Removed. This was a change that I later backed out, and forgot to
1021 correct the ChangeLog entry (that I had efficiently created) before
1022 committing the later change.]
1024 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1025 address errors (4xx response to RCPT) over the same connection as other
1026 messages routed to the same hosts. Such deliveries are always "forced",
1027 so retry times are not inspected. This resulted in far too many retries
1028 for the affected addresses. The effect occurred only when there were more
1029 hosts than the hosts_max_try setting in the smtp transport when it had
1030 the 4xx errors. Those hosts that it had tried were not added to the list
1031 of hosts for which the message was waiting, so if all were tried, there
1032 was no problem. Two fixes have been applied:
1034 (i) If there are any address or message errors in an SMTP delivery, none
1035 of the hosts (tried or untried) are now added to the list of hosts
1036 for which the message is waiting, so the message should not be a
1037 candidate for sending over the same connection that was used for a
1038 successful delivery of some other message. This seems entirely
1039 reasonable: after all the message is NOT "waiting for some host".
1040 This is so "obvious" that I'm not sure why it wasn't done
1041 previously. Hope I haven't missed anything, but it can't do any
1042 harm, as the worst effect is to miss an optimization.
1044 (ii) If, despite (i), such a delivery is accidentally attempted, the
1045 routing retry time is respected, so at least it doesn't keep
1046 hammering the server.
1048 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1049 in ${readsocket because some servers need this prod.
1051 PH/32 Added some extra debug output when updating a wait-xxx database.
1053 PH/33 The hint "could be header name not terminated by colon", which has been
1054 given for certain expansion errors for a long time, was not being given
1055 for the ${if def:h_colon_omitted{... case.
1057 PH/34 The spec says: "With one important exception, whenever a domain list is
1058 being scanned, $domain contains the subject domain." There was at least
1059 one case where this was not true.
1061 PH/35 The error "getsockname() failed: connection reset by peer" was being
1062 written to the panic log as well as the main log, but it isn't really
1063 panic-worthy as it just means the connection died rather early on. I have
1064 removed the panic log writing for the ECONNRESET error when getsockname()
1067 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1068 runs only) independently of the message's sender address. This meant
1069 that, if the 4xx error was in fact related to the sender, a different
1070 message to the same recipient with a different sender could confuse
1071 things. In particualar, this can happen when sending to a greylisting
1072 server, but other circumstances could also provoke similar problems.
1073 I have changed the default so that the retry time for these errors is now
1074 based a combination of the sender and recipient addresses. This change
1075 can be overridden by setting address_retry_include_sender=false in the
1078 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1079 remote server are returned as part of bounce messages. This was not
1080 happening for LMTP over a pipe (the lmtp transport), but now it is the
1081 same for both kinds of LMTP.
1083 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1084 in header lines that were in fact CNAMEs. This is no longer the case.
1086 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1087 and queue runs started by the daemon processed all messages. This has
1088 been fixed so that -R and -S can now usefully be given with -q<time>.
1090 PH/40 Import PCRE release 6.7 (fixes some bugs).
1092 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1094 PH/42 Give an error if -q is specified more than once.
1096 PH/43 Renamed the variables $interface_address and $interface_port as
1097 $received_ip_address and $received_port, to make it clear that these
1098 values apply to message reception, and not to the outgoing interface when
1099 a message is delivered. (The old names remain recognized, of course.)
1101 PH/44 There was no timeout on the connect() call when using a Unix domain
1102 socket in the ${readsocket expansion. There now is.
1104 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1105 be meaningful with "accept".
1107 SC/01 Eximstats V1.43
1108 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1110 SC/02 Eximstats V1.44
1111 Use a glob alias rather than an array ref in the generated
1112 parser. This improves both readability and performance.
1114 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1115 Collect SpamAssassin and rejection statistics.
1116 Don't display local sender or destination tables unless
1117 there is data to show.
1118 Added average volumes into the top table text output.
1120 SC/04 Eximstats V1.46
1121 Collect data on the number of addresses (recipients)
1122 as well as the number of messages.
1124 SC/05 Eximstats V1.47
1125 Added 'Message too big' to the list of mail rejection
1126 reasons (thanks to Marco Gaiarin).
1128 SC/06 Eximstats V1.48
1129 Mainlog lines which have GMT offsets and are too short to
1130 have a flag are now skipped.
1132 SC/07 Eximstats V1.49 (Alain Williams)
1133 Added the -emptyok flag.
1135 SC/08 Eximstats V1.50
1136 Fixes for obtaining the IP address from reject messages.
1138 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1139 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1140 whitesspace changes from 4.64-PH/27
1142 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1145 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1146 are found, allow negative numbers in numeric criteria)
1148 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1150 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1151 to match changes made in 4.64-PH/43
1153 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1155 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1156 "domain" is actually an IP address, and force a failure. This locks out
1157 those revolvers/nameservers that support "A-for-A" lookups, in
1158 contravention of the specifications.
1160 PH/48 When a host name was looked up from an IP address, and the subsequent
1161 forward lookup of the name timed out, the host name was left in
1162 $sender_host_name, contrary to the specification.
1164 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1165 restricted to single-key lookups, Exim was not diagnosing an error if
1166 * or *@ was used with a query-style lookup.
1168 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1170 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1171 long ago, but noone interested enough thought of it. Let's just say that
1172 the "1.1" means that there are some new functions that weren't there at
1173 some point in the past.
1175 PH/51 Error processing for expansion failure of helo_data from an smtp
1176 transport during callout processing was broken.
1178 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1179 tested/used via the -bh/-bhc/-bs options.
1181 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1182 bug, fixed in subsequent PCRE releases).
1184 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1185 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1187 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1193 SC/01 Use a glob alias rather than an array ref in eximstats generated
1194 parser. This improves both readability and performance.
1196 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1197 Don't display local sender or destination tables in eximstats unless
1198 there is data to show.
1199 Added average volumes into the eximstats top table text output.
1201 SC/03 Collect data on the number of addresses (recipients) as well
1202 as the number of messages in eximstats.
1204 TF/01 Correct an error in the documentation for the redirect router. Exim
1205 does (usually) call initgroups() when daemonizing.
1207 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1208 with consistent privilege compared to when running as a daemon.
1210 TF/03 Note in the spec that $authenticated_id is not set for local
1211 submissions from trusted users.
1213 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1214 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1216 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1217 by adding some example configuration directives to the default
1218 configuration file. A little bit of work is required to uncomment the
1219 directives and define how usernames and passwords are checked, but
1220 there is now a framework to start from.
1222 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1223 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1224 without this. I don't know how relevant this is to other LDAP libraries.
1226 PH/02 Add the verb name to the "unknown ACL verb" error.
1228 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1230 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1232 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1233 directory not expanded when it should be if an expanded home directory
1234 was set for the address (which is overridden by the transport).
1236 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1239 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1240 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1241 because it is too late at that time, and has no effect.
1243 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1244 security issue with \' (bugzilla #107). I could not use the
1245 PQescapeStringConn() function, because it needs a PGconn value as one of
1248 PH/08 When testing addresses using -bt, indicate those final addresses that
1249 are duplicates that would not cause an additional delivery. At least one
1250 person was confused, thinking that -bt output corresponded to deliveries.
1251 (Suppressing duplicates isn't a good idea as you lose the information
1252 about possibly different redirections that led to the duplicates.)
1254 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1255 systems where poll() doesn't work, in particular OS X.
1257 PH/10 Added more information to debugging output for retry time not reached.
1259 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1260 operations in malware.c.
1262 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1265 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1266 syslog_duplication set false, log lines that would normally be written
1267 both the the main log and to the reject log were not written to syslog at
1270 PH/14 In the default configuration, change the use of "message" in ACL warn
1271 statements to "add_header".
1273 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1274 not followed by a command (e.g. "seen endif").
1276 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1277 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1280 PH/17 Added extra conditions to the default value of delay_warning_condition
1284 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1285 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1286 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1289 The Auto-Submitted: and various List- headers are standardised, whereas I
1290 don't think Precedence: ever was.
1292 PH/18 Refactored debugging code in route_finduser() to show more information,
1293 in particular, the error code if getpwnam() issues one.
1295 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1296 This is apparently needed in addition to the PH/07 change above to avoid
1297 any possible encoding problems.
1299 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1300 but not after initializing Perl.
1302 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1303 output them only if debugging. By default they are written stderr,
1304 apparently, which is not desirable.
1306 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1309 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1312 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1314 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1315 authenticated or an ident call has been made. Suppress the default
1316 values for $authenticated_id and $authenticated_sender (but permit -oMai
1317 and -oMas) when testing with -bh.
1319 PH/24 Re-jigged the order of the tests in the default configuration so that the
1320 tests for valid domains and recipients precede the DNS black list and CSA
1321 tests, on the grounds that those ones are more expensive.
1323 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1324 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1325 This bug exists in every version of Exim that I still have, right back to
1328 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1329 However, an attempt to turn on case-sensitivity in a regex key by
1330 including (?-i) didn't work because the subject string was already
1331 lowercased, and the effects were non-intuitive. It turns out that a
1332 one-line patch can be used to allow (?-i) to work as expected.
1338 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1339 other effects) broke the use of negated acl sub-conditions.
1341 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1344 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1345 "Deny" causes Exim to reject the incoming connection with a 554 error.
1346 Unfortunately, if there is a major crisis, such as a disk failure,
1347 tcp-wrappers gives "deny", whereas what one would like would be some
1348 kind of temporary error. A kludge has been added to help with this.
1349 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1350 554 error is used if errno is still zero or contains ENOENT (which occurs
1351 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1354 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1356 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1357 errors. Otherwise a message that provokes a temporary error (when other
1358 messages do not) can cause a whole host to time out.
1360 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1361 addresses were routed directly to files or pipes from a redirect router.
1362 File deliveries just didn't batch; pipe deliveries might have suffered
1365 PH/06 A failure to get a lock for a hints database would erroneously always say
1366 "Failed to get write lock", even when it was really a read lock.
1368 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1369 of 0600. This has been changed to use the value of the lockfile_mode
1370 option (which defaults to 0600).
1372 PH/08 Applied small patch from the Sieve maintainer.
1374 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1375 folder from quota calculations, a direct delivery into this folder messed
1376 up the contents of the maildirsize file. This was because the regex was
1377 used only to exclude .Trash (or whatever) when the size of the mailbox
1378 was calculated. There was no check that a delivery was happening into an
1379 excluded directory. This bug has been fixed by ignoring all quota
1380 processing for deliveries into excluded directories.
1382 PH/10 Added the maildirfolder_create_regex option to appendfile.
1388 PH/01 The code for finding all the local interface addresses on a FreeBSD
1389 system running IPv6 was broken. This may well have applied to all BSD
1390 systems, as well as to others that have similar system calls. The broken
1391 code found IPv4 interfaces correctly, but gave incorrect values for the
1392 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1393 that it would not match correctly against @[] and not recognize the IPv6
1396 PH/02 The ipliteral router was not recognizing addresses of the form user@
1397 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1399 PH/03 Added disable_ipv6.
1401 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1402 decoded form, because it is most often used to construct To: headers
1403 lines in autoreplies, and the decoded form may well be syntactically
1404 invalid. However, $reply_address has leading white space removed, and all
1405 newlines turned into spaces so that the autoreply transport does not
1408 PH/05 If group was specified without a user on a router, and no group or user
1409 was specified on a transport, the group from the router was ignored.
1411 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1412 for visible compile-time settings that can be used to change these
1413 numbers, for those that want even more. Backwards compatibility with old
1414 spool files has been maintained. However, going back to a previous Exim
1415 release will lost any variables that are in spool files.
1417 PH/07 Two small changes when running in the test harness: increase delay when
1418 passing a TCP/IP connection to a new process, in case the original
1419 process has to generate a bounce, and remove special handling of
1420 127.0.0.2 (sic), which is no longer necessary.
1422 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1423 be the same on different OS.
1425 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1428 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1429 whether --show-vars was specified or not
1431 JJ/02 exipick: Added support for new ACL variable spool format introduced
1434 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1435 syntactically invalid From: or Reply-to: line, and a filter used this to
1436 generate an autoreply, and therefore failed to obtain an address for the
1437 autoreply, Exim could try to deliver to a non-existent relative file
1438 name, causing unrelated and misleading errors. What now happens is that
1439 it logs this as a hard delivery error, but does not attempt to create a
1442 PH/11 The exinext utility has a -C option for testing purposes, but although
1443 the given file was scanned by exinext itself; it wasn't being passed on
1444 when Exim was called.
1446 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1447 an end-of-file indication when reading a command response.
1449 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1450 compiled. In many other places in Exim, IPv6 addresses are always
1451 recognized, so I have changed this. It also means that IPv4 domain
1452 literals of the form [IPV4:n.n.n.n] are now always recognized.
1454 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1455 used if the router is not running as root, for example, when verifying at
1456 ACL time, or when using -bh. The debugging output from this situation was
1457 non-existent - all you got was a failure to exec. I have made two
1460 (a) Failures to set uid/gid, the current directory, or a process leader
1461 in a subprocess such as that created by queryprogram now generate
1462 suitable debugging ouput when -d is set.
1464 (b) The queryprogram router detects when it is not running as root,
1465 outputs suitable debugging information if -d is set, and then runs
1466 the subprocess without attempting to change uid/gid.
1468 PH/15 Minor change to Makefile for building test_host (undocumented testing
1471 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1472 additional section of a DNS packet that returns MX or SRV records.
1473 Instead, it always explicitly searches for A/AAAA records. This avoids
1474 major problems that occur when a DNS server includes only records of one
1475 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1476 fixed another bug: if SRV records were looked up and the corresponding
1477 address records were *not* found in the additional section, the port
1478 values from the SRV records were lost.
1480 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1481 using the correct key (the original address) when searching the retry
1482 rules in order to find which one to use for generating the retry hint.
1484 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1485 adding the default one. Similarly, if it contains a Reply-To: header, the
1486 errors_reply_to option, if set, is not used.
1488 PH/19 When calculating a retry time, Exim used to measure the "time since
1489 failure" by looking at the "first failed" field in the retry record. Now
1490 it does not use this if it is later than than the arrival time of the
1491 message. Instead it uses the arrival time. This makes for better
1492 behaviour in cases where some deliveries succeed, thus re-setting the
1493 "first failed" field. An example is a quota failure for a huge message
1494 when small messages continue to be delivered. Without this change, the
1495 "time since failure" will always be short, possible causing more frequent
1496 delivery attempts for the huge message than are intended.
1497 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1499 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1500 $1, $2, $3) because the numerical variables can be reset during some
1501 expansion items (e.g. "match"), thereby losing the authentication data.
1503 PH/21 Make -bV show the size of off_t variables so that the test suite can
1504 decide whether to run tests for quotas > 2G.
1506 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1507 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1508 filecount value is greater than 2G or if a quota value is greater than 2G
1509 on a system where the size of off_t is not greater than 4, a panic error
1512 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1513 never match. The debug and -bh output now contains an explicit error
1514 message indicating a malformed IPv4 address or mask.
1516 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1517 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1518 PH/23 above applies.
1520 PH/25 Do not write to syslog when running in the test harness. The only
1521 occasion when this arises is a failure to open the main or panic logs
1522 (for which there is an explicit test).
1524 PH/26 Added the /no_tell option to "control=freeze".
1526 PH/27 If a host name lookup failed very early in a connection, for example, if
1527 the IP address matched host_lookup and the reverse lookup yielded a name
1528 that did not have a forward lookup, an error message of the form "no IP
1529 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1530 could be logged. Now it outputs the IP address instead of "NULL".
1532 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1533 allows the sender and the authenticated sender to be set when
1534 submitting a message from within Exim. Since child_open_exim() is
1535 documented for local_scan(), the new function should be too.
1537 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1538 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1539 results in an empty string is now treated as unset.
1541 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1543 PH/31 Added sender_verify_fail logging option.
1545 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1546 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1547 tidied the source and removed it altogether.
1549 PH/33 When a queue run was abandoned because the load average was too high, a
1550 log line was always written; now it is written only if the queue_run log
1551 selector is set. In addition, the log line for abandonment now contains
1552 information about the queue run such as the pid. This is always present
1553 in "start" and "stop" lines but was omitted from the "abandon" line.
1555 PH/34 Omit spaces between a header name and the colon in the error message that
1556 is given when verify = headers_syntax fails (if there are lots of them,
1557 the message gets confusing).
1559 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1560 names, as there are now some PTR records that contain slashes. This check
1561 is only to protect against broken name servers that fall over on strange
1562 characters, so the fact that it applies to all lookups doesn't matter.
1564 PH/36 Now that the new test suite is complete, we can remove some of the
1565 special code in Exim that was needed for the old test suite. For example,
1566 sorting DNS records because real resolvers return them in an arbitrary
1567 order. The new test suite's fake resolver always returns records in the
1570 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1571 bounces) except when queue_only is set, to avoid logging races between
1572 the different processes.
1574 PH/38 Panic-die if .include specifies a non-absolute path.
1576 PH/39 A tweak to the "H" retry rule from its user.
1578 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1579 a label. They prevented compilation on older perls.
1581 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1582 a warning to be raised on newish perls.
1584 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1585 on queue. Changes to match documented behaviour of showing count of
1586 messages matching specified criteria.
1588 PH/40 Changed the default ident timeout from 30s to 5s.
1590 PH/41 Added support for the use of login_cap features, on those BSD systems
1591 that have them, for controlling the resources used by pipe deliveries.
