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 dependancy 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
37 TF/01 Failure to get a lock on a hints database can have serious
38 consequences so log it to the panic log.
40 TF/02 Log LMTP confirmation messages in the same way as SMTP,
41 controlled using the smtp_confirmation log selector.
43 TF/03 Include the error message when we fail to unlink a spool file.
45 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
46 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
47 for maintaining out-of-tree patches for some time.
49 PP/01 Bugzilla 139: Documentation and portability issues.
50 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
51 Handle per-OS dynamic-module compilation flags.
53 PP/02 Let /dev/null have normal permissions.
54 The 4.73 fixes were a little too stringent and complained about the
55 permissions on /dev/null. Exempt it from some checks.
56 Reported by Andreas M. Kirchwitz.
58 PP/03 Report version information for many libraries, including
59 Exim version information for dynamically loaded libraries. Created
60 version.h, now support a version extension string for distributors
61 who patch heavily. Dynamic module ABI change.
63 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
64 privilege escalation vulnerability whereby the Exim run-time user
65 can cause root to append content of the attacker's choosing to
68 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
71 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
72 If dropping privileges for untrusted macros, we disabled normal logging
73 on the basis that it would fail; for the Exim run-time user, this is not
74 the case, and it resulted in successful deliveries going unlogged.
75 Fixed. Reported by Andreas Metzler.
81 PP/01 Date: & Message-Id: revert to normally being appended to a message,
82 only prepend for the Resent-* case. Fixes regression introduced in
83 Exim 4.70 by NM/22 for Bugzilla 607.
85 PP/02 Include check_rfc2047_length in configure.default because we're seeing
86 increasing numbers of administrators be bitten by this.
88 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
90 PP/03 Bugzilla 994: added openssl_options main configuration option.
92 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
94 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
96 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
98 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
99 without a peer certificate, leading to a segfault because of an
100 assumption that peers always have certificates. Be a little more
101 paranoid. Problem reported by Martin Tscholak.
103 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
104 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
105 NB: ClamAV planning to remove STREAM in "middle of 2010".
106 CL also introduces -bmalware, various -d+acl logging additions and
107 more caution in buffer sizes.
109 PP/09 Implemented reverse_ip expansion operator.
111 PP/10 Bugzilla 937: provide a "debug" ACL control.
113 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
115 PP/12 Bugzilla 973: Implement --version.
117 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
119 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
121 PP/15 Bugzilla 816: support multiple condition rules on Routers.
123 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
124 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
125 ignore trailing whitespace.
127 JJ/02 prevent non-panic DKIM error from being sent to paniclog
129 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
132 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
133 Notification from Dr Andrew Aitchison.
135 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
136 ExtendedDetectionInfo response format.
137 Notification from John Horne.
139 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
142 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
143 XSL and documented dependency on system catalogs, with examples of how
146 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
149 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
150 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
151 configuration file which is writeable by the Exim user or group.
153 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
154 of configuration files to cover files specified with the -C option if
155 they are going to be used with root privileges, not just the default
158 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
159 option (effectively making it always true).
161 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
162 files to be used while preserving root privileges.
164 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
165 that rogue child processes cannot use them.
167 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
168 run-time user, instead of root.
170 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
171 Exim run-time user without dropping privileges.
173 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
174 result string, instead of calling string_vformat() twice with the same
177 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
178 for other users. Others should always drop root privileges if they use
179 -C on the command line, even for a whitelisted configure file.
181 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
183 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
189 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
190 $data_path, and $header_path variables; fixed documentation bugs and
193 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
194 exipick to access non-standard spools, including the "frozen" queue
197 NM/01 Bugzilla 965: Support mysql stored procedures.
198 Patch from Alain Williams
200 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
202 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
203 Patch from Andreas Metzler
205 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
206 Patch from Kirill Miazine
208 NM/05 Bugzilla 671: Added umask to procmail example.
210 JJ/03 installed exipick 20100323.0, fixing doc bug
212 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
213 directory. Notification and patch from Dan Rosenberg.