1593 PH/42 The content-scanning code uses fopen() to create files in which to put
1594 message data. Previously it was not paying any attention to the mode of
1595 the files. Exim runs with umask(0) because the rest of the code creates
1596 files with open(), and sets the required mode explicitly. Thus, these
1597 files were ending up world-writeable. This was not a big issue, because,
1598 being within the spool directory, they were not world-accessible. I have
1599 created a function called modefopen, which takes an additional mode
1600 argument. It sets umask(777), creates the file, chmods it to the required
1601 mode, then resets the umask. All the relevant calls to fopen() in the
1602 content scanning code have been changed to use this function.
1604 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1605 to 24 hours. This avoids potential overflow problems when processing G
1606 and H retry rules. I suspect nobody ever tinkers with this value.
1608 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1610 PH/45 When the plaintext authenticator is running as a client, the server's
1611 challenges are checked to ensure they are valid base64 strings. By
1612 default, the authentication attempt is cancelled if an invalid string is
1613 received. Setting client_ignore_invalid_base64 true ignores these errors.
1614 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1615 they are received. Thus, the responses can be made to depend on the
1616 challenges. If an invalid string is ignored, an empty string is placed in
1619 PH/46 Messages that are created by the autoreply transport now contains a
1620 References: header, in accordance with RFCs 2822 and 3834.
1622 PH/47 Added authenticated_sender_force to the smtp transport.
1624 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1626 PH/49 Installed latest patch from the Sieve maintainer.
1628 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1629 was also set, the appendfile transport was unnecessarily scanning a
1630 directory of message files (e.g. for maildir delivery) to find the count
1631 of files (along with the size), even though it did not need this
1632 information. It now does the scan only if it needs to find either the
1633 size of the count of files.
1635 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1637 PH/52 Two bugs concerned with error handling when the smtp transport is
1640 (i) Exim was not creating retry information for temporary errors given
1641 for individual recipients after the DATA command when the smtp transport
1642 was used in LMTP mode. This meant that they could be retried too
1643 frequently, and not timed out correctly.
1645 (ii) Exim was setting the flag that allows error details to be returned
1646 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1647 recipients that were returned after the DATA command.
1649 PH/53 This is related to PH/52, but is more general: for any failing address,
1650 when detailed error information was permitted to be returned to the
1651 sender, but the error was temporary, then after the final timeout, only
1652 "retry timeout exceeded" was returned. Now it returns the full error as
1653 well as "retry timeout exceeded".
1655 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1656 do this, and (what is worse) MTAs that accept it.
1658 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1659 will now be deprecated.
1661 PH/56 New os.c-cygwin from the Cygwin maintainer.
1663 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1664 formats (previously only available in exim formats via -bpr, -bpru,
1665 and -bpra. Now also available in native and exiqgrep formats)
1667 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1668 with very large, slow to parse queues
1670 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1672 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1674 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1675 responses to authentication challenges, though it was showing the
1676 challenges; (ii) I've removed the CR characters from the debug output for
1679 PH/58 Allow for the insertion of a newline as well as a space when a string
1680 is turned into more than one encoded-word during RFC 2047 encoding. The
1681 Sieve code now uses this.
1683 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1684 data_4xx, lost_connection, tls_required.
1686 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1687 message was being sent as an SMTP response.
1689 PH/61 Add -l and -k options to exicyclog.
1691 PH/62 When verifying, if an address was redirected to one new address, so that
1692 verification continued, and the new address failed or deferred after
1693 having set something in $address_data, the value of $address_data was not
1694 passed back to the ACL. This was different to the case when no
1695 redirection occurred. The value is now passed back in both cases.
1697 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1698 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1699 use login_cap.h, so on its own it isn't the distinguishing feature. The
1700 new name refers directly to the setclassresources() function.
1702 PH/65 Added configuration files for NetBSD3.
1704 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1706 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1707 is preferred over IPv4.
1709 PH/68 The bounce_return_message and bounce_return_body options were not being
1710 honoured for bounces generated during the reception of non-SMTP messages.
1711 In particular, this applied to messages rejected by the ACL. This bug has
1712 been fixed. However, if bounce_return_message is true and bounce_return_
1713 body is false, the headers that are returned for a non-SMTP message
1714 include only those that have been read before the error was detected.
1715 (In the case of an ACL rejection, they have all been read.)
1717 PH/69 The HTML version of the specification is now built in a directory called
1718 spec_html instead of spec.html, because the latter looks like a path with
1719 a MIME-type, and this confuses some software.
1721 PH/70 Catch two compiler warnings in sieve.c.
1723 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1724 function verify_get_ident() calls ip_connect() to connect a socket, but
1725 if the "connect()" function timed out, ip_connect() used to close the
1726 socket. However, verify_get_ident() also closes the socket later, and in
1727 between Exim writes to the log, which may get opened at this point. When
1728 the socket was closed in ip_connect(), the log could get the same file
1729 descriptor number as the socket. This naturally causes chaos. The fix is
1730 not to close the socket in ip_connect(); the socket should be closed by
1731 the function that creates it. There was only one place in the code where
1732 this was missing, in the iplookup router, which I don't think anybody now
1733 uses, but I've fixed it anyway.
1735 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1736 well as to direct DNS lookups. Otherwise the handling of names in host
1737 lists is inconsistent and therefore confusing.
1743 PH/01 Two changes to the default runtime configuration:
1745 (1) Move the checks for relay_from_hosts and authenticated clients from
1746 after to before the (commented out) DNS black list checks.
1748 (2) Add control=submission to the relay_from_hosts and authenticated
1749 clients checks, on the grounds that messages accepted by these
1750 statements are most likely to be submissions.
1752 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1754 (1) Generate an error if the third argument for the ${prvs expansion is
1757 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1760 (3) Reset the variables that are obtained from the first argument of
1761 ${prvscheck and used in the second argument before leaving the code,
1762 because their memory is reclaimed, so using them afterwards may do
1765 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1766 one (it's much easier than Tom thought :-).
1768 (5) Because of (4), we can now allow for the use of $prvscheck_result
1769 inside the third argument.
1771 PH/03 For some reason, the default setting of PATH when running a command from
1772 a pipe transport was just "/usr/bin". I have changed it to
1775 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1776 anything to be listed in the output from -bV.
1778 PH/05 When a filter generated an autoreply, the entire To: header line was
1779 quoted in the delivery log line, like this:
1781 => >A.N.Other <ano@some.domain> <original@ddress> ...
1783 This has been changed so that it extracts the operative address. There
1784 may be more than one such address. If so, they are comma-separated, like
1787 => >ano@some.domain,ona@other.domain <original@ddress> ...
1789 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1790 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1791 not being looked up in the rDNS to get a host name, Exim was showing the
1792 IP address twice in Received: lines, even though the IP addresses were
1793 identical. For example:
1795 Received: from [1.2.3.4] (helo=[1.2.3.4])
1797 However, if the real host name was known, it was omitting the HELO data
1798 if it matched the actual IP address. This has been tidied up so that it
1799 doesn't show the same IP address twice.
1801 PH/07 When both +timestamp and +memory debugging was on, the value given by
1802 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1803 called by the debug printing, thereby overwriting the timestamp buffer.
1804 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1806 PH/08 When the original message was included in an autoreply transport, it
1807 always said "this is a copy of the message, including all the headers",
1808 even if body_only or headers_only was set. It now gives an appropriate
1811 PH/09 Applied a patch from the Sieve maintainer which:
1813 o fixes some comments
1814 o adds the (disabled) notify extension core
1815 o adds some debug output for the result of if/elsif tests
1816 o points to the current vacation draft in the documentation
1817 and documents the missing references header update
1821 o fixes a bug in processing the envelope test (when testing
1822 multiple envelope elements, the last element determinted the
1825 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1826 Electronic Mail") by including:
1828 Auto-submitted: auto-generated
1830 in the messages that it generates (bounce messages and others, such as
1831 warnings). In the case of bounce messages for non-SMTP mesages, there was
1832 also a typo: it was using "Auto_submitted" (underscore instead of
1833 hyphen). Since every message generated by Exim is necessarily in response
1834 to another message, thes have all been changed to:
1836 Auto-Submitted: auto-replied
1838 in accordance with these statements in the RFC:
1840 The auto-replied keyword:
1842 - SHOULD be used on messages sent in direct response to another
1843 message by an automatic process,
1845 - MUST NOT be used on manually-generated messages,
1847 - MAY be used on Delivery Status Notifications (DSNs) and Message
1848 Disposition Notifications (MDNs),
1850 - MUST NOT be used on messages generated by automatic or periodic
1851 processes, except for messages which are automatic responses to
1854 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1855 to the default Received: header definition.
1857 PH/12 Added log selector acl_warn_skipped (default on).
1859 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1860 variables because (a) they are in the wrong storage pool and (b) even if
1861 they were copied, it wouldn't work properly because of the caching.
1863 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1864 checking when decoding. Apparently there are clients that generate
1865 overlong encoded strings. Why am I not surprised?
1867 PH/15 If the first argument of "${if match_address" was not empty, but did not
1868 contain an "@" character, Exim crashed. Now it writes a panic log message
1869 and treats the condition as false.
1871 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1873 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1874 extension "envelope-auth". The code is finished and in agreement with
1875 other implementations, but there is no documentation so far and in fact,
1876 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1877 not changing the active code.
1879 Print executed "if" and "elsif" statements when debugging is used. This
1880 helps a great deal to understand what a filter does.
1882 Document more things not specified clearly in RFC3028. I had all this
1883 sorted out, when out of a sudden new issues came to my mind. Oops."
1885 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1888 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1889 canonical form (as documented). However, after a host name lookup from
1890 the IP address, check_host() was doing a simple string comparison with
1891 addresses acquired from the DNS when checking that the found name did
1892 have the original IP as one of its addresses. Since any found IPv6
1893 addresses are likely to be in abbreviated form, the comparison could
1894 fail. Luckily, there already exists a function for doing the comparison
1895 by converting both addresses to binary, so now that is used instead of
1896 the text comparison.
1898 PH/20 There was another similar case to PH/19, when a complete host name was
1899 given in a host list; looking up its IP address could give an abbreviated
1900 form, whereas the current host's name might or might not be abbreviated.
1901 The same fix has been applied.
1907 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1908 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1911 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1912 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1914 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1916 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1917 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1918 Also, for "Auto-Submitted", treat anything other than "no" as
1919 non-personal, in accordance with RFC 3834. (Previously it treated
1920 anything starting "auto-" as non-personal.)
1922 TF/01 The control=submission/name=... option had a problem with syntax
1923 errors if the name included a slash character. The /name= option
1924 now slurps the rest of the string, so it can include any characters
1925 but it must come last in the list of options (after /sender_retain
1928 PH/05 Some modifications to the interface to the fake nameserver for the new
1936 TK/01 Added the "success_on_redirect" address verification option. See
1937 NewStuff for rationale and an example.
1939 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1941 PH/02 Patch to exigrep to allow it to work on syslog lines.
1943 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1944 fread() to skip over the body file's header line, because in Cygwin the
1945 header line is locked and is inaccessible.
1947 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1948 co-exist for some time) to make it clear that it is the Exim ID that is
1949 referenced, not the Message-ID: header line.
1951 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1952 string_format() function, because snprintf() does not exist on all
1955 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1956 ${stat: expansion item.
1958 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1959 protocol synchronization error", to keep the pedants happy.
1961 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1962 well as for IRIX systems, when gcc is being used. See the host.c source
1965 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1967 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1970 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1971 transport and to the smtp transport in LMTP mode.
1973 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1975 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1976 run a filter in a subprocess. This could lead to confusion in subsequent
1977 lookups in the parent process. There should also be a search_tidyup() at
1978 the end of the subprocess.
1980 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1981 only if the host matched helo_try_verify_hosts, which caused the
1982 verification to occur when the EHLO/HELO command was issued. The ACL just
1983 tested the remembered result. Now, if a previous verification attempt has
1984 not happened, "verify = helo" does it there and then.
1986 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1988 TK/03 Fix log output including CR from clamd.
1990 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1991 exist provoked a memory error which could cause a segfault.
1993 PH/15 Installed PCRE 6.2
1995 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1997 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1998 of the problem. Specifically, suggested +O2 rather than +O1 for the
2001 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2003 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2004 record, the port was not correctly logged when the outgoing_port log
2005 selector was set (it logged the transort's default port).
2007 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2008 fallback_hosts, and "hosts" in the smtp transport.
2010 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2011 host errors such as "Connection refused".
2013 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2014 authentication with radiusclient 0.4.9:
2016 - Error returned from rc_read_config was caught wrongly
2017 - Username/password not passed on to radius server due to wrong length.
2019 The presumption is that some radiusclient API changes for 4.51/PH/17
2020 were not taken care of correctly. The code is still untested by me (my
2021 Linux distribution still has 0.3.2 of radiusclient), but it was
2022 contributed by a Radius user.
2024 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2025 expanding the "port" option of the smtp transport.
2027 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2028 while reading a MIME header. Thanks to Tom Hughes for a patch.
2030 PH/24 Include config.h inside local_scan.h so that configuration settings are
2033 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2034 that in an ACL for RCPT (for example), you can examine exactly what was
2037 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2038 commands, but it was not correctly comparing the address with the actual
2039 client host address. Thus, it would show the EHLO address in Received:
2040 header lines when this was not necessary.
2042 PH/27 Added the % operator to ${eval:}.
2044 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2045 it should be ignoring failures (because with -C, for example, it has lost
2046 privilege). It wasn't ignoring creation failures other than "already
2049 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2052 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2053 because an input line was too long, either on its own, or by virtue of
2054 too many continuations, the temporary file was not being removed, and the
2055 return code was incorrect.
2057 PH/31 Missing "BOOL" in function definition in filtertest.c.
2059 PH/32 Applied Sieve patches from the maintainer.
2061 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2063 PH/33 Added "verify = not_blind".
2065 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2066 Local/Makefile (with some defaults set). These are used in built scripts
2067 such as exicyclog, but they have never been used in the exim_install
2068 script (though there are many overriding facilities there). I have
2069 arranged that the exim_install script now takes note of these two
2072 PH/35 Installed configuration files for Dragonfly.
2074 PH/36 When a locally submitted message by a trusted user did not contain a
2075 From: header, and the sender address was obtained from -f or from an SMTP
2076 MAIL command, and the trusted user did not use -F to supply a sender
2077 name, $originator_name was incorrectly used when constructing a From:
2078 header. Furthermore, $originator_name was used for submission mode
2079 messages from external hosts without From: headers in a similar way,
2080 which is clearly wrong.
2082 PH/37 Added control=suppress_local_fixups.
2084 PH/38 When log_selector = +received_sender was set, and the addition of the
2085 sender made the log line's construction buffer exactly full, or one byte
2086 less than full, an overflow happened when the terminating "\n" was
2089 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2090 when the result of a list match is failure because a DNS lookup failed.
2092 PH/40 RM_COMMAND is now used in the building process.
2094 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2095 the "build-* directories that it finds.
2097 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2098 address in a domain literal was a prefix of an interface address.
2100 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2101 when verifying a sender address, unless rewrite_headers is false.
2103 PH/44 Wrote a long comment about why errors_to addresses are verified as
2104 recipients, not senders.
2106 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2107 the ratelimit ACL was added.
2109 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2111 PH/46 Added extra information about PostgreSQL errors to the error string.
2113 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2114 avoiding the need to install special zones in a real server. This is
2115 backwards compatible; if it can't find the fake resolver, it drops back.
2116 Thus, both old and new test suites can be run.
2118 TF/02 Added util/ratelimit.pl
2120 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2121 clock is set back in time.
2123 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2124 Candler <B.Candler@pobox.com>.
2126 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2127 OK for addresses that are the result of redirections.
2129 PH/48 A number of further additions for the benefit of the new test suite,
2130 including a fake gethostbyname() that interfaces to the fake DNS resolver
2133 TF/06 The fix for widen_domains has also been applied to qualify_single and
2134 search_parents which are the other dnslookup options that can cause
2137 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2140 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2142 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2143 service" if it encountered an MX record with an empty target hostname.
2144 The message is now "an MX or SRV record indicated no SMTP service".
2146 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2147 if the DNS of the sending site is misconfigured. This is quite a
2148 common situation. This change restores the behaviour of treating a
2149 helo verification defer as a failure.
2151 PH/51 If self=fail was set on a router, the bounce message did not include the
2152 actual error message.
2158 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2160 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2161 command itself ended in error, the underlying message about the transport
2162 filter timeout was being overwritten with the pipe command error. Now the
2163 underlying error message should be appended to the second error message.
2165 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2167 PH/02 Reduce the amount of output that "make" produces by default. Full output
2168 can still be requested.
2170 PH/03 The warning log line about a condition test deferring for a "warn" verb
2171 was being output only once per connection, rather than after each
2172 occurrence (because it was using the same function as for successful
2173 "warn" verbs). This seems wrong, so I have changed it.
2175 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2176 it should not have, which might have caused a crash in the right
2177 circumstances, but probably never did.
2179 PH/04 Installed a modified version of Tony Finch's patch to make submission
2180 mode fix the return path as well as the Sender: header line, and to
2181 add a /name= option so that you can make the user's friendly name appear
2184 TF/03 Added the control = fakedefer ACL modifier.
2186 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2187 Mark Lowes for thorough testing.
2189 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2191 TK/03 Merged latest SRS patch from Miles Wilton.
2193 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2194 with the definition in sysexits.h (which is #included earlier).
2195 Fortunately, Exim does not actually use EX_OK. The code used to try to
2196 preserve the sysexits.h value, by assumimg that macro definitions were
2197 scanned for macro replacements. I have been disabused of this notion,
2198 so now the code just undefines EX_OK before #including unistd.h.
2200 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2201 in the smtp transport. When a block could not be written in a single
2202 write() function, the timeout was being re-applied to each part-write.
2203 This seems wrong - if the receiver was accepting one byte at a time it
2204 would take for ever. The timeout is now adjusted when this happens. It
2205 doesn't have to be particularly precise.
2207 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2208 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2210 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2211 without a domain if the check to <postmaster@domain> fails.
2213 SC/01 Eximstats: added -xls and the ability to specify output files
2214 (patch written by Frank Heydlauf).
2216 SC/02 Eximstats: use FileHandles for outputing results.
2218 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2220 SC/04 Eximstats: fixed display of large numbers with -nvr option
2222 SC/05 Eximstats: fixed merging of reports with empty tables.
2224 SC/06 Eximstats: added the -include_original_destination flag
2226 SC/07 Eximstats: removed tabs and trailing whitespace.
2228 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2230 TK/06 MBOX spool code: Add real "From " MBOX separator line
2231 so the .eml file is really in mbox format (even though
2232 most programs do not really care). Patch from Alex Miller.
2234 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2235 The latter is generated from $received_to and is only set if the
2236 message has one envelope recipient. SA can use these headers,
2237 obviously out-of-the-box. Patch from Alex Miller.
2239 PH/08 The ${def test on a variable was returning false if the variable's
2240 value was "0", contrary to what the specification has always said!
2241 The result should be true unless the variable is empty.
2243 PH/09 The syntax error of a character other than { following "${if
2244 def:variable_name" (after optional whitespace) was not being diagnosed.
2245 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2246 accidental colon was present, for example, could give incorrect results.
2248 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2249 result is used (not including appendfile, where other changes are about
2252 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2253 This involved changing a lot of size variables from int to off_t. It
2254 should work with maildirs and everything.
2256 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2257 spamd dying while we are connected to it.
2259 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2262 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2263 function for BDB 4.3.
2265 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2267 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2268 the sources for PCRE have been re-arranged and more files are now
2271 PH/15 The code I had for printing potentially long long variables in PH/11
2272 above was not the best (it lost precision). The length of off_t variables
2273 is now inspected at build time, and an appropriate printing format (%ld
2274 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2275 to be "long long int" or "long int". This is needed for the internal
2276 formatting function string_vformat().
2278 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2279 the configuration file to be ":syslog", then the script "guesses" where
2280 the logs files are, rather than using the compiled in default. In our
2281 case the guess is not the same as the compiled default, so the script
2282 suddenly stopped working when I started to use syslog. The patch checks
2283 to see if log_file_path is "". If so, it attempts to read it from exim
2284 with no configuration file to get the compiled in version, before it
2285 falls back to the previous guessing code."
2287 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2288 implementing BATV in an Exim configuration. See NewStuff for the gory
2291 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2292 Makefile that are specific to HP-UX.
2294 PH/18 If the "use_postmaster" option was set for a recipient callout together
2295 with the "random" option, the postmaster address was used as the MAIL
2296 FROM address for the random test, but not for the subsequent recipient
2297 test. It is now used for both.
2299 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2300 patch removes a few documentation additions to RFC 3028, because the
2301 latest draft now contains them. It adds the new en;ascii-case comparator
2302 and a new error check for 8bit text in MIME parts. Comparator and
2303 require names are now matched exactly. I enabled the subaddress
2304 extension, but it is not well tested yet (read: it works for me)."
2306 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2307 rework some of the code of TK/09 above to avoid the hardwired use of
2308 "%lld" and "long long". Replaced the call to snprintf() with a call to
2311 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2312 records point to non-existent hosts", "retry timeout exceeded", and
2313 "retry time not reached for any host after a long failure period".
2315 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2316 experimental DomainKeys support:
2318 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2319 (2) On an error such as an illegally used "control", the wrong name for
2320 the control was given.
2322 These problems did NOT occur unless DomainKeys support was compiled.
2324 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2326 PH/24 Added ${if match_ip condition.
2328 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2329 databases so that it will be absolutely obvious if a crash occurs in the
2330 DB library. This is a regular occurrence (often caused by mis-matched
2333 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2334 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2335 picked up on a user's system that detects such things. There doesn't seem
2336 to be a gcc warning option for this - only an attribute that has to be
2337 put on the function's prototype. It seems that in Fedora Core 4 they have
2338 set this on a number of new functions. No doubt there will be more in due
2341 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2342 specify a transport. However, if an address that was verified by such a
2343 router was the subject of a callout, Exim crashed because it tried to
2344 read the rcpt_include_affixes from the non-existent transport. Now it
2345 just assumes that the setting of that option is false. This bug was
2346 introduced by 4.51/PH/31.
2348 PH/28 Changed -d+all to exclude +memory, because that information is very
2349 rarely of interest, but it makes the output a lot bigger. People tend to
2350 do -d+all out of habit.
2352 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2353 code in os-type was giving problems when libc.so lives in lib64, like on
2356 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2357 aren't the modern standard, and it seems that some systems' include files
2358 don't always have them. Exim was already checking for some of the newer
2359 ones like T_AAAA, and defining it itself. I've added checks for all the
2360 record types that Exim uses.
2362 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2363 not automatically generating a new one, as it is supposed to. This
2364 prevented TLS from working. If the file did exist, but contained invalid
2365 data, a new version was generated, as expected. It was only the case of a
2366 non-existent file that was broken.
2368 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2369 with a change in libdomainkeys > 0.64.
2371 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2372 from DNS. If the selector record carries the flag, it now has
2373 precedence over the domain-wide flag.
2375 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2377 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2378 the use of a port name that isn't defined in /etc/services) occurred, the
2379 message was deferred as in a normal delivery, and thus remained on the
2380 spool, instead of being failed because of the mua_wrapper setting. This
2381 is now fixed, and I tidied up some of the mua_wrapper messages at the
2384 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2385 the messages in a hash of arrays rather than using individual hashes.
2386 This is a bit cleaner and results in dramatic memory savings, albeit
2387 at a slight CPU cost.
2389 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2390 as requested by Marc Sherman.
2392 SC/10 Eximstats: added histograms for user specified patterns as requested
2395 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2397 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2398 fopen() in the content-scanning modules that did not already have it.
2404 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2405 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2407 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2409 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2411 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2412 to test Sieve filters that use "vacation".
2414 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2415 that changes the way the GnuTLS parameters are stored in the cache file.
2416 The new format can be generated externally. For backward compatibility,
2417 if the data in the cache doesn't make sense, Exim assumes it has read an
2418 old-format file, and it generates new data and writes a new file. This
2419 means that you can't go back to an older release without removing the
2422 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2423 work if there are any delivery delays because "one_time" forces the
2424 parent to be marked "delivered", so its unseen clone is never tried
2425 again. For this reason, Exim now forbids the simultaneous setting of
2428 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2429 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2430 entry if you want to know the details. The fix, however, neglected to
2431 consider the case when local delivery batching is involved. The test for
2432 "previously delivered" was not happening when checking to see if an
2433 address could be batched with a previous (undelivered) one; under
2434 certain circumstances this could lead to multiple deliveries to the same
2437 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2438 in its include files, and this causes problems building Exim.
2440 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2441 header_syntax) but Exim was just ignoring anything given after a slash.
2442 In particular, this caused confusion with an attempt to use "verify =
2443 reverse_host_lookup/defer_ok". An error is now given when options are
2444 supplied for verify items that do not have them. (Maybe reverse_host_
2445 lookup should have a defer_ok option, but that's a different point.)
2447 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2448 defined by RFC 821) to 2048, because there were problems with some AUTH
2449 commands, and RFC 1869 says the size should be increased for extended
2450 SMTP commands that take arguments.
2452 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2455 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2456 "unknown" error; now it says that the functionality isn't in the binary.
2458 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2459 an address' error message when a string expansion fails (syntax or
2460 whatever). Otherwise the password may appear in the log. Following change
2461 PH/42 below, there is no longer a chance of it appearing in a bounce
2464 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2466 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2467 of its addresses was ever tried. (Bugzilla bug #2.)
2469 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2470 the result incorrectly in the debug output. (It correctly added a newline
2471 to what was transported.)
2473 TF/01 Added $received_time.
2475 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2476 commented out examples of how to interface to a virus scanner and to
2477 SpamAssassin. Also added commented examples of av_scanner and
2478 spamd_address settings.
2480 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2481 and controls are allowed in which ACLs. There were a couple of minor
2482 errors. Some of the entries in the conditions table (which is a table of
2483 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2484 negation of where the condition IS allowed.
2486 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2488 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2489 header file does not have a version number, so I've had to invent a new
2490 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2491 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2492 radiusclient), but it was contributed by a Radius user.
2494 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2495 files or directories, for OpenSSL.
2497 PH/19 When an Exim process that is running as root has to create an Exim log
2498 file, it does so in a subprocess that runs as exim:exim so as to get the
2499 ownership right at creation (otherwise, other Exim processes might see
2500 the file with the wrong ownership). There was no test for failure of this
2501 fork() call, which would lead to the process getting stuck as it waited
2502 for a non-existent subprocess. Forks do occasionally fail when resources
2503 run out. I reviewed all the other calls to fork(); they all seem to check
2506 PH/20 When checking for unexpected SMTP input at connect time (before writing
2507 the banner), Exim was not dealing correctly with a non-positive return
2508 from the read() function. If the client had disconnected by this time,
2509 the result was a log entry for a synchronization error with an empty
2510 string after "input=" when read() returned zero. If read() returned -1
2511 (an event I could not check), uninitialized data bytes were printed.
2512 There were reports of junk text (parts of files, etc) appearing after
2515 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2517 PH/22 Added support for macro redefinition, and (re)definition in between
2518 driver and ACL definitions.
2520 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2521 forgetting to use the resulting value; it was using the unexpanded value.
2523 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2524 hadn't been configured. The fix is from Juergen Kreileder, who
2525 understands it better than I do:
2527 "Here's what I see happening with three configured cyrus_sasl
2528 authenticators configured (plain, login, cram-md5):
2530 On startup auth_cyrus_sasl_init() gets called for each of these.
2531 This means three calls to sasl_listmech() without a specified mech_list.
2532 => SASL tests which mechs of all available mechs actually work
2533 => three warnings about OTP not working
2534 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2536 With the patch, sasl_listmech() also gets called three times. But now
2537 SASL's mech_list option is set to the server_mech specified in the the
2538 authenticator. Or in other words, the answer from sasl_listmech()
2539 gets limited to just the mech you're testing for (which is different
2541 => the return list contains just 'plain' or 'login', 'cram-md5' or
2542 nothing depending on the value of ob->server_mech.
2544 I've just tested the patch: Authentication still works fine,
2545 unavailable mechs specified in the exim configuration are still
2546 caught, and the auth.log warnings about OTP are gone."
2548 PH/25 When debugging is enabled, the contents of the command line are added
2549 to the debugging output, even when log_selector=+arguments is not
2552 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2553 answer is "GNU", and only if the return is "GNU/something" is the answer
2556 PH/27 $acl_verify_message is now set immediately after the failure of a
2557 verification in an ACL, and so is available in subsequent modifiers. In
2558 particular, the message can be preserved by coding like this:
2560 warn !verify = sender
2561 set acl_m0 = $acl_verify_message
2563 Previously, $acl_verify_message was set only while expanding "message"
2564 and "log_message" when a very denied access.
2566 PH/28 Modified OS/os.c-Linux with
2568 -#ifndef OS_LOAD_AVERAGE
2569 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2571 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2572 nomenclature these days.)
2574 PH/29 Installed patch from the Sieve maintainer that adds the options
2575 sieve_useraddress and sieve_subaddress to the redirect router.
2577 PH/30 In these circumstances:
2578 . Two addresses routed to the same list of hosts;
2579 . First host does not offer TLS;
2580 . First host accepts first address;
2581 . First host gives temporary error to second address;
2582 . Second host offers TLS and a TLS session is established;
2583 . Second host accepts second address.
2584 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2585 and peerdn, if requested) that were in fact used only for the second
2588 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2589 attention to any local part prefix or suffix that was matched by the
2590 router that accepted the address. It now behaves in the same way as it
2591 does for delivery: the affixes are removed from the local part unless
2592 rcpt_include_affixes is set on the transport.
2594 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2595 timeout during the DATA phase of an incoming message.
2597 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2598 applied a patch sanctioned by the Sieve maintainer.
2600 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2601 the uid or gid is negative. A case of a negative gid caused this to be
2602 noticed. The fix allows for either to be negative.
2604 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2605 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2608 PH/36 The change PH/12 above was broken. Fixed it.
2610 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2611 the grounds that routing the same address twice would always produce the
2612 same answer. This might have been true once, but it is certainly no
2613 longer true now. Routing a child address may depend on the previous
2614 routing that produced that child. Some complicated redirection strategies
2615 went wrong when messages had multiple recipients, and made Exim's
2616 behaviour dependent on the order in which the addresses were given.
2618 I have moved the duplicate checking until after the routing is complete.
2619 Exim scans the addresses that are assigned to local and remote
2620 transports, and removes any duplicates. This means that more work will be
2621 done, as duplicates will always all be routed, but duplicates are
2622 presumably rare, so I don't expect this is of any significance.
2624 For deliveries to pipes, files, and autoreplies, the duplicate checking
2625 still happens during the routing process, since they are not going to be
2628 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2629 It corrects a timeout issue with spamd. This is Ian's comment: "The
2630 background is that sometimes spamd either never reads data from a
2631 connection it has accepted, or it never writes response data. The exiscan
2632 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2633 blindly assumes that writes won't block so it may never time out."
2635 PH/39 Allow G after quota size as well as K and M.
2637 PH/40 The value set for $authenticated_id in an authenticator may not contain
2638 binary zeroes or newlines because the value is written to log lines and
2639 to spool files. There was no check on this. Now the value is run through
2640 the string_printing() function so that such characters are converted to
2641 printable escape sequences.
2643 PH/41 $message_linecount is a new variable that contains the total number of
2644 lines in the message. Compare $body_linecount, which is the count for the
2647 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2648 bounce and delay warning messages, except in certain special cases, which
2651 (a) An SMTP error message from a remote host;
2652 (b) A message specified in a :fail: redirection;
2653 (c) A message specified in a "fail" command in a system filter;
2654 (d) A message specified in a FAIL return from the queryprogram router;
2655 (e) A message specified by the cannot_route_message router option.
2657 In these cases only, Exim does include the error details in bounce and
2658 warning messages. There are also a few cases where bland messages such
2659 as "unrouteable address" or "local delivery error" are given.
2661 PH/43 $value is now also set for the "else" part of a ${run expansion.
2663 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2664 being worked on, but at least Exim now implements the latest version to
2667 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2668 process to complete was treated as a delivery failure, a timeout while
2669 writing the message to the pipe was logged, but erroneously treated as a
2670 successful delivery. Such timeouts include transport filter timeouts. For
2671 consistency with the overall process timeout, these timeouts are now
2672 treated as errors, giving rise to delivery failures by default. However,
2673 there is now a new Boolean option for the pipe transport called
2674 timeout_defer, which, if set TRUE, converts the failures into defers for
2675 both kinds of timeout. A transport filter timeout is now identified in
2678 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2679 systems where "make" and "gmake" are different, calling "gmake" at top
2680 level broke things. I've arranged for the value of $(MAKE) to be passed
2681 from the Makefile to this script so that it can call the same version of
2685 A note about Exim versions 4.44 and 4.50
2686 ----------------------------------------
2688 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2689 changes of various kinds. As a consequence, a big documentation update was
2690 needed. This delayed the release for rather longer than seemed good, especially
2691 in the light of a couple of (minor) security issues. Therefore, the changes
2692 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2693 release. So 4.44 and 4.50 are in effect two different branches that both start
2696 I have left the 4.50 change log unchanged; it contains all the changes since
2697 4.43. The change log for 4.44 is below; many of its items are identical to
2698 those for 4.50. This seems to be the most sensible way to preserve the
2699 historical information.
2705 1. Minor wording change to the doc/README.SIEVE file.
2707 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2708 computation of the current number of files was incorrect.
2710 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2711 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2714 4. Give more explanation in the error message when the command for a transport
2715 filter fails to execute.
2717 5. There are several places where Exim runs a non-Exim command in a
2718 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2719 was being done only for the command run by the queryprogram router. It is
2720 now done for all such subprocesses. The other cases are: ${run, transport
2721 filters, and the commands run by the lmtp and pipe transports.
2723 6. Added CONFIGURE_GROUP build-time option.
2725 7. Some older OS have a limit of 256 on the maximum number of file
2726 descriptors. Exim was using setrlimit() to set 1000 as a large value
2727 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2728 systems. I've change it so that if it can't get 1000, it tries for 256.
2730 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2731 was an oversight, and furthermore, ever since the addition of extra
2732 controls (e.g. 4.43/32), the checks on when to allow different forms of
2733 "control" were broken. There should now be diagnostics for all cases when a
2734 control that does not make sense is encountered.
2736 9. Added the /retain_sender option to "control=submission".
2738 10. $recipients is now available in the predata ACL (oversight).
2740 11. Tidy the search cache before the fork to do a delivery from a message
2741 received from the command line. Otherwise the child will trigger a lookup
2742 failure and thereby defer the delivery if it tries to use (for example) a
2743 cached ldap connection that the parent has called unbind on.
2745 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2746 of $address_data from the recipient verification was clobbered by the
2747 sender verification.
2749 13. The value of address_data from a sender verification is now available in
2750 $sender_address_data in subsequent conditions in the ACL statement.
2752 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2754 15. Added a new option "connect=<time>" to callout options, to set a different
2757 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2758 was its contents. (It was OK if the option was not defined at all.)
2760 17. A "Completed" log line is now written for messages that are removed from
2761 the spool by the -Mrm option.