215 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
217 TK/02 Improve log output when DKIM signing operation fails.
219 MH/01 Treat the transport option dkim_domain as a colon separated
220 list, not as a single string, and sign the message with each element,
221 omitting multiple occurences of the same signer.
223 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
224 Bugzilla 985, 986. Patch by Simon Arlott
226 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
227 Patch by Simon Arlott
229 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
230 MBX locking. Notification from Dan Rosenberg.
236 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
238 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
240 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
242 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
244 NM/04 Bugzilla 915: Flag broken perl installation during build.
250 TK/01 Added patch by Johannes Berg that expands the main option
251 "spamd_address" if it starts with a dollar sign.
253 TK/02 Write list of recipients to X-Envelope-Sender header when building
254 the mbox-format spool file for content scanning (suggested by Jakob
257 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
258 (http://www.dcc-servers.net/) support via dccifd. Activated by
259 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
261 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
262 by Mark Daniel Reidel <mr@df.eu>.
264 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
265 When building exim an external PCRE library is now needed -
266 PCRE is a system library on the majority of modern systems.
267 See entry on PCRE_LIBS in EDITME file.
269 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
270 conversation. Added nologin parameter to request.
271 Patch contributed by Kirill Miazine.
273 TF/01 Do not log submission mode rewrites if they do not change the address.
275 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
277 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
278 log files in place. Contributed by Roberto Lima.
280 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
282 TF/03 Bugzilla 615: When checking the local_parts router precondition
283 after a local_part_suffix or local_part_prefix option, Exim now
284 does not use the address's named list lookup cache, since this
285 contains cached lookups for the whole local part.
287 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
288 Robert Millan. Documentation is in experimental-spec.txt.
290 TF/04 Bugzilla 668: Fix parallel build (make -j).
292 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
294 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
295 Patch provided by Jan Srzednicki.
297 TF/05 Leading white space used to be stripped from $spam_report which
298 wrecked the formatting. Now it is preserved.
300 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
301 that they are available at delivery time.
303 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
305 TF/08 TLS error reporting now respects the incoming_interface and
306 incoming_port log selectors.
308 TF/09 Produce a more useful error message if an SMTP transport's hosts
309 setting expands to an empty string.
311 NM/06 Bugzilla 744: EXPN did not work under TLS.
312 Patch provided by Phil Pennock.
314 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
315 Patch provided by Richard Godbee.
317 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
318 acl_smtp_notquit, added index entry.
320 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
321 Patch provided by Eugene Bujak.
323 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
324 accept(). Patch provided by Maxim Dounin.
326 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
327 Patch provided by Phil Pennock.
329 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
331 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
332 Patch provided by Brad "anomie" Jorsch.
334 NM/14 Bugzilla 622: Added timeout setting to transport filter.
335 Patch provided by Dean Brooks.
337 TK/05 Add native DKIM support (does not depend on external libraries).
339 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
340 Patch provided by Graeme Fowler.
342 NM/16 Bugzilla 851: Documentation example syntax fix.
344 NM/17 Changed NOTICE file to remove references to embedded PCRE.
346 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
349 NM/19 Bugzilla 745: TLS version reporting.
350 Patch provided by Phil Pennock.
352 NM/20 Bugzilla 167: bool: condition support.
353 Patch provided by Phil Pennock.
355 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
356 clients. Patch provided by Phil Pennock.
358 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
359 Patch provided by Brad "anomie" Jorsch.
361 NM/23 Bugzilla 687: Fix misparses in eximstats.
362 Patch provided by Heiko Schlittermann.
364 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
365 Patch provided by Heiko Schlittermann.
367 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
368 plus update to original patch.