2763 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2764 information about exactly what failed.
2766 19. Added -dd to debug only the daemon process.
2768 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2769 handles timeouts, both on the server side and network timeouts. Renamed the
2770 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2772 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2773 It is now set to "smtps".
2775 22. $host_address is now set to the target address during the checking of
2776 ignore_target_hosts.
2778 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2779 being passed; this would have caused $sender_host_name to have been used if
2780 matching the list had actually called for a host name (not very likely,
2781 since this list is usually IP addresses). A host name is now passed as
2784 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2785 code that specifies a non-restarting handler (typically sigaction() in
2786 modern systems) in an attempt to fix a rare and obscure crash bug.
2788 25. Narrowed the window for a race in the daemon that could cause it to ignore
2789 SIGCHLD signals. This is not a major problem, because they are used only to
2790 wake it up if nothing else does.
2792 26. A malformed maildirsize file could cause Exim to calculate negative values
2793 for the mailbox size or file count. Odd effects could occur as a result.
2794 The maildirsize information is now recalculated if the size or filecount
2797 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2798 support for a long time. Removed HAVE_SYS_VFS_H.
2800 28. Installed the latest version of exipick from John Jetmore.
2802 29. In an address list, if the pattern was not a regular expression, an empty
2803 subject address (from a bounce message) matched only if the pattern was an
2804 empty string. Non-empty patterns were not even tested. This was the wrong
2805 because it is perfectly reasonable to use an empty address as part of a
2806 database query. An empty address is now tested by patterns that are
2807 lookups. However, all the other forms of pattern expect the subject to
2808 contain a local part and a domain, and therefore, for them, an empty
2809 address still always fails if the pattern is not itself empty.
2811 30. Exim went into a mad DNS loop when attempting to do a callout where the
2812 host was specified on an smtp transport, and looking it up yielded more
2813 than one IP address.
2815 31. Re-factored the code for checking spool and log partition space into a
2816 function that finds that data and another that does the check. The former
2817 is then used to implement four new variables: $spool_space, $log_space,
2818 $spool_inodes, and $log_inodes.
2820 32. The RFC2047 encoding function was originally intended for short strings
2821 such as real names; it was not keeping to the 75-character limit for
2822 encoded words that the RFC imposes. It now respects the limit, and
2823 generates multiple encoded words if necessary. To be on the safe side, I
2824 have increased the buffer size for the ${rfc2047: expansion operator from
2827 33. It is now permitted to omit both strings after an "if" condition; if the
2828 condition is true, the result is "true". As before, when the second string
2829 is omitted, a false condition yields an empty string. This makes it less
2830 cumbersome to write custom ACL and router conditions.
2832 34. Failure to deliver a bounce message always caused it to be frozen, even if
2833 there was an errors_to setting on the router. The errors_to setting is now
2836 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2837 canonical form (fully expanded) before being placed in
2838 $sender_host_address.
2840 36. The table in the code that translates DNS record types into text (T_A to
2841 "A" for instance) was missing entries for NS and CNAME. It is just possible
2842 that this could have caused confusion if both these types were looked up
2843 for the same domain, because the text type is used as part of Exim's
2844 per-process caching. But the chance of anyone hitting this buglet seems
2847 37. The dnsdb lookup has been extended in a number of ways.
2849 (1) There is a new type, "zns", which walks up the domain tree until it
2850 finds some nameserver records. It should be used with care.
2852 (2) There is a new type, "mxh", which is like "mx" except that it returns
2853 just the host names, not the priorities.
2855 (3) It is now possible to give a list of domains (or IP addresses) to be
2856 looked up. The behaviour when one of the lookups defers can be
2857 controlled by a keyword.
2859 (4) It is now possible to specify the separator character for use when
2860 multiple records are returned.
2862 38. The dnslists ACL condition has been extended: it is now possible to supply
2863 a list of IP addresses and/or domains to be looked up in a particular DNS
2866 39. Added log_selector=+queue_time_overall.
2868 40. When running the queue in the test harness, wait just a tad after forking a
2869 delivery process, to get repeatability of debugging output.
2871 41. Include certificate and key file names in error message when GnuTLS fails
2872 to set them up, because the GnuTLS error message doesn't include the name
2873 of the failing file when there is a problem reading it.
2875 42. Allow both -bf and -bF in the same test run.
2877 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2879 44. The "Exiscan patch" is now merged into the mainline Exim source.
2881 45. Sometimes the final signoff response after QUIT could fail to get
2882 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2883 < 0 before doing a fflush(). This bug looks as though it goes back to the
2884 introduction of TLS in release 3.20, but "sometimes" must have been rare
2885 because the tests only now provoked it.
2887 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2888 (this can affect the format of dates).
2890 47. exim_tidydb, when checking for the continued existence of a message for
2891 which it has found a message-specific retry record, was not finding
2892 messages that were in split spool directories. Consequently, it was
2893 deleting retry records that should have stayed in existence.
2895 48. Steve fixed some bugs in eximstats.
2897 49. The SPA authentication driver was not abandoning authentication and moving
2898 on to the next authenticator when an expansion was forced to fail,
2899 contradicting the general specification for all authenticators. Instead it
2900 was generating a temporary error. It now behaves as specified.
2902 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2903 (the order specifies the preference for clients). The order is now AES256,
2904 AES128, 3DES, ARCFOUR128.
2906 51. Small patch to Sieve code - explicitly set From: when generating an
2909 52. Exim crashed if a remote delivery caused a very long error message to be
2910 recorded - for instance if somebody sent an entire SpamAssassin report back
2911 as a large number of 550 error lines. This bug was coincidentally fixed by
2912 increasing the size of one of Exim's internal buffers (big_buffer) that
2913 happened as part of the Exiscan merge. However, to be on the safe side, I
2914 have made the code more robust (and fixed the comments that describe what
2917 53. Now that there can be additional text after "Completed" in log lines (if
2918 the queue_time_overall log selector is set), a one-byte patch to exigrep
2919 was needed to allow it to recognize "Completed" as not the last thing in
2922 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2923 patch that reportedly fixes this has been added. I am not expert enough to
2924 create a test for it. This is what the patch creator wrote:
2926 "I found a little strange behaviour of ldap code when working with
2927 Windows 2003 AD Domain, where users was placed in more than one
2928 Organization Units. When I tried to give exim partial DN, the exit code
2929 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2930 But simultaneously result of request was absolutely normal ldap result,
2931 so I produce this patch..."
2933 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2934 REFERENCE, so I have modified the code to exclude the patch when that macro
2937 55. Some experimental protocols are using DNS PTR records for new purposes. The
2938 keys for these records are domain names, not reversed IP addresses. The
2939 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2940 leaves it alone. Component reversal etc. now happens only for IP addresses.
2943 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2945 57. Double the size of the debug message buffer (to 2048) so that more of very
2946 long debug lines gets shown.
2948 58. The exicyclog utility now does better if the number of log files to keep
2949 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2951 59. Two changes related to the smtp_active_hostname option:
2953 (1) $smtp_active_hostname is now available as a variable.
2954 (2) The default for smtp_banner uses $smtp_active_hostname instead
2955 of $primary_hostname.
2957 60. The host_aton() function is supposed to be passed a string that is known
2958 to be a valid IP address. However, in the case of IPv6 addresses, it was
2959 not checking this. This is a hostage to fortune. Exim now panics and dies
2960 if the condition is not met. A case was found where this could be provoked
2961 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2962 components; fortuitously, this particular loophole had already been fixed
2963 by change 4.50/55 above.
2965 If there are any other similar loopholes, the new check in host_aton()
2966 itself should stop them being exploited. The report I received stated that
2967 data on the command line could provoke the exploit when Exim was running as
2968 exim, but did not say which command line option was involved. All I could
2969 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2970 running as the user.
2973 61. There was a buffer overflow vulnerability in the SPA authentication code
2974 (which came originally from the Samba project). I have added a test to the
2975 spa_base64_to_bits() function which I hope fixes it.
2978 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2979 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2981 63. The daemon start-up calls getloadavg() while still root for those OS that
2982 need the first call to be done as root, but it missed one case: when
2983 deliver_queue_load_max is set with deliver_drop_privilege. This is
2984 necessary for the benefit of the queue runner, because there is no re-exec
2985 when deliver_drop_privilege is set.
2987 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2988 This has been fixed.
2990 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2991 was in use, was not putting the data itself into the right store pool;
2992 consequently, it could be overwritten for a subsequent message in the same
2993 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2996 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2998 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2999 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3000 the function were treating the return as a boolean value, which happened to
3001 work because 0=false and not-0=true, but is not correct code.
3003 68. The host_aton() function was not handling scoped IPv6 addresses (those
3004 with, for example, "%eth0" on the end) correctly.
3006 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3007 negated items (that is, ~something) in unsigned ints. Some compilers
3008 apparently mutter when there is no cast.
3010 70. If an address verification called from an ACL failed, and did not produce a
3011 user-specific message (i.e. there was only a "system" message), nothing was
3012 put in $acl_verify_message. In this situation, it now puts the system
3015 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3016 session; change 4.31/43 added the unwanted input to the log line - except
3017 that it did not do this in the start of session case. It now does.
3019 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3020 This is wrong and can cause the other end to generate a synchronization
3021 error if it is another Exim or anything else that does the synchronization
3022 check. A QUIT command is no longer sent after a timeout.
3024 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3025 during host lookups.
3027 74. The defer_ok option of callout verification was not working if it was used
3028 when verifying addresses in header lines, that is, for this case:
3030 verify = header_sender/callout=defer_ok
3032 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3033 those file descriptors could be used for SMTP connections. If anything
3034 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3035 could be sent to the SMTP client, causing chaos. The daemon now opens
3036 stdin, stdout, and stderr to /dev/null when it puts itself into the
3039 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3040 log by default. The user can override this with suitable Perl magic.
3042 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3043 the log message when discard triggers, was not working for the DATA ACL or
3044 for the non-SMTP ACL.
3046 78. Error message wording change in sieve.c.
3048 79. If smtp_accept_max_per_host was set, the number of connections could be
3049 restricted to fewer than expected, because the daemon was trying to set up
3050 a new connection before checking whether the processes handling previous
3051 connections had finished. The check for completed processes is now done
3052 earlier. On busy systems, this bug wouldn't be noticed because something
3053 else would have woken the daemon, and it would have reaped the completed
3056 80. If a message was submitted locally by a user whose login name contained one
3057 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3058 It caused a spool format error. I have fixed the spool reading code. A
3059 related problem was that the "from" clause in the Received: line became
3060 illegal because of the space(s). It is now covered by ${quote_local_part.
3062 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3065 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3067 83. Patch from Sieve maintainer to fix unterminated string problem in
3068 "vacation" handling.
3070 84. Some minor changes to the Linux configuration files to help with other
3071 OS variants using glibc.
3073 85. One more patch for Sieve to update vacation handling to latest spec.
3076 ----------------------------------------------------
3077 See the note above about the 4.44 and 4.50 releases.
3078 ----------------------------------------------------
3084 1. Change 4.43/35 introduced a bug that caused file counts to be
3085 incorrectly computed when quota_filecount was set in an appendfile
3088 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3089 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3092 3. Give more explanation in the error message when the command for a transport
3093 filter fails to execute.
3095 4. There are several places where Exim runs a non-Exim command in a
3096 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3097 was being done only for the command run by the queryprogram router. It is
3098 now done for all such subprocesses. The other cases are: ${run, transport
3099 filters, and the commands run by the lmtp and pipe transports.
3101 5. Some older OS have a limit of 256 on the maximum number of file
3102 descriptors. Exim was using setrlimit() to set 1000 as a large value
3103 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3104 systems. I've change it so that if it can't get 1000, it tries for 256.
3106 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3107 was an oversight, and furthermore, ever since the addition of extra
3108 controls (e.g. 4.43/32), the checks on when to allow different forms of
3109 "control" were broken. There should now be diagnostics for all cases when a
3110 control that does not make sense is encountered.
3112 7. $recipients is now available in the predata ACL (oversight).
3114 8. Tidy the search cache before the fork to do a delivery from a message
3115 received from the command line. Otherwise the child will trigger a lookup
3116 failure and thereby defer the delivery if it tries to use (for example) a
3117 cached ldap connection that the parent has called unbind on.
3119 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3120 of $address_data from the recipient verification was clobbered by the
3121 sender verification.
3123 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3124 was its contents. (It was OK if the option was not defined at all.)
3126 11. A "Completed" log line is now written for messages that are removed from
3127 the spool by the -Mrm option.
3129 12. $host_address is now set to the target address during the checking of
3130 ignore_target_hosts.
3132 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3133 being passed; this would have caused $sender_host_name to have been used if
3134 matching the list had actually called for a host name (not very likely,
3135 since this list is usually IP addresses). A host name is now passed as
3138 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3139 code that specifies a non-restarting handler (typically sigaction() in
3140 modern systems) in an attempt to fix a rare and obscure crash bug.
3142 15. Narrowed the window for a race in the daemon that could cause it to ignore
3143 SIGCHLD signals. This is not a major problem, because they are used only to
3144 wake it up if nothing else does.
3146 16. A malformed maildirsize file could cause Exim to calculate negative values
3147 for the mailbox size or file count. Odd effects could occur as a result.
3148 The maildirsize information is now recalculated if the size or filecount
3151 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3152 support for a long time. Removed HAVE_SYS_VFS_H.
3154 18. Updated exipick to current release from John Jetmore.
3156 19. Allow an empty sender to be matched against a lookup in an address list.
3157 Previously the only cases considered were a regular expression, or an
3160 20. Exim went into a mad DNS lookup loop when doing a callout where the
3161 host was specified on the transport, if the DNS lookup yielded more than
3164 21. The RFC2047 encoding function was originally intended for short strings
3165 such as real names; it was not keeping to the 75-character limit for
3166 encoded words that the RFC imposes. It now respects the limit, and
3167 generates multiple encoded words if necessary. To be on the safe side, I
3168 have increased the buffer size for the ${rfc2047: expansion operator from
3171 22. Failure to deliver a bounce message always caused it to be frozen, even if
3172 there was an errors_to setting on the router. The errors_to setting is now
3175 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3176 canonical form (fully expanded) before being placed in
3177 $sender_host_address.
3179 24. Updated eximstats to version 1.33
3181 25. Include certificate and key file names in error message when GnuTLS fails
3182 to set them up, because the GnuTLS error message doesn't include the name
3183 of the failing file when there is a problem reading it.
3185 26. Expand error message when OpenSSL has problems setting up cert/key files.
3188 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3189 (this can affect the format of dates).
3191 28. exim_tidydb, when checking for the continued existence of a message for
3192 which it has found a message-specific retry record, was not finding
3193 messages that were in split spool directories. Consequently, it was
3194 deleting retry records that should have stayed in existence.
3196 29. eximstats updated to version 1.35
3197 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3198 1.35 - bugfix such that pie charts by volume are generated correctly
3200 30. The SPA authentication driver was not abandoning authentication and moving
3201 on to the next authenticator when an expansion was forced to fail,
3202 contradicting the general specification for all authenticators. Instead it
3203 was generating a temporary error. It now behaves as specified.
3205 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3206 (the order specifies the preference for clients). The order is now AES256,
3207 AES128, 3DES, ARCFOUR128.
3209 31. Small patch to Sieve code - explicitly set From: when generating an
3212 32. Exim crashed if a remote delivery caused a very long error message to be
3213 recorded - for instance if somebody sent an entire SpamAssassin report back
3214 as a large number of 550 error lines. This bug was coincidentally fixed by
3215 increasing the size of one of Exim's internal buffers (big_buffer) that
3216 happened as part of the Exiscan merge. However, to be on the safe side, I
3217 have made the code more robust (and fixed the comments that describe what
3220 33. Some experimental protocols are using DNS PTR records for new purposes. The
3221 keys for these records are domain names, not reversed IP addresses. The
3222 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3223 leaves it alone. Component reversal etc. now happens only for IP addresses.
3226 34. The host_aton() function is supposed to be passed a string that is known
3227 to be a valid IP address. However, in the case of IPv6 addresses, it was
3228 not checking this. This is a hostage to fortune. Exim now panics and dies
3229 if the condition is not met. A case was found where this could be provoked
3230 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3231 components; fortuitously, this particular loophole had already been fixed
3232 by change 4.50/55 or 4.44/33 above.
3234 If there are any other similar loopholes, the new check in host_aton()
3235 itself should stop them being exploited. The report I received stated that
3236 data on the command line could provoke the exploit when Exim was running as
3237 exim, but did not say which command line option was involved. All I could
3238 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3239 running as the user.
3242 35. There was a buffer overflow vulnerability in the SPA authentication code
3243 (which came originally from the Samba project). I have added a test to the
3244 spa_base64_to_bits() function which I hope fixes it.
3247 36. The daemon start-up calls getloadavg() while still root for those OS that
3248 need the first call to be done as root, but it missed one case: when
3249 deliver_queue_load_max is set with deliver_drop_privilege. This is
3250 necessary for the benefit of the queue runner, because there is no re-exec
3251 when deliver_drop_privilege is set.
3253 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3254 was in use, was not putting the data itself into the right store pool;
3255 consequently, it could be overwritten for a subsequent message in the same
3256 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3259 38. Sometimes the final signoff response after QUIT could fail to get
3260 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3261 < 0 before doing a fflush(). This bug looks as though it goes back to the
3262 introduction of TLS in release 3.20, but "sometimes" must have been rare
3263 because the tests only now provoked it.
3269 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3270 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3271 Now it returns an integer. A number of calls were still expecting a T/F
3272 return. Fortuitously, in all cases, the tests worked in OK situations,
3273 which is the norm. However, things would have gone wrong on any write
3274 failures on the smtp file descriptor. This function is used when sending
3275 messages over SMTP and also when doing verify callouts.
3277 2. When Exim is called to do synchronous delivery of a locally submitted
3278 message (the -odf or -odi options), it no longer closes stderr before doing
3281 3. Implemented the mua_wrapper option.
3283 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3285 5. Implemented the functions header_remove(), header_testname(),
3286 header_add_at_position(), and receive_remove_recipient(), and exported them
3289 6. If an ACL "warn" statement specified the addition of headers, Exim already
3290 inserted X-ACL-Warn: at the start if there was no header name. However, it
3291 was not making this test for the second and subsequent header lines if
3292 there were newlines in the string. This meant that an invalid header could
3293 be inserted if Exim was badly configured.
3295 7. Allow an ACL "warn" statement to add header lines at the start or after all
3296 the Received: headers, as well as at the end.
3298 8. Added the rcpt_4xx retry error code.
3300 9. Added postmaster_mailfrom=xxx to callout verification option.
3302 10. Added mailfrom=xxxx to the callout verification option, for verify=
3305 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3306 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3308 12. Inserted some casts to stop certain compilers warning when using pointer
3309 differences as field lengths or precisions in printf-type calls (mostly
3310 affecting debugging statements).
3312 13. Added optional readline() support for -be (dynamically loaded).
3314 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3315 same clock tick as a message's arrival, so that its received time was the
3316 same as the "first fail" time on the retry record, and that message
3317 remained on the queue past the ultimate address timeout, every queue runner
3318 would try a delivery (because it was past the ultimate address timeout) but
3319 after another failure, the ultimate address timeout, which should have then
3320 bounced the address, did not kick in. This was a "< instead of <=" error;
3321 in most cases the first failure would have been in the next clock tick
3322 after the received time, and all would be well.
3324 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3325 being recognized when the domain list was tested by the match_domain
3326 condition in an expansion string.
3328 16. Added the ${str2b64: operator.
3330 17. Exim was always calling setrlimit() to set a large limit for the number of
3331 processes, without checking whether the existing limit was already
3332 adequate. (It did check for the limit on file descriptors.) Furthermore,
3333 errors from getrlimit() and setrlimit() were being ignored. Now they are
3334 logged to the main and panic logs, but Exim does carry on, to try to do its
3335 job under whatever limits there are.
3337 18. Imported PCRE 5.0.
3339 19. Trivial typo in log message " temporarily refused connection" (the leading
3342 20. If the log selector return_path_on_delivery was set and an address was
3343 redirected to /dev/null, the delivery process crashed because it assumed
3344 that a return path would always be set for a "successful" delivery. In this
3345 case, the whole delivery is bypassed as an optimization, and therefore no
3348 21. Internal re-arrangement: the function for sending a challenge and reading
3349 a response while authentication was assuming a zero-terminated challenge
3350 string. It's now changed to take a pointer and a length, to allow for
3351 binary data in such strings.
3353 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3355 23. Exim was not respecting finduser_retries when seeking the login of the
3356 uid under which it was called; it was always trying 10 times. (The default
3357 setting of finduser_retries is zero.) Also, it was sleeping after the final
3358 failure, which is pointless.
3360 24. Implemented tls_on_connect_ports.
3362 25. Implemented acl_smtp_predata.
3364 26. If the domain in control=submission is set empty, Exim assumes that the
3365 authenticated id is a complete email address when it generates From: or
3366 Sender: header lines.
3368 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3369 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3370 chown and chgrp in /bin and hostname in /usr/bin.
3372 28. Exim was keeping the "process log" file open after each use, just as it
3373 does for the main log. This opens the possibility of it remaining open for
3374 long periods when the USR1 signal hits a daemon. Occasional processlog
3375 errors were reported, that could have been caused by this. Anyway, it seems
3376 much more sensible not to leave this file open at all, so that is what now
3379 29. The long-running daemon process does not normally write to the log once it
3380 has entered its main loop, and it closes the log before doing so. This is
3381 so that log files can straightforwardly be renamed and moved. However,
3382 there are a couple of unusual error situations where the daemon does write
3383 log entries, and I had neglected to close the log afterwards.
3385 30. The text of an SMTP error response that was received during a remote
3386 delivery was being truncated at 512 bytes. This is too short for some of
3387 the long messages that one sometimes sees. I've increased the limit to
3390 31. It is now possible to make retry rules that apply only when a message has a
3391 specific sender, in particular, an empty sender.
3393 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3394 it possible to be selective about when SMTP synchronization is enforced.
3396 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3398 32. Implemented hosts_connection_nolog.
3400 33. Added an ACL for QUIT.
3402 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3405 35. Added mailbox_size and mailbox_filecount to appendfile.
3407 36. Added control = no_multiline_responses to ACLs.
3409 37. There was a bug in the logic of the code that waits for the clock to tick
3410 in the case where the clock went backwards by a substantial amount such
3411 that the microsecond fraction of "now" was more than the microsecond
3412 fraction of "then" (but the whole seconds number was less).
3414 38. Added support for the libradius Radius client library this is found on
3415 FreeBSD (previously only the radiusclient library was supported).
3421 1. When certain lookups returned multiple values in the form name=value, the
3422 quoting of the values was not always being done properly. Specifically:
3423 (a) If the value started with a double quote, but contained no whitespace,
3425 (b) If the value contained whitespace other than a space character (i.e.
3426 tabs or newlines or carriage returns) it was not quoted.
3427 This fix has been applied to the mysql and pgsql lookups by writing a
3428 separate quoting function and calling it from the lookup code. The fix
3429 should probably also be applied to nisplus, ibase and oracle lookups, but
3430 since I cannot test any of those, I have not disturbed their existing code.
3432 2. A hit in the callout cache for a specific address caused a log line with no
3433 reason for rejecting RCPT. Now it says "Previous (cached) callout
3434 verification failure".
3436 3. There was an off-by-one bug in the queryprogram router. An over-long
3437 return line was truncated at 256 instead of 255 characters, thereby
3438 overflowing its buffer with the terminating zero. As well as fixing this, I
3439 have increased the buffer size to 1024 (and made a note to document this).
3441 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3442 when Exim is waiting for an SMTP response from a remote server, Exim
3443 restarts its select() call on the socket, thereby resetting its timeout.
3444 This is not a problem when such interrupts are rare. Somebody set up a cron
3445 job to run exiwhat every 2 minutes, which is less than the normal select()
3446 timeout (5 or 10 minutes). This meant that the select() timeout never
3447 kicked in because it was always reset. I have fixed this by comparing the
3448 time when an interrupt arrives with the time at the start of the first call
3449 to select(). If more time than the timeout has elapsed, the interrupt is
3450 treated as a timeout.
3452 5. Some internal re-factoring in preparation for the addition of Sieve
3453 extensions (by MH). In particular, the "personal" test is moved to a
3454 separate function, and given an option for scanning Cc: and Bcc: (which is
3455 not set for Exim filters).
3457 6. When Exim created an email address using the login of the caller as the
3458 local part (e.g. when creating a From: or Sender: header line), it was not
3459 quoting the local part when it contained special characters such as @.
3461 7. Installed new OpenBSD configuration files.
3463 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3464 try to make them clearer.
3466 9. Callout options, other than the timeout value, were being ignored when
3467 verifying sender addresses in header lines. For example, when using
3469 verify = header_sender/callout=no_cache
3471 the cache was (incorrectly) being used.
3473 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3474 only the Cygwin environment.
3476 11. When return_path_on_delivery was set as a log selector, if different remote
3477 addresses in the same message used different return paths and parallel
3478 remote delivery occurred, the wrong values would sometimes be logged.
3479 (Whenever a remote delivery process finished, the return path value from
3480 the most recently started remote delivery process was logged.)
3482 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3483 lines when AUTH and/or TLS are in use. This is the "received protocol"
3484 field. Exim used to use "asmtp" for authenticated SMTP, without any
3485 indication (in the protocol name) for TLS use. Now it follows the RFC and
3486 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3487 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3488 names appear in log lines as well as in Received: header lines.
3490 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3491 extensions, and comparison tests, and to fix some bugs.
3493 14. Changes to the "personal" filter test:
3495 (1) The test was buggy in that it was just doing the equivalent of
3496 "contains" tests on header lines. For example, if a user's address was
3497 anne@some.where, the "personal" test would incorrectly be true for
3499 To: susanne@some.where
3501 This test is now done by extracting each address from the header in turn,
3502 and checking the entire address. Other tests that are part of "personal"
3503 are now done using regular expressions (for example, to check local parts
3504 of addresses in From: header lines).
3506 (2) The list of non-personal local parts in From: addresses has been
3507 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3508 taken from the Sieve specification recommendations.
3510 (3) If the message contains any header line starting with "List-" it is
3511 treated as non-personal.
3513 (4) The test for "circular" in the Subject: header line has been removed
3514 because it now seems ill-conceived.
3516 15. Minor typos in src/EDITME comments corrected.
3518 16. Installed latest exipick from John Jetmore.
3520 17. If headers_add on a router specified a text string that was too long for
3521 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3522 of string_sprintf() is now avoided.
3524 18. $message_body_size was not set (it was always zero) when running the DATA
3525 ACL and the local_scan() function.
3527 19. For the "mail" command in an Exim filter, no default was being set for
3528 the once_repeat time, causing a random time value to be used if "once" was
3529 specified. (If the value happened to be <= 0, no repeat happened.) The
3530 default is now 0s, meaning "never repeat". The "vacation" command was OK
3531 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3532 (I found it when inspecting the code).
3534 20. There is now an overall timeout for performing a callout verification. It
3535 defaults to 4 times the callout timeout, which applies to individual SMTP
3536 commands during the callout. The overall timeout applies when there is more
3537 than one host that can be tried. The timeout is checked before trying the
3538 next host. This prevents very long delays if there are a large number of
3539 hosts and all are timing out (e.g. when the network connections are timing
3540 out). The value of the overall timeout can be changed by specifying an
3541 additional sub-option for "callout", called "maxwait". For example:
3543 verify = sender/callout=5s,maxwait=20s
3545 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3546 to the end before writing, but this should make it even safer).
3548 22. Exim was forgetting that it had advertised PIPELINING for the second and
3549 subsequent messages on an SMTP connection. It was also not resetting its
3550 memory on STARTTLS and an internal HELO.
3552 23. When Exim logs an SMTP synchronization error within a session, it now
3553 records whether PIPELINING has been advertised or not.
3555 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3556 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3557 rather than long int.
3559 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3561 26. Added the never_mail option to autoreply.
3567 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3568 crash if the getsockname() call failed; this can happen if a connection is
3569 closed very soon after it is established. The problem was simply in the
3570 order in which certain operations were done, causing Exim to try to write
3571 to the SMTP stream before it had set up the file descriptor. The bug has
3572 been fixed by making things happen in the correct order.
3578 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3579 before the connection was closed, thus losing the rejection response.
3581 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3582 some early Solaris releases, but causes trouble in current releases where
3583 socklen_t is defined.
3585 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3588 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3591 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3592 that releasing the top of it at the end releases what was used for sub-
3593 expansions (unless the block got too big). However, discard this block if
3594 the first thing is a variable or header, so that we can use its block when
3595 it is dynamic (useful for very large $message_headers, for example).
3597 6. Lookups now cache *every* query, not just the most recent. A new, separate
3598 store pool is used for this. It can be recovered when all lookup caches are
3599 flushed. Lookups now release memory at the end of their result strings.
3600 This has involved some general refactoring of the lookup sources.
3602 7. Some code has been added to the store_xxx() functions to reduce the amount
3603 of flapping under certain conditions.
3605 8. log_incoming_interface used to affect only the <= reception log lines. Now
3606 it causes the local interface and port to be added to several more SMTP log
3607 lines, for example "SMTP connection from", and rejection lines.
3609 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3611 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3613 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3614 could be overwritten at the end of the current message (or the start of a
3615 new message if it was set in a HELO ACL). The value is now preserved for
3616 the duration of the SMTP connection.
3618 12. If a transport had a headers_rewrite setting, and a matching header line
3619 contained an unqualified address, that address was qualified, even if it
3620 did not match any rewriting rules. The underlying bug was that the values
3621 of the flags that permit the existence of unqualified sender and recipient
3622 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3623 non-local messages, and by -bnq for local messages) were not being
3624 preserved with the message after it was received.
3626 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3627 "next input=" as part of the text comprising the host identity instead of
3628 the correct text. The code was using the same buffer for two different
3629 strings. However, depending on which order the printing function evaluated
3630 its arguments, the bug did not always show up. Under Linux, for example, my
3631 test suite worked just fine.
3633 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3634 doesn't work with some older versions of Perl. It has been changed to "my",
3635 which in any case is probably the better facility to use.
3637 15. A really picky compiler found some instances of statements for creating
3638 error messages that either had too many or two few arguments for the format
3641 16. The size of the buffer for calls to the DNS resolver has been increased
3642 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3643 for addresses that have a lot of PTR records. This alleviates a problem; it
3644 does not fully solve it.
3646 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3647 buffer now truncates the list and logs the incident, which is the same
3648 action as happens when Exim is looking up a host name and its aliases.
3649 Previously in this situation something unpredictable would happen;
3650 sometimes it was "internal error: store_reset failed".
3652 18. If a server dropped the connection unexpectedly when an Exim client was
3653 using GnuTLS and trying to read a response, the client delivery process
3654 crashed while trying to generate an error log message.
3656 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3657 string, for example:
3659 warn message = H1: something\nH2: something
3661 the text was added as a single header line from Exim's point of view
3662 though it ended up OK in the delivered message. However, searching for the
3663 second and subsequent header lines using $h_h2: did not work. This has been
3664 fixed. Similarly, if a system filter added multiple headers in this way,
3665 the routers could not see them.
3667 20. Expanded the error message when iplsearch is called with an invalid key to
3668 suggest using net-iplsearch in a host list.
3670 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3671 ACLs are no longer actually imposed (and a message to that effect is
3674 22. If a "gecos" field in a passwd entry contained escaped characters, in
3675 particular, if it contained a \" sequence, Exim got it wrong when building
3676 a From: or a Sender: header from that name. A second bug also caused
3677 incorrect handling when an unquoted " was present following a character
3678 that needed quoting.
3680 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3681 was not being matched caselessly.
3683 24. Arranged for all hyphens in the exim.8 source to be escaped with
3686 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3687 properly. Recipient callout cache records were still being keyed to include
3688 the sender, even when use_sender was set false. This led to far more
3689 callouts that were necessary. The sender is no longer included in the key
3690 when use_sender is false.
3692 26. Added "control = submission" modifier to ACLs.
3694 27. Added the ${base62d: operator to decode base 62 numbers.
3696 28. dnsdb lookups can now access SRV records.
3698 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3699 the configuration file.
3701 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3702 (-v) mode. This makes the output for a verbose queue run more intelligible.
3704 31. Added a use_postmaster feature to recipient callouts.
3706 32. Added the $body_zerocount variable, containing the number of binary zero
3707 bytes in the message body.
3709 33. The time of last modification of the "new" subdirectory is now used as the
3710 "mailbox time last read" when there is a quota error for a maildir
3713 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3715 35. Added +ignore_unknown as a special item in host lists.
3717 36. Code for decoding IPv6 addresses in host lists is now included, even if
3718 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3719 address was recognized as an IP address, but was then not correctly decoded
3720 into binary, causing unexpected and incorrect effects when compared with
3727 1. Very minor rewording of debugging text in manualroute to say "list of
3728 hosts" instead of "hostlist".
3730 2. If verify=header_syntax was set, and a header line with an unqualified
3731 address (no domain) and a large number of spaces between the end of the
3732 name and the colon was received, the reception process suffered a buffer
3733 overflow, and (when I tested it) crashed. This was caused by some obsolete
3734 code that should have been removed. The fix is to remove it!
3736 3. When running in the test harness, delay a bit after writing a bounce
3737 message to get a bit more predictability in the log output.
3739 4. Added a call to search_tidyup() just before forking a reception process. In
3740 theory, someone could use a lookup in the expansion of smtp_accept_max_
3741 per_host which, without the tidyup, could leave open a database connection.
3743 5. Added the variables $recipient_data and $sender_data which get set from a
3744 lookup success in an ACL "recipients" or "senders" condition, or a router
3745 "senders" option, similar to $domain_data and $local_part_data.
3747 6. Moved the writing of debug_print from before to after the "senders" test
3750 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3751 problems for message scanning, either using a data ACL, or using
3752 local_scan() because the Received: header was not generated till after they
3753 were called (in order to set the time as the time of reception completion).
3754 I have revised the way this works. The header is now generated after the
3755 body is received, but before the ACL or local_scan() are called. After they
3756 are run, the timestamp in the header is updated.
3762 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3763 before starting a queue runner without re-exec. This happened only when
3764 deliver_drop_privilege was set or when the Exim user was set to root. The
3765 effect of the bug was that timeouts during subsequent deliveries caused
3766 crashes instead of being properly handled. The handler is now left at its
3767 default (and expected) setting.
3769 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3770 message, again when deliver_drop_privilege is set or Exim is run as root.
3771 The bug described in (1) was not present in this case, but the tidying up
3772 of the other signals was missing. I have made the two cases consistent.
3774 3. The ignore_target_hosts setting on a manualroute router was being ignored
3775 for hosts that were looked up using the /MX notation.
3777 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3780 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3781 operated on the sender address. After changing the $sender_address to <>
3782 for the sender address verify, Exim was re-instated it as the original
3783 (before rewriting) address, but remembering that it had rewritten it, so it
3784 wasn't rewriting it again. This bug also had the effect of breaking the
3785 sender address verification caching when the sender address was rewritten.