370 NM/26 Bugzilla 799: Documentation correction for ratelimit.
372 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
373 Patch provided by David Brownlee.
375 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
377 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
379 NM/30 Bugzilla 888: TLS documentation bugfixes.
381 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
383 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
384 Unlike the original bugzilla I have changed all shell scripts in src tree.
386 NM/33 Bugzilla 898: Transport filter timeout fix.
387 Patch by Todd Rinaldo.
389 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
390 Patch by Serge Demonchaux.
392 NM/35 Bugzilla 39: Base64 decode bug fixes.
393 Patch by Jakob Hirsch.
395 NM/36 Bugzilla 909: Correct connect() call in dcc code.
397 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
399 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
401 NM/39 Bugzilla 911: Fixed MakeLinks build script.
407 TK/01 Add preliminary DKIM support. Currently requires a forked version of
408 ALT-N's libdkim that I have put here:
409 http://duncanthrax.net/exim-experimental/
411 Note to Michael Haardt: I had to rename some vars in sieve.c. They
412 were called 'true' and it seems that C99 defines that as a reserved
413 keyword to be used with 'bool' variable types. That means you could
414 not include C99-style headers which use bools without triggering
415 build errors in sieve.c.
417 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
418 as mailq or other aliases. Changed the --help handling significantly
419 to do whats expected. exim_usage() emits usage/help information.
421 SC/01 Added the -bylocaldomain option to eximstats.
423 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
425 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
427 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
433 PH/01 Another patch from the Sieve maintainer.
435 PH/02 When an IPv6 address is converted to a string for single-key lookup
436 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
437 dots are used instead of colons so that keys in lsearch files need not
438 contain colons. This was done some time before quoting was made available
439 in lsearch files. However, iplsearch files do require colons in IPv6 keys
440 (notated using the quote facility) so as to distinguish them from IPv4
441 keys. This meant that lookups for IP addresses in host lists did not work
442 for iplsearch lookups.
444 This has been fixed by arranging for IPv6 addresses to be expressed with
445 colons if the lookup type is iplsearch. This is not incompatible, because
446 previously such lookups could never work.
448 The situation is now rather anomolous, since one *can* have colons in
449 ordinary lsearch keys. However, making the change in all cases is
450 incompatible and would probably break a number of configurations.
452 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
455 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
456 conversion specification without a maximum field width, thereby enabling
457 a rogue spamd server to cause a buffer overflow. While nobody in their
458 right mind would setup Exim to query an untrusted spamd server, an
459 attacker that gains access to a server running spamd could potentially
460 exploit this vulnerability to run arbitrary code as the Exim user.
462 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
463 $primary_hostname instead of what libspf2 thinks the hosts name is.
465 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
466 a directory entry by the name of the lookup key. Previously, if a
467 symlink pointed to a non-existing file or a file in a directory that
468 Exim lacked permissions to read, a lookup for a key matching that
469 symlink would fail. Now it is enough that a matching directory entry
470 exists, symlink or not. (Bugzilla 503.)
472 PH/03 The body_linecount and body_zerocount variables are now exported in the
475 PH/04 Added the $dnslist_matched variable.
477 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
478 This means they are set thereafter only if the connection becomes
481 PH/06 Added the client_condition to authenticators so that some can be skipped
482 by clients under certain conditions.
484 PH/07 The error message for a badly-placed control=no_multiline_responses left
485 "_responses" off the end of the name.
487 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
489 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
490 (without spaces) instead of just copying the configuration text.
492 PH/10 Added the /noupdate option to the ratelimit ACL condition.
494 PH/11 Added $max_received_linelength.
496 PH/12 Added +ignore_defer and +include_defer to host lists.
498 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
499 way in which PCRE > 7.0 is built.
501 PH/14 Implemented queue_only_load_latch.
503 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
504 MAIL command. The effect was to mangle the value on 64-bit systems.
506 PH/16 Another patch from the Sieve maintainer.
508 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
510 PH/18 If a system quota error occurred while trying to create the file for
511 a maildir delivery, the message "Mailbox is full" was not appended to the
512 bounce if the delivery eventually timed out. Change 4.67/27 below applied
513 only to a quota excession during the actual writing of the file.
515 PH/19 It seems that peer DN values may contain newlines (and other non-printing
516 characters?) which causes problems in log lines. The DN values are now
517 passed through string_printing() before being added to log lines.
519 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
520 and InterBase are left for another time.)
522 PH/21 Added message_body_newlines option.
524 PH/22 Guard against possible overflow in moan_check_errorcopy().
526 PH/23 POSIX allows open() to be a macro; guard against that.
528 PH/24 If the recipient of an error message contained an @ in the local part
529 (suitably quoted, of course), incorrect values were put in $domain and
530 $local_part during the evaluation of errors_copy.
536 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
537 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
540 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
541 issue a MAIL command.
543 PH/02 In an ACL statement such as
545 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
547 if a client was not listed at all, or was listed with a value other than
548 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
549 the condition was not true (as it should be), so access was not denied.
550 The bug was that the ! inversion was incorrectly passed on to the second
551 item. This has been fixed.
553 PH/03 Added additional dnslists conditions == and =& which are different from
554 = and & when the dns lookup returns more than one IP address.
556 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
557 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
559 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
560 FSYNC, which compiles an option called disable_fsync that allows for
561 bypassing fsync(). The documentation is heavily laced with warnings.
563 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
565 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
566 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
567 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
568 including adding "make clean"; (3) Added -fPIC when compiling the test
569 dynamically loaded module, to get rid of a warning.
571 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
572 message fails, move_frozen_messages = true and ignore_bounce_errors_after
573 = 0s. The bug is otherwise harmless.
575 PH/07 There was a bug in the dovecot authenticator such that the value of
576 $auth1 could be overwritten, and so not correctly preserved, after a
577 successful authentication. This usually meant that the value preserved by
578 the server_setid option was incorrect.
580 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
582 PH/09 Installed PCRE release 7.0.
584 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
585 run for batched SMTP input. It is now run at the start of every message
586 in the batch. While fixing this I discovered that the process information
587 (output by running exiwhat) was not always getting set for -bs and -bS
588 input. This is fixed, and it now also says "batched" for BSMTP.
590 PH/11 Added control=no_pipelining.
592 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
593 patch, slightly modified), and move the expansion of helo_data till after
594 the connection is made in the smtp transport (so it can use these
597 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
599 PH/14 Added log_selector = +pid.
601 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
603 PH/16 Add ${if forany and ${if forall.
605 PH/17 Added dsn_from option to vary the From: line in DSNs.
607 PH/18 Flush SMTP output before performing a callout, unless control =
608 no_callout_flush is set.
610 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
611 was true (the default) a successful delivery failed to delete the retry
612 item, thus causing premature timeout of the address. The bug is now
615 PH/20 Added hosts_avoid_pipelining to the smtp transport.
617 PH/21 Long custom messages for fakedefer and fakereject are now split up
618 into multiline reponses in the same way that messages for "deny" and
619 other ACL rejections are.
621 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
622 with slight modification.
624 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
625 draft, changing the syntax and factoring some duplicate code".
627 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
628 for deliveries of the second and subsequent messages over the same SMTP
631 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
632 ${reduce, with only minor "tidies".
634 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
636 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
637 expansion side effects.
639 PH/27 When a message times out after an over-quota error from an Exim-imposed
640 quota, the bounce message says "mailbox is full". This message was not
641 being given when it was a system quota that was exceeded. It now should
644 MH/03 Made $recipients available in local_scan(). local_scan() already has
645 better access to the recipient list through recipients_list[], but
646 $recipients can be useful in postmaster-provided expansion strings.
648 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
649 in the case of a MAIL command with additional options following the
650 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
651 were accidentally chopped off.
653 PH/29 SMTP synchronization checks are implemented when a command is read -
654 there is a check that no more input is waiting when there shouldn't be
655 any. However, for some commands, a delay in an ACL can mean that it is
656 some time before the response is written. In this time, more input might
657 arrive, invalidly. So now there are extra checks after an ACL has run for
658 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
659 pipelining has not been advertised.