3787 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3788 This has been changed so that if the ip literal address matches
3789 ignore_target_hosts, the router declines.
3791 7. Added expansion conditions match_domain, match_address, and match_local_
3792 part (NOT match_host).
3794 8. The placeholder for the Received: header didn't have a length field set.
3796 9. Added code to Exim itself and to exim_lock to test for a specific race
3797 condition that could lead to file corruption when using MBX delivery. The
3798 issue is with the lockfile that is created in /tmp. If this file is removed
3799 after a process has opened it but before that process has acquired a lock,
3800 there is the potential for a second process to recreate the file and also
3801 acquire a lock. This could lead to two Exim processes writing to the file
3802 at the same time. The added code performs the same test as UW imapd; it
3803 checks after acquiring the lock that its file descriptor still refers to
3804 the same named file.
3806 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3807 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3810 11. Added the smtp_active_hostname option. If used, this will typically be made
3811 to depend on the incoming interface address. Because $interface_address is
3812 not set up until the daemon has forked a reception process, error responses
3813 that can happen earlier (such as "too many connections") no longer contain
3816 12. If an expansion in a condition on a "warn" statement fails because a lookup
3817 defers, the "warn" statement is abandoned, and the next ACL statement is
3818 processed. Previously this caused the whole ACL to be aborted.
3820 13. Added the iplsearch lookup type.
3822 14. Added ident_timeout as a log selector.
3824 15. Added tls_certificate_verified as a log selector.
3826 16. Added a global option tls_require_ciphers (compare the smtp transport
3827 option of the same name). This controls incoming TLS connections.
3829 17. I finally figured out how to make tls_require_ciphers do a similar thing
3830 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3831 before starting the TLS session.
3833 18. Tabs are now shown as \t in -bP output.
3835 19. If the log selector return_path_on_delivery was set, Exim crashed when
3836 bouncing a message because it had too many Received: header lines.
3838 20. If two routers both had headers_remove settings, and the first one included
3839 a superfluous trailing colon, the final name in the first list and the
3840 first name in the second list were incorrectly joined into one item (with a
3841 colon in the middle).
3847 1. Added -C and -D options to the exinext utility, mainly to make it easier
3848 to include in the automated testing, but these could be helpful when
3849 multiple configurations are in use.
3851 2. The exinext utility was not formatting the output nicely when there was
3852 an alternate port involved in the retry record key, nor when there was a
3853 message id as well (for retries that were specific to a specific message
3854 and a specific host). It was also confused by IPv6 addresses, because of
3855 the additional colons they contain. I have fixed the IPv4 problem, and
3856 patched it up to do a reasonable job for IPv6.
3858 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3859 delivery, the log line now contains "pipelined" if PIPELINING was used.
3861 4. An SMTP transport process used to panic and die if the bind() call to set
3862 an explicit outgoing interface failed. This has been changed; it is now
3863 treated in the same way as a connect() failure.
3865 5. A reference to $sender_host_name in the part of a conditional expansion
3866 that was being skipped was still causing a DNS lookup. This no longer
3869 6. The def: expansion condition was not recognizing references to header lines
3870 that used bh_ and bheader_.
3872 7. Added the _cache feature to named lists.
3874 8. The code for checking quota_filecount in the appendfile transport was
3875 allowing one more file than it should have been.
3877 9. For compatibility with Sendmail, the command line option
3885 and sets the incoming protocol and host name (for trusted callers). The
3886 host name and its colon can be omitted when only the protocol is to be set.
3887 Note the Exim already has two private options, -pd and -ps, that refer to
3888 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3889 "s", but I don't think that's a major issue.
3891 10. A number of refactoring changes to the code, none of which should affect
3894 (a) The number of logging options was getting close to filling up the
3895 32-bit word that was used as a bit map. I have split them into two classes:
3896 those that are passed in the argument to log_write(), and those that are
3897 only ever tested independently outside of that function. These are now in
3898 separate 32-bit words, so there is plenty of room for expansion again.
3899 There is no change in the user interface or the logging behaviour.
3901 (b) When building, for example, log lines, the code previously used a
3902 macro that called string_cat() twice, in order to add two strings. This is
3903 not really sufficiently general. Furthermore, there was one instance where
3904 it was actually wrong because one of the argument was used twice, and in
3905 one call a function was used. (As it happened, calling the function twice
3906 did not affect the overall behaviour.) The macro has been replaced by a
3907 function that can join an arbitrary number of extra strings onto a growing
3910 (c) The code for expansion conditions now uses a table and a binary chop
3911 instead of a serial search (which was left over from when there were very
3912 few conditions). Also, it now recognizes conditions like "pam" even when
3913 the relevant support is not compiled in: a suitably worded error message is
3914 given if an attempt is made to use such a condition.
3916 11. Added ${time_interval:xxxxx}.
3918 12. A bug was causing one of the ddress fields not to be passed back correctly
3919 from remote delivery subprocesses. The field in question was not being
3920 subsequently used, so this caused to problems in practice.
3922 13. Added new log selectors queue_time and deliver_time.
3924 14. Might have fixed a bug in maildirsizefile handling that threw up
3925 "unexpected character" debug warnings, and recalculated the data
3926 unnecessarily. In any case, I expanded the warning message to give more
3929 15. Added the message "Restricted characters in address" to the statements in
3930 the default ACL that block characters like @ and % in local parts.
3932 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3933 Three changes have been made:
3935 (a) There was a serious bug; a negative response to MAIL caused the whole
3936 recipient domain to be cached as invalid, thereby blocking all messages
3937 to all local parts at the same domain, from all senders. This bug has
3938 been fixed. The domain is no longer cached after a negative response to
3939 MAIL if the sender used is not empty.
3941 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3944 (c) A new callout option, "use_sender" has been added for people who want
3945 the modified behaviour.
3951 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3954 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3955 indeed breaks things for older releases.
3957 3. Added additional logging to the case where there is a problem reading data
3958 from a filter that is running in a subprocess using a pipe, in order to
3959 try to track down a specific problem.
3961 4. Testing facility fudge: when running in the test harness and attempting
3962 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3963 getting "No route to host". Convert this to a timeout.
3965 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3968 6. Some OS don't have socklen_t but use size_t instead. This affects the
3969 fifth argument of getsockopt() amongst other things. This is now
3970 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3971 can be set for individual OS. I have set it for SunOS5, OSF1, and
3972 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3973 some earlier ones do not.
3975 7. Change 4.30/15 was not doing the test caselessly.
3977 8. The standard form for an IPv6 address literal was being rejected by address
3978 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3979 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3980 this, as well as the form without the "IPv6" on the front (but only when
3981 address literals are enabled, of course).
3983 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3985 10. Exim crashed if a message with an empty sender address specified by -f
3986 encountered a router with an errors_to setting. This could be provoked only
3987 by a command such as
3991 where an empty string was supplied; "<>" did not hit this bug.
3993 11. Installed PCRE release 4.5.
3995 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3996 remained set. It is now erased.
3998 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3999 times from message ids (which are base 36 rather than the normal 62).
4001 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4002 were being counted as actual protocol errors, and logged if the log
4003 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4004 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4005 and DATA following a set of rejected RCPTs do not count as protocol errors.
4006 In other words, Exim assumes they were pipelined, though this may not
4007 actually be the case. Of course, in all cases the client gets an
4008 appropriate error code.
4010 15. If a lookup fails in an ACL condition, a message about the failure may
4011 be available; it is used if testing the ACL cannot continue, because most
4012 such messages specify what the cause of the deferral is. However, some
4013 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4014 that caused an old message to be retained and used if a later statement
4015 caused a defer, replacing the real cause of the deferral.
4017 16. If an IP address had so many PTR records that the DNS lookup buffer
4018 was not large enough to hold them, Exim could crash while trying to process
4019 the truncated data. It now detects and logs this case.
4021 17. Further to 4.21/58, another change has been made: if (and only if) the
4022 first line of a message (the first header line) ends with CRLF, a bare LF
4023 in a subsequent header line has a space inserted after it, so as not to
4024 terminate the header.
4026 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4027 unnecessarily, used atoi() instead of strtol(), and didn't check the
4028 termination when getting file sizes from file names by regex.
4030 19. Completely re-implemented the support for maildirsize files, in the light
4031 of a number of problems with the previous contributed implementation
4032 (4.30/29). In particular:
4034 . If the quota is zero, the maildirsize file is maintained, but no quota is
4037 . If the maildir directory does not exist, it is created before any attempt
4038 to write a maildirsize file.
4040 . The quota value in the file is just a cache; if the quota is changed in
4041 the transport, the new value overrides.
4043 . A regular expression is available for excluding directories from the
4046 20. The autoreply transport checks the characters in options that define the
4047 message's headers; it allows continued headers, but it was checking with
4048 isspace() after an embedded newline instead of explicitly looking for a
4051 21. If all the "regular" hosts to which an address was routed had passed their
4052 expiry times, and had not reached their retry times, the address was
4053 bounced, even if fallback hosts were defined. Now Exim should go on to try
4056 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4057 equivalent code in the SMTP transport. Some hosts send humungous responses
4058 to HELO/EHLO, more than 1024 it seems.
4060 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4061 gives compiler warnings in some environments. I've now done it "properly",
4064 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4065 (because of problems with the built-in one) was declared to return uschar *
4066 instead of char *, causing compiler failure.
4068 25. Fixed a file descriptor leak when processing alias/forward files.
4070 26. Fixed a minor format string issue in dbfn.c.
4072 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4074 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4075 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4078 29. When a sender address is verified, it is cached, to save repeating the test
4079 when there is more than one recipient in a message. However, when the
4080 verification involves a callout, it is possible for different callout
4081 options to be set for different recipients. It is too complicated to keep
4082 track of this in the cache, so now Exim always runs a verification when a
4083 callout is required, relying on the callout cache for the optimization.
4084 The overhead is duplication of the address routing, but this should not be
4087 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4088 to be verified with callout=postmaster, and the main callout worked but the
4089 postmaster check failed, the verification correctly failed. However, if a
4090 subsequent RCPT command asked for sender verification *without* the
4091 postmaster check, incorrect caching caused this verification also to fail,
4094 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4095 it was not caching the DNS options (qualify_single, search_parents) that
4096 were used when the lookup failed. A subsequent lookup with different
4097 options therefore always gave the same answer, though there were cases
4098 where it should not have. (Example: a "domains = !$mx_any" option on a
4099 dnslookup router: the "domains" option is always processed without any
4100 widening, but the router might have qualify_single set.) Now Exim uses the
4101 cached value only when the same options are set.
4103 32. Added John Jetmore's "exipick" utility to the distribution.
4105 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4106 than a timeout (e.g. a certificate is required, and is not provided), an
4107 Exim server now closes the connection immediately. Previously it waited for
4108 the client to close - but if the client is SSL, it seems that they each
4109 wait for each other, leading to a delay before one of them times out.
4111 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4112 maintained 0.8.x compatibility because I don't think many are using it, and
4113 it is clearly obsolete.
4115 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4118 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4119 client certificate was expired. A simple patch fixes this, though I don't
4120 understand the full logic of why the verify callback is called multiple
4123 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4124 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4125 which causes problems with some clients (such as the Certicom SSL Plus
4126 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4127 disables the coutermeasure allowing Eudora to connect."
4129 38. Exim was not checking that a write() to a log file succeeded. This could
4130 lead to Bad Things if a log got too big, in particular if it hit a file
4131 size limit. Exim now panics and dies if it cannot write to a log file, just
4132 as it does if it cannot open a log file.
4134 39. Modified OS/Makefile-Linux so that it now contains
4136 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4138 The two -D definitions ensure that Exim is compiled with large file
4139 support, which makes it possible to handle log files that are bigger than
4142 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4143 instance) a domain was checked against a named list that involved a lookup,
4144 causing $domain_data to be set, then another domain was checked against the
4145 same list, then the first domain was re-checked, the value of $domain_data
4146 after the final check could be wrong. In particular, if the second check
4147 failed, it could be set empty. This bug probably also applied to
4150 41. The strip_trailing_dot option was not being applied to the address given
4151 with the -f command-line option.
4153 42. The code for reading a message's header from the spool was incrementing
4154 $received_count, but never initializing it. This meant that the value was
4155 incorrect (doubled) while delivering a message in the same process in which
4156 it was received. In the most common configuration of Exim, this never
4157 happens - a fresh exec is done - but it can happen when
4158 deliver_drop_privilege is set.
4160 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4161 it now includes up to 150 characters of the unexpected data in the log
4164 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4165 and building data strings. The size of both of these buffers was 10 000
4166 bytes - far larger than anybody would *ever* want, thought I. Needless to
4167 say, somebody hit the limit. I have increased the maximum line length to
4168 20 000 and the maximum data length of concatenated lines to 100 000. I have
4169 also fixed two bugs, because there was no checking on these buffers. Tsk,
4170 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4171 buffer is too small.
4173 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4174 lsearch lookups. Now it does.
4176 46. When parsing a route_list item in a manualroute router, a fixed-length
4177 buffer was used for the list of hosts. I made this 1024 bytes long,
4178 thinking that nobody would ever have a list of hosts that long. Wrong.
4179 Somebody had a whole pile of complicated expansion conditions, and the
4180 string was silently truncated, leading to an expansion error. It turns out
4181 that it is easier to change to an unlimited length (owing to other changes
4182 that have happened since this code was originally written) than to build
4183 structure for giving a limitation error. The length of the item that
4184 expands into the list of hosts is now unlimited.
4186 47. The lsearch lookup could not handle data where the length of text line was
4187 more than 4095 characters. Such lines were truncated, leading to shortened
4188 data being returned. It should now handle lines of any length.
4190 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4191 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4194 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4195 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4196 no longer happens in comment lines. A list of replacements is now placed
4197 at the head of all of the source files, except those whose only change is
4198 to replace PERL_COMMAND in the very first #! line.
4200 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4201 messages on the queue, with a bottom-up merge sort, using code contributed
4202 by Michael Haardt. This should make operations like -bp somewhat faster on
4203 large queues. It won't affect queue runners, except when queue_run_in_order
4206 51. Installed eximstats 1.31 in the distribution.
4208 52. Added support for SRV lookups to the dnslookup router.
4210 53. If an ACL referred to $message_body or $message_body_end, the value was not
4211 reset for any messages that followed in the same SMTP session.
4213 54. The store-handling optimization for building very long strings was not
4214 differentiating between the different store pools. I don't think this
4215 actually made any difference in practice, but I've tidied it.
4217 55. While running the routers to verify a sender address, $sender_address
4218 was still set to the sender address. This is wrong, because when routing to
4219 send a bounce to the sender, it would be empty. Therefore, I have changed
4220 it so that, while verifying a sender address, $sender_address is set to <>.
4221 (There is no change to what happens when verifying a recipient address.)
4223 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4224 target A or AAAA records (if not already returned) without resetting the
4225 qualify_single or search_parents options of the DNS resolver. These are
4226 inappropriate in this case because the targets of MX and SRV records must
4227 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4228 target that, when qualified, matched something in the local domain. These
4229 two options are now turned off when doing these lookups.
4231 57. It seems that at least some releases of Reiserfs (which does not have the
4232 concept of a fixed number of inodes) returns zero and not -1 for the
4233 number of available inodes. This interacted badly with check_spool_inodes,
4234 which assumed that -1 was the "no such thing" setting. What I have done is
4235 to check that the total number of inodes is greater than zero before doing
4236 the test of how many are available.
4238 58. When a "warn" ACL statement has a log_message modifier, the message is
4239 remembered, and not repeated. This is to avoid a lot of repetition when a
4240 message has many recipients that cause the same warning to be written.
4241 Howewer, Exim was preserving the list of already written lines for an
4242 entire SMTP session, which doesn't seem right. The memory is now reset if a
4243 new message is started.
4245 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4246 debugging output unless log_rewrite was also set.
4248 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4249 of (int)(handle) when we know that handle contains (void *)(-1).
4251 61. The Exim daemon panic-logs an error return when it closes the incoming
4252 connection. However "connection reset by peer" seems to be common, and
4253 isn't really an error worthy of noting specially, so that particular error
4256 62. When Exim is trying to find all the local interfaces, it used to panic and
4257 die if the ioctl to get the interface flags failed. However, it seems that
4258 on at least one OS (Solaris 9) it is possible to have an interface that is
4259 included in the list of interfaces, but for which you get a failure error
4260 for this call. This happens when the interface is not "plumbed" into a
4261 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4262 failure of the "get flags" call assumes that the interface is down.
4264 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4265 makes life easier for people who are doing arithmetic on fields extracted
4266 from dates, where you often get leading zeros that should not be
4267 interpreted as octal.
4269 64. Added qualify_domain to the redirect router, to override the global
4272 65. If a pathologically long header line contained very many addresses (the
4273 report of this problem mentioned 10 000) and each of them was rewritten,
4274 Exim could use up a very large amount of memory. (It kept on making new
4275 copies of the header line as it rewrote, and never released the old ones.)
4276 At the expense of a bit more processing, the header rewriting function has
4277 been changed so that it no longer eats memory in this way.
4279 66. The generation of the Received: header has been moved from the time that a
4280 message starts to be received, to the time that it finishes. The timestamp
4281 in the Received: header should now be very close to that of the <= log
4282 line. There are two side-effects of this change:
4284 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4285 logged header lines no longer include the local Received: line, because
4286 it has not yet been created. The same applies to a copy of the message
4287 that is returned to a non-SMTP sender when a message is rejected.
4289 (b) When a filter file is tested using -bf, no additional Received: header
4290 is added to the test message. After some thought, I decided that this
4293 This change does not affect the value of $received_for. It is still set
4294 after address rewriting, but before local_scan() is called.
4296 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4298 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4299 gave an unhelpful panic error message, and a defer error. I have managed to
4300 change this behaviour so that it now rejects any supplied certificate,
4301 which seems right, as the list of acceptable certificates is empty.
4303 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4304 gave an unhelpful defer error. I have not managed to make this reject any
4305 supplied certificates, but the error message it gives is "no certificate
4306 supplied", which is not helpful.
4308 70. exigrep's output now also includes lines that are not associated with any
4309 message, but which match the given pattern. Implemented by a patch from
4310 Martin Sluka, which also tidied up the Perl a bit.
4312 71. Recipient callout verification, like sender verification, was using <> in
4313 the MAIL FROM command. This isn't really the right thing, since the actual
4314 sender may affect whether the remote host accepts the recipient or not. I
4315 have changed it to use the actual sender in the callout; this means that
4316 the cache record is now keyed on a recipient/sender pair, not just the
4317 recipient address. There doesn't seem to be a real danger of callout loops,
4318 since a callout by the remote host to check the sender would use <>.
4319 [SEE ABOVE: changed after hitting problems.]
4321 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4322 temporary errors. However, in the case of such a code being given after
4323 the end of a data transmission (i.e. after ".") Exim was failing to write
4324 a retry record for the message. (Yes, there was some broken host that was
4325 actually sending 8xx at this point.)
4327 73. An unknown lookup type in a host list could cause Exim to panic-die when
4328 the list was checked. (An example that provoked this was putting <; in the
4329 middle of a list instead of at the start.) If this happened during a DATA
4330 ACL check, a -D file could be left lying around. This kind of configuration
4331 error no longer causes Exim to die; instead it causes a defer errror. The
4332 incident is still logged to the main and panic logs.
4334 74. Buglet left over from Exim 3 conversion. The message "too many messages
4335 in one connection" was written to the rejectlog but not the mainlog, except
4336 when address rewriting (yes!) was being logged.
4338 75. Added write_rejectlog option.
4340 76. When a system filter was run not as root (that is, when system_filter_user
4341 was set), the values of the $n variables were not being returned to the
4342 main process; thus, they were not subsequently available in the $sn
4345 77. Added +return_path_on_delivery log selector.
4347 78. A connection timeout was being treated differently from recipients deferred
4348 when testing hosts_max_try with a message that was older than the host's
4349 retry timeout. (The host should not be counted, thus allowing all hosts to
4350 be tried at least once before bouncing.) This may have been the cause of an
4351 occasionally reported bug whereby a message would remain on the queue
4352 longer than the retry timeout, but would be bounced if a delivery was
4353 forced. I say "may" because I never totally pinned down the problem;
4354 setting up timeout/retry tests is difficult. See also the next item.
4356 79. The ultimate address timeout was not being applied to errors that involved
4357 a combination of host plus message (for example, a timeout on a MAIL
4358 command). When an address resolved to a number of possible hosts, and they
4359 were not all tried for each delivery (e.g. because of hosts_max_try), a
4360 message could remain on the queue longer than the retry timeout.
4362 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4365 81. Fixed an obscure SMTP outgoing bug which required at least the following
4366 conditions: (a) there was another message waiting for the same server;
4367 (b) the server returned 5xx to all RCPT commands in the first message so
4368 that the message was not completed; (c) the server dropped the connection
4369 or gave a negative response to the RSET that Exim sends to abort the
4370 transaction. The observed case was a dropped connection after DATA that had
4371 been sent in pipelining mode. That is, the server had advertised PIPELINING
4372 but was not implementing it correctly. The effect of the bug was incorrect
4373 behaviour, such as trying another host, and this could lead to a crash.
4379 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4380 and daemon.c were passed as pointers to ints; they should have been
4381 pointers to socklen_t variables (which are typically unsigned ints).
4383 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4386 3. Fixed a really odd bug that affected only the testing scheme; patching a
4387 certain fixed string in the binary changed the value of another string that
4388 happened to be identical to the end of the original first string.
4390 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4391 name", it returns that address as the IP address. On some operating
4392 systems (e.g. Solaris), it also passes back the IP address string as the
4393 "host name". However, on others (e.g. Linux), it passes back an empty
4394 string. Exim wasn't checking for this, and was changing the host name to an
4395 empty string, assuming it had been canonicized.
4397 5. Although rare, it is permitted to have more than one PTR record for a given
4398 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4399 all the names associated with an address, because they do in Solaris.
4400 However, it seems that they do not in Linux for data that comes from the
4401 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4402 I found this out when I moved to a new Linux workstation and tried to run
4403 the Exim test suite.
4405 To get round this problem I have changed the code so that it now does its
4406 own call to the DNS to look up PTR records when searching for a host name.
4407 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4408 addresses that are only in /etc/hosts are still found.
4410 This behaviour is, however, controlled by an option called host_lookup_
4411 order, which defaults to "bydns:byaddr". If people want to use the other
4412 order, or indeed, just use one or the other means of lookup, they can
4413 specify it in this variable.
4415 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4416 some operating systems, this comes back from gethostbyaddr() as an empty
4417 string, and this is what Exim used to test for. However, it seems that in
4418 other systems, "." is yielded. Exim now tests for this case too.
4420 7. The values of check_spool_space and check_log_space are now held internally
4421 as a number of kilobytes instead of an absolute number of bytes. If a
4422 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4423 kilobyte. This means that much larger values can be stored.
4425 8. Exim monitor: an attempt to get the action menu when not actually pointing
4426 at a message produces an empty menu entitled "No message selected". This
4427 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4428 no entries in it ("Shell widget menu has zero width and/or height"). So I
4429 have added a single, blank menu entry in this case.
4431 9. Added ${quote_local_part.
4433 10. MIME decoding is now applied to the contents of Subject: header lines when
4436 11. Now that a reference to $sender_host_address automatically causes a reverse
4437 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4438 host lookup before query-style lookups in lists that might use this
4439 variable. This has therefore been abolished, and the "net-" prefix is no
4440 longer necessary for query-style lookups.
4442 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4443 have been a typo for LFLAGS, so it has been changed.
4445 13. The install script calls Exim with "-C /dev/null" in order to find the
4446 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4447 to be output. Howeve, since Exim outputs its version number before the
4448 error, it didn't break the script. It just looked ugly. I fixed this by
4449 always allowing "-C /dev/null" if the caller is root.
4451 14. Ignore overlarge ACL variable number when reading spool file - insurance
4452 against a later release with more variables having written the file.
4454 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4455 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4456 this, as well as the form without the "IPv6" on the front.
4458 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4459 OS/Makefile-Darwin file.
4461 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4462 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4464 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4465 variable. However, some OS use ints for inodes. I've added cast to long int
4466 to get rid of the compiler warning.
4468 19. I had forgotten to lock out "/../" in configuration file names when
4469 ALT_CONFIG_PREFIX was set.
4471 20. Routers used for verification do not need to specify transports. However,
4472 if such a router generated a host list, and callout was configured, Exim
4473 crashed, because it could not find a port number from the (non-existent)
4474 transport. It now assumes port 25 in this circumstance.
4476 21. Added the -t option to exigrep.
4478 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4479 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4480 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4481 them is set, LOOKUP_LSEARCH is forced.
4483 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4485 24. Added sender and host information to the "rejected by local_scan()" log
4486 line; previously there was no indication of these.
4488 25. Added .include_if_exists.
4490 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4491 Linux. It turns out that not all file systems support this. Apparently some
4492 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4493 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4494 not supported on the file descriptor, is now ignored when Exim is trying to
4495 sync a directory. This applies only to Linux.
4497 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4499 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4500 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4503 29. Added support for maildirsize files from supplied patch (modified a bit).
4505 30. The use of :fail: followed by an empty string could lead Exim to respond to
4506 sender verification failures with (e.g.):
4508 550 Verification failed for <xxx>
4509 550 Sender verify failed
4511 where the first response line was missing the '-' that indicates it is not
4512 the final line of the response.
4514 31. The loop for finding the name of the user that called Exim had a hardwired
4515 limit of 10; it now uses the value of finduser_retries, which is used for
4516 all other user lookups.
4518 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4521 33. Exim was neglecting to zero errno before one call of strtol() when
4522 expanding a string and expecting an integer value. On some systems this
4523 resulted in spurious "integer overflow" errors. Also, it was casting the
4524 result into an int without checking.
4526 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4527 did not work. The code looks as if it has *never* worked, though it appears
4528 to have been documented since at least releast 1.62. I have made it work.
4530 35. The "timeout_DNS" error in retry rules, also documented since at least
4531 1.62, also never worked. As it isn't clear exactly what this means, and
4532 clearly it isn't a major issue, I have abolished the feature by treating it
4533 as "timeout", and writing a warning to the main and panic logs.
4535 36. The display of retry rules for -brt wasn't always showing the error code
4538 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4539 timeout_connect_A, timeout_connect_MX.
4541 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4542 to the empty sender.
4544 39. The daemon was not analysing the content of -oX till after it had closed
4545 stderr and disconnected from the controlling terminal. This meant that any
4546 syntax errors were only noted on the panic log, and the return code from
4547 the command was 0. By re-arranging the code a little, I've made the
4548 decoding happen first, so such errors now appear on stderr, and the return
4549 code is 1. However, the actual setting up of the sockets still happens in
4550 the disconnected process, so errors there are still only recorded on the
4553 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4554 connections (as happens on some IP stacks) was logged at start up time as
4555 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4556 it from "IPv6 and IPv4", which means that two separate sockets are being
4559 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4560 says whether AF_INET or AF_INET6 was passed as an argument.
4562 42. Exiwhat output was messed up when time zones were included in log
4565 43. Exiwhat now gives more information about the daemon's listening ports,
4566 and whether -tls-on-connect was used.
4568 44. The "port" option of the smtp transport is now expanded.
4570 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4571 silently ignored. Now an error message is written to the main and panic
4574 46. There's a new ACL modifier called "logwrite" which writes to a log file
4575 as soon as it is encountered.
4577 47. Added $local_user_uid and $local_user_gid at routing time.
4579 48. Exim crashed when trying to verify a sender address that was being
4582 49. Exim was recognizing only a space character after ".include". It now also
4583 recognizes a tab character.
4585 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4586 extracting the relevant information from the specification. The man page no
4587 longer contains scrambled data for the -d option, and I've added a section
4588 at the front about calling Exim under different names.
4590 51. Added "extra_headers" argument to the "mail" command in filter files.
4592 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4595 53. Installed eximstats 1.29.
4597 54. Added transport_filter_timeout as a generic transport option.
4599 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4600 Cc: header lines. This was required by RFC 822, but it not required by RFC
4603 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4604 incoming messages that did not have them. Now it does so only if the
4605 message originates locally, that is, if there is no associated remote host
4606 address. When Resent- header lines are present, this applies to the Resent-
4607 lines rather than the non-Resent- lines.
4609 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4610 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4612 58. Messages for configuration errors now include the name of the main
4613 configuration file - useful now that there may be more than one file in a
4614 list (.included file names were always shown).
4616 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4617 for those rare installations that do not start the daemon as root or run it
4618 setuid root. I've cut out the call to initgroups() if the daemon is not
4621 60. The Exim user and group can now be bound into the binary as text strings
4622 that are looked up at the start of Exim's processing.
4624 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4626 62. Added $mailstore_basename variable.
4628 63. Installed patch to sieve.c from Michael Haardt.
4630 64. When Exim failed to open the panic log after failing to open the main log,
4631 the original message it was trying to log was written to stderr and debug
4632 output, but if they were not available (the usual case in production), it
4633 was lost. Now it is written to syslog before the two lines that record the
4634 failures to open the logs.
4636 65. Users' Exim filters run in subprocesses under the user's uid. It is
4637 possible for a "deliver" command or an alias in a "personal" command to
4638 provoke an address rewrite. If logging of address rewriting is configured,
4639 this fails because the process is not running as root or exim. There may be
4640 a better way of dealing with this, but for the moment (because 4.30 needs
4641 to be released), I have disabled address rewrite logging when running a
4642 filter in a non-root, non-exim process.
4648 1. The buildconfig auxiliary program wasn't quoting the value set for
4649 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4650 not defined. This bug was masked in 4.22 by the effect that was fixed in
4653 2. Some messages that were rejected after a message id was allocated were
4654 shown as "incomplete" by exigrep. It no longer does this for messages that
4655 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4657 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4658 have allow_domain_literals set, the ID did not get logged in the <= line.
4659 Domain literals are now always recognized in Message-ID: header lines.
4661 4. The first argument for a ${extract expansion item is the key name or field
4662 number. Leading and trailing spaces in this item were not being ignored,
4663 causing some misleading effects.
4665 5. When deliver_drop_privilege was set, single queue runner processes started
4666 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4667 same command in the process it spins off) were not dropping privilege.
4669 6. When the daemon running as "exim" started a queue runner, it always
4670 re-executed Exim in the spun-off process. This is a waste of effort when
4671 deliver_drop_privilege is set. The new process now just calls the
4672 queue-runner function directly.
4678 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4681 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4682 ignored. Though the use of -oP was forcing the writing of a pid file, it
4683 was always written to the default place.
4685 3. If the message "no IP address found for host xxxx" is generated during
4686 incoming verification, it is now followed by identification of the incoming
4687 connection (so you can more easily find what provoked it).
4689 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4691 5. Added some features to "harden" Exim a bit more against certain attacks:
4693 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4694 be put in Local/Makefile. This is like the never_users runtime option,
4695 but it cannot be overridden. The default setting is "root".
4697 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4698 prefix string with which any file named in a -C command line option
4701 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4702 is retained for -C and -D only if the caller of Exim is root. Without
4703 it, the exim user may also use -C and -D and retain privilege.
4705 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4706 command line option is disabled.
4708 6. Macro names set by the -D option must start with an upper case letter, just
4709 like macro names defined in the configuration file.
4711 7. Added "dereference=" facility to LDAP.
4713 8. Two instances of the typo "uknown" in the source files are fixed.
4715 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4716 the Configure-Makefile script screwed up while processing it.
4718 10. Incorporated PCRE 4.4.
4720 11. The SMTP synchronization check was not operating right at the start of an
4721 SMTP session. For example, it could not catch a HELO sent before the client
4722 waited for the greeting. There is now a check for outstanding input at the
4723 point when the greeting is written. Because of the duplex, asynchronous
4724 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4725 way, but not yet received, when the check is performed.
4727 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4728 on TCP/IP sockets, because this apparently causes some broken clients to
4731 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4732 unchanged) from the Cygwin maintainer.
4734 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4735 was supported instead of testing for mbx. Effectively a typo.
4737 15. The spa authenticator server code was not checking that the input it
4738 received was valid base64.
4740 16. The debug output line for the "set" modifier in ACLs was not showing the
4741 name of the variable that was being set.
4743 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4745 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4746 lookup on its own. Something else was needed to trigger the lookup. For
4747 example, a match in host_lookup or the need for a host name in a host list.
4748 Now, if $sender_host_name is referenced and the host name has not yet been
4749 looked up, a lookup is performed. If the lookup fails, the variable remains
4750 empty, and $host_lookup_failed is set to "1".
4752 19. Added "eqi" as a case-independent comparison operator.
4754 20. The saslauthd authentication condition could segfault if neither service
4755 nor realm was specified.
4757 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4758 error message that was logged was misleading, and incoming SMTP
4759 connections were dropped. The message is now more accurate, and temporary
4760 errors are given to SMTP connections.
4762 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4763 (except RSET) with a 421 error, until QUIT is received. However, it was
4764 failing to send a response to QUIT.
4766 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4767 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4768 ACL is now run after the helo_try_verify_hosts code.
4770 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4771 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4772 case-independent, so other case variants are also recognized). Apparently
4773 some systems use these upper case variants.
4775 25. If more than two messages were waiting for the same host, and a transport
4776 filter was specified for the transport, Exim sent two messages over the
4777 same TCP/IP connection, and then failed with "socket operation on non-
4778 socket" when it tried to send the third.
4780 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4782 27. The extern definition of crypt16() in expand.c was not being excluded when
4783 the OS had its own crypt16() function.
4785 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4786 as a preferred synonym for return_size_limit, both as an option and as an
4789 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4791 30. Changed the default configuration ACL to relax the local part checking rule
4792 for addresses that are not in any local domains. For these addresses,
4793 slashes and pipe symbols are allowed within local parts, but the sequence
4794 /../ is explicitly forbidden.
4796 31. SPA server authentication was not clearing the challenge buffer before
4799 32. log_message in a "warn" ACL statement was writing to the reject log as
4800 well as to the main log, which contradicts the documentation and doesn't
4801 seem right (because no rejection is happening). So I have stopped it.
4803 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4804 However, I am unable to do any testing of this.
4806 34. Fixed an infelicity in the appendfile transport. When checking directories
4807 for a mailbox, to see if any needed to be created, it was accidentally
4808 using path names with one or more superfluous leading slashes; tracing
4809 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4811 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4812 contents are added to the log line that is written for every discarded
4813 recipient. (Previously a log_message setting was ignored.)
4815 36. The ${quote: operator now quotes the string if it is empty.
4817 37. The install script runs exim in order to find its version number. If for
4818 some reason other than non-existence or emptiness, which it checks, it
4819 could not run './exim', it was installing it with an empty version number,
4820 i.e. as "exim-". This error state is now caught, and the installation is
4823 38. An argument was missing from the function that creates an error message
4824 when Exim fails to connect to the socket for saslauthd authentication.
4825 This could cause Exim to crash, or give a corrupted message.
4827 39. Added isip, isip4, and isip6 to ${if conditions.