661 PH/30 MH's patch to allow iscntrl() characters to be list separators.
663 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
664 returned in the SMTP response when smtp_return_error_details was false.
667 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
668 instead of the C I/O that was originally supplied, because problems were
671 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
672 Exim which did not show up earlier: it was assuming that a call to
673 SSL_CTX_set_info_callback() might give an error value. In fact, there is
674 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
675 was a macro that became an assignment, so it seemed to work. This has
676 changed to a proper function call with a void return, hence the compile
677 error. Exim's code has been fixed.
679 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
682 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
684 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
686 PH/37 If a message is not accepted after it has had an id assigned (e.g.
687 because it turns out to be too big or there is a timeout) there is no
688 "Completed" line in the log. When some messages of this type were
689 selected by exigrep, they were listed as "not completed". Others were
690 picked up by some special patterns. I have improved the selection
691 criteria to be more general.
693 PH/38 The host_find_failed option in the manualroute router can now be set
694 to "ignore", to completely ignore a host whose IP address cannot be
695 found. If all hosts are ignored, the behaviour is controlled by the new
696 host_all_ignored option.
698 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
699 homing or because of multiple MX records with /mx) generated more than
700 one IP address, and the following item turned out to be the local host,
701 all the secondary addresses of the first item were incorrectly removed
702 from the list, along with the local host and any following hosts (which
703 is what is supposed to happen).
705 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
706 whoever called Exim into the -H file. In the case of the daemon it was
707 behaving confusingly. When first started, it used values for whoever
708 started the daemon, but after a SIGHUP it used the Exim user (because it
709 calls itself on a restart). I have changed the code so that it now always
712 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
713 message are rejected with the same error (e.g. no authentication or bad
714 sender address), and a DATA command is nevertheless sent (as can happen
715 with PIPELINING or a stupid MUA), the error message that was given to the
716 RCPT commands is included in the rejection of the DATA command. This is
717 intended to be helpful for MUAs that show only the final error to their
720 PH/42 Another patch from the Sieve maintainer.
722 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
723 Eximstats - Fixed some broken HTML links and added missing column headers
725 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
726 columns for Rejects, Temp Rejects, Ham, and Spam rows.
728 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
730 PH/43 Yet another patch from the Sieve maintainer.
732 PH/44 I found a way to check for a TCP/IP connection going away before sending
733 the response to the final '.' that terminates a message, but only in the
734 case where the client has not sent further data following the '.'
735 (unfortunately, this is allowed). However, in many cases there won't be
736 any further data because there won't be any more messages to send. A call
737 to select() can be used: if it shows that the input is "ready", there is
738 either input waiting, or the socket has been closed. An attempt to read
739 the next input character can distinguish the two cases. Previously, Exim
740 would have sent an OK response which the client would never have see.
741 This could lead to message repetition. This fix should cure that, at
742 least in a lot of common cases.
744 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
745 advertised in response to EHLO.
751 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
752 fixed by 4.65/MH/01 (is this a record?) are fixed:
754 (i) An empty string was always treated as zero by the numeric comparison
755 operators. This behaviour has been restored.
757 (ii) It is documented that the numeric comparison operators always treat
758 their arguments as decimal numbers. This was broken in that numbers
759 starting with 0 were being interpreted as octal.
761 While fixing these problems I realized that there was another issue that
762 hadn't been noticed. Values of message_size_limit (both the global option
763 and the transport option) were treated as octal if they started with 0.
764 The documentation was vague. These values are now always treated as
765 decimal, and I will make that clear in the documentation.
771 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
772 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
775 MH/01 Don't check that the operands of numeric comparison operators are
776 integers when their expansion is in "skipping" mode (fixes bug
777 introduced by 4.64-PH/07).
779 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
780 child addresses, Exim now panics and dies. Previously, because the count
781 is held in a short int, deliveries were likely to be lost. As such a
782 large number of recipients for a single message is