4829 40. The ACL variables $acl_xx are now saved with the message, and can be
4830 accessed later in routers, transports, and filters.
4832 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4833 strings in the file are not string-expanded.
4835 42. If a MAIL command specified a SIZE value that was too large to fit into an
4836 int variable, the check against message_size_limit failed. Such values are
4837 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4838 day this will have to be increased, but I don't think I want to be around
4839 when emails are that large.
4846 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4847 iconv() is not standard in FreeBSD.
4849 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4850 IPv6 enabled. The observed symptom was a segmentation fault on return from
4851 the function os_common_find_running_interfaces() in src/os.c.
4853 3. In the check_special_case() function in daemon.c I had used "errno" as an
4854 argument name, which causes warnings on some systems. This was basically a
4855 typo, since it was named "eno" in the comments!
4857 4. The code that waits for the clock to tick (at a resolution of some fraction
4858 of a second) so as to ensure message-id uniqueness was always waiting for
4859 at least one whole tick, when it could have waited for less. [This is
4860 almost certainly not relevant at current processor speeds, where it is
4861 unlikely to ever wait at all. But we try to future-proof.]
4863 5. The function that sleeps for a time interval that includes fractions of a
4864 second contained a race. It did not block SIGALRM between setting the
4865 timer, and suspending (a couple of lines later). If the interval was short
4866 and the sigsuspend() was delayed until after it had expired, the suspension
4867 never ended. On busy systems this could lead to processes getting stuck for
4870 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4871 process, before it forks any delivery processes. The open lookup caching
4872 mechanism meant that the open file or database connection was passed into
4873 the delivery process. The problem was that delivery processes always tidy
4874 up cached lookup data. This could cause a problem for the next delivery
4875 process started by the queue runner, because the external queue runner
4876 process does not know about the closure. So the next delivery process
4877 still has data in the lookup cache. In the case of a file lookup, there was
4878 no problem because closing a file descriptor in a subprocess doesn't affect
4879 the parent. However, if the lookup was caching a connection to a database,
4880 the connection was closed, and the second delivery process was likely to
4881 see errors such as "PGSQL: query failed: server closed the connection
4882 unexpectedly". The problem has been fixed by closing all cached lookups
4883 in a queue runner before running a delivery process.
4885 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4886 seem to have the "const" qualifier which it has on other OS. I've
4889 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4890 *of the same type* (client or server) with the same public name that an
4891 error should be diagnosed.
4893 9. When Exim looked up a host name for an IP address, but failed to find the
4894 original IP address when looking up the host name (a safety check), it
4895 output the message "<ip address> does not match any IP for NULL", which was
4896 confusing, to say the least. The bug was that the host name should have
4897 appeared instead of "NULL".
4899 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4900 user that is built into the binary, and the -C or -D options is used, root
4901 privilege is dropped before the configuration file is read. In addition,
4902 logging is switched to stderr instead of the normal log files. If the
4903 configuration then re-defines the Exim user, the unprivileged environment
4904 is probably not what is expected, so Exim logs a panic warning message (but
4907 However, if deliver_drop_privilege is set, the unprivileged state may well
4908 be exactly what is intended, so the warning has been cut out in that case,
4909 and Exim is allowed to try to write to its normal log files.
4915 1. smtp_return_error_details was not giving details for temporary sender
4916 or receiver verification errors.
4918 2. Diagnose a configuration error if two authenticators have the same public
4921 3. Exim used not to create the message log file for a message until the first
4922 delivery attempt. This could be confusing when incoming messages were held
4923 for policy or load reasons. The message log file is now created at the time
4924 the message is received, and an initial "Received" line is written to it.
4926 4. The automatically generated man page for command line options had a minor
4927 bug that caused no ill effects; however, a more serious problem was that
4928 the procedure for building the man page automatically didn't always
4929 operate. Consequently, release 4.20 contains an out-of-date version. This
4930 shouldn't happen again.
4932 5. When building Exim with embedded Perl support, the script that builds the
4933 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4934 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4936 6. The freeze_tell option was not being used for messages that were frozen on
4937 arrival, either by an ACL or by local_scan().
4939 7. Added the smtp_incomplete_transaction log selector.
4941 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4942 was accepting AUTH without a new EHLO.
4944 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4945 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4948 10. Logging of TCP/IP connections (when configured) now happens in the main
4949 daemon process instead of the child process, so that the TCP/IP connection
4950 count is more accurate (but it can never be perfect).
4952 11. The use of "drop" in a nested ACL was not being handled correctly in the
4953 outer ACL. Now, if condition failure induced by the nested "drop" causes
4954 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4955 or "require"), the connection is dropped.
4957 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4958 that yield "discard" can now be used with an "accept" or a "discard" verb,
4959 but an error is generated for any others (because I can't see a useful way
4960 to define what should happen).
4962 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4963 are now processed in the same way as lines in the Exim configuration file.
4964 In particular, continuation lines are supported.
4966 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4968 15. Added -ti meaning -t -i.
4970 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4971 domains, and warn by logging if others are found.
4973 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4974 returned from GIFCONF when Exim is trying to find the list of interfaces on
4975 a host. The code in os.c has been modified to copy each ifreq to an aligned
4976 structure in all cases.
4978 Also, in some cases, the returned ifreq's were being copied to a 'struct
4979 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4980 means the last couple of bytes of an IPv6 address could be chopped if the
4981 ifreq contained only a normal sockaddr (14 bytes storage).
4983 18. Named domain lists were not supported in the hosts_treat_as_local option.
4984 An entry such as +xxxx was not recognized, and was treated as a literal
4987 19. Ensure that header lines added by a DATA ACL are included in the reject log
4988 if the ACL subsequently rejects the message.
4990 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4991 MD5 (which is deprecated).
4993 21. When testing a filter file using -bf, Exim was writing a message when it
4994 took the sender from a "From " line in the message, but it was not doing so
4995 when it took $return_path from a Return-Path: header line. It now does.
4997 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4998 with a valid header line field name (a series of printing characters
4999 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5001 23. Changed "disc" in the source to "disk" to conform to the documentation and
5002 the book and for uniformity.
5004 24. Ignore Sendmail's -Ooption=value command line item.
5006 25. When execve() failed while trying to run a command in a pipe transport,
5007 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5008 could be confused with a return value of 69 from the command itself. This
5009 has been changed to 127, the value the shell returns if it is asked to run
5010 a non-existent command. The wording for the related log line suggests a
5011 non-existent command as the problem.
5013 26. If received_header_text expands to an empty string, do not add a Received:
5014 header line to the message. (Well, it adds a token one on the spool, but
5015 marks it "old" so that it doesn't get used or transmitted.)
5017 27. Installed eximstats 1.28 (addition of -nt option).
5019 28. There was no check for failure on the call to getsockname() in the daemon
5020 code. This can fail if there is a shortage of resources on the system, with
5021 ENOMEM, for example. A temporary error is now given on failure.
5023 29. Contrary to the C standard, it seems that in some environments, the
5024 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5025 program. Exim now does this explicitly; it affects the formatting of
5026 timestamps using strftime().
5028 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5029 complaints. I've now initialized all the variables, to avoid this.
5031 32. Header lines added by a system filter were not being "seen" during
5032 transport-time rewrites.
5034 33. The info_callback() function passed to OpenSSL is set up with type void
5035 (*)(SSL *, int, int), as described somewhere. However, when calling the
5036 function (actually a macro) that sets it up, the type void(*)() is
5037 expected. I've put in a cast to prevent warnings from picky compilers.
5039 34. If a DNS black list lookup found a CNAME record, but there were no A
5040 records associated with the domain it pointed at, Exim crashed.
5042 35. If a DNS black list lookup returned more than one A record, Exim ignored
5043 all but the first. It now scans all returned addresses if a particular IP
5044 value is being sought. In this situation, the contents of the
5045 $dnslist_value variable are a list of all the addresses, separated by a
5048 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5049 to accept a host name and all its aliases if the forward lookup for any of
5050 them yielded the IP address of the incoming connection. Now it accepts only
5051 those names whose forward lookup yields the correct IP address. Any other
5052 names are discarded. This closes a loophole whereby a rogue DNS
5053 administrator could create reverse DNS records to break through a
5054 wildcarded host restriction in an ACL.
5056 37. If a user filter or a system filter that ran in a subprocess used any of
5057 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5058 the wrong values were passed to the pipe command ($thisaddress had the
5059 value of $0, $0 had the value of $1, etc). This bug was introduced by
5060 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5062 38. Improved the line breaking for long SMTP error messages from ACLs.
5063 Previously, if there was no break point between 40 and 75 characters, Exim
5064 left the rest of the message alone. Two changes have been made: (a) I've
5065 reduced the minimum length to 35 characters; (b) if it can't find a break
5066 point between 35 and 75 characters, it looks ahead and uses the first one
5067 that it finds. This may give the occasional overlong line, but at least the
5068 remaining text gets split now.
5070 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5071 file descriptors might be low, and that setting 1000 would always raise it.
5072 It turns out that in some environments, the limit is already over 1000 and
5073 that lowering it causes trouble. So now Exim takes care not to decrease it.
5075 40. When delivering a message, the value of $return_path is set to $sender_
5076 address at the start of routing (routers may change the value). By an
5077 oversight, this default was not being set up when an address was tested by
5078 -bt or -bv, which affected the outcome if any router or filter referred to
5081 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5082 log only when "message" or "log_message" are set. However, if one of the
5083 conditions was an address verification, or a call to a nested ACL, the
5084 messages generated by the underlying test were being passed through. This
5085 no longer happens. The underlying message is available in $acl_verify_
5086 message for both "message" and "log_message" expansions, so it can be
5087 passed through if needed.
5089 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5090 new expansion $bh_ to give the encoded byte string without charset
5091 translation. Translation happens only if iconv() is available; HAVE_ICONV
5092 indicates this at build time. HEADERS_CHARSET gives the charset to
5093 translate to; headers_charset can change it in the configuration, and
5094 "headers charset" can change it in an individual filter file.
5096 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5097 that creates RFC 2047 encoded "words" labels them as that charset instead
5098 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5099 expansion operator; (ii) when Exim creates a From: line for a local
5100 message; (iii) when a header line is rewritten to include a "phrase" part.
5102 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5103 buggy, causing it to skip the first lines of messages whose message ID
5104 ended in 'D'. This would not have bitten before Exim release 4.14, because
5105 message IDs were unlikely to end in 'D' before then. The effect was to have
5106 incorrect size information for certain domains.
5108 45. #include "config.h" was missing at the start of the crypt16.c module. This
5109 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5112 46. If there was a timeout during a "random" callout check, Exim treated it as
5113 a failure of the random address, and carried on sending RSET and the real
5114 address. If the delay was just some slowness somewhere, the response to the
5115 original RCPT would be taken as a response to RSET and so on, causing
5116 mayhem of various kinds.
5118 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5119 when I implemented it. It didn't allow for the fact that some option values
5120 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5121 the right test for positive values.
5123 48. Domain names in DNS records are case-independent. Exim always looks them up
5124 in lower case. Some resolvers return domain names in exactly the case they
5125 appear in the zone file, that is, they may contain uppercase letters. Not
5126 all resolvers do this - some return always lower case. Exim was treating a
5127 change of case by a resolver as a change of domain, similar to a widening
5128 of a domain abbreviation. This triggered its re-routing code and so it was
5129 trying to route what was effectively the same domain again. This normally
5130 caused routing to fail (because the router wouldn't handle the domain
5131 twice). Now Exim checks for this case specially, and just changes the
5132 casing of the domain that it ultimately uses when it transmits the message
5135 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5138 50. If a filter generated a file delivery with a non-absolute name (possible if
5139 no home directory exists for the router), the forbid_file option was not
5142 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5143 the existing equality matching.
5145 52. Exim was using ints instead of ino_t variables in some places where it was
5146 dealing with inode numbers.
5148 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5149 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5150 called TMPDIR, and if it finds it is different, it changes its value.
5152 54. The smtp_printf() function is now made available to local_scan() so
5153 additional output lines can be written before returning. There is also an
5154 smtp_fflush() function to enable the detection of a dropped connection.
5155 The variables smtp_input and smtp_batched_input are exported to
5158 55. Changed the default runtime configuration: the message "Unknown user"
5159 has been removed from the ACL, and instead placed on the localuser router,
5160 using the cannot_route_message feature. This means that any verification
5161 failures that generate their own messages won't get overridden. Similarly,
5162 the "Unrouteable address" message that was in the ACL for unverifiable
5163 relay addresses has also been removed.
5165 56. Added hosts_avoid_esmtp to the smtp transport.
5167 57. The exicyclog script was not checking for the esoteric option
5168 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5169 will work only if exicyclog is run under the appropriate euid.
5171 58. Following a discussion on the list, the rules by which Exim recognises line
5172 endings on incoming messages have been changed. The -dropcr and drop_cr
5173 options are now no-ops, retained only for backwards compatibility. The
5174 following line terminators are recognized: LF CRLF CR. However, special
5175 processing applies to CR:
5177 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5178 nor a local message in the state where . is a terminator.
5180 (ii) If a bare CR is encountered in a header line, an extra space is added
5181 after the line terminator so as not to end the header. The reasoning
5182 behind this is that bare CRs in header lines are most likely either
5183 to be mistakes, or people trying to play silly games.
5185 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5186 was being incorrectly given as 18 bytes larger than it should have been.
5187 This is a VOB (very old bug).
5189 60. This may never have affected anything current, but just in case it has:
5190 When the local host is found other than at the start of a list of hosts,
5191 the local host, those with the same MX, and any that follow, are discarded.
5192 When the list in question was part of a longer list of hosts, the following
5193 hosts (not currently being processed) were also being discarded. This no
5194 longer happens. I'm not sure if this situation could ever has previously
5197 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5200 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5201 header. This is something that is recommended in a new Internet Draft, and
5202 is something that is documented as being done by Sendmail. There are two
5203 possible values. For messages generated by the autoreply transport, Exim
5206 Auto-Submitted: auto-replied
5208 whereas for all other generated messages (e.g. bounces) it adds
5210 Auto-Submitted: auto-generated
5212 63. The "personal" condition in filters now includes a test for the
5213 Auto-Submitted: header. If it contains the string "auto-" the message it
5214 not considered personal.
5216 64. Added rcpt_include_affixes as a generic transport option.
5218 65. Added queue_only_override (default true).
5220 66. Added the syslog_duplication option.
5222 67. If what should have been the first header line of a message consisted of
5223 a space followed by a colon, Exim was mis-interpreting it as a header line.
5224 It isn't of course - it is syntactically invalid and should therefore be
5225 treated as the start of the message body. The misbehaviour could have
5226 caused a number of strange effects, including loss of data in subsequent
5227 header lines, and spool format errors.
5229 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5230 client host had authenticated. This control can now be exercised by an ACL
5231 for more flexibility.
5233 69. By default, callouts do not happen when testing with -bh. There is now a
5234 variant, -bhc, which does actually run the callout code, including
5235 consulting and updating the callout cache.
5237 70. Added support for saslauthd authentication, courtesy of Alexander
5240 71. If statvfs() failed on the spool or log directories while checking their
5241 size for availability, Exim confusingly gave the error "space shortage".
5242 Furthermore, in debugging mode it crashed with a floating point exception.
5243 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5244 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5245 serious problem, Exim now writes to the main and panic logs when this
5246 happens, with details of the failure. It then refuses to accept the
5247 incoming message, giving the message "spool directory problem" or "log
5248 directory problem" with a 421 code for SMTP messages.
5250 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5251 0, 1, and 2 exist, because some OS complain for execs without them (see
5252 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5253 descriptors. However, the code omitted to check that the open succeeded,
5254 causing mysterious errors if for some reason the permissions on /dev/null
5255 got screwed. Now Exim writes a message to the main and panic logs, and
5256 bombs out if it can't open /dev/null.
5258 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5259 interact so that it is all more flexible. It is supposed to remain
5260 backwards compatible. Also added extra_local_interfaces.
5262 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5263 to bomb out with an assertion failure - to the client this appears as a
5264 connection drop. This problem occurs in the part of the code that was taken
5265 from the Samba project. Fortunately, the assertion is in a very simple
5266 function, so I have fixed this by reproducing the function inline in the
5267 one place where it is called, and arranging for authentication to fail
5268 instead of killing the process with assert().
5270 75. The SPA client code was not working when the server requested OEM rather
5271 than Unicode encoding.
5273 76. Added code to make require_files with a specific uid setting more usable in
5274 the case where statting the file as root fails - usually a non-root-mounted
5275 NFS file system. When this happens and the failure is EACCES, Exim now
5276 forks a subprocess and does the per-uid checking as the relevant uid.
5278 77. Added process_log_path.
5280 78. If log_file_path was not explicitly set, a setting of check_log_space or
5281 check_log_inodes was ignored.
5283 79. If a space check for the spool or log partitions fails, the incident is now
5284 logged. Of course, in the latter case the data may get lost...
5286 80. Added the %p formatting code to string_format() so that it can be used to
5287 print addresses in debug_print(). Adjusted all the address printing in the
5288 debugging in store.c to use %p rather than %d.
5290 81. There was a concern that a line of code in smtp_in.c could overflow a
5291 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5292 initially expressed, the concern was not well-founded, because trailing
5293 spaces are removed early. However, if the trailing spaces were followed by
5294 a NULL, they did not get removed, so the overflow was possible. Two fixes
5297 (a) I re-wrote the offending code in a cleaner fashion.
5298 (b) If an incoming SMTP command contains a NULL character, it is rejected
5301 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5302 runs initgroups(), so that if the Exim user is in any additional groups,
5303 they will be used during message reception.
5309 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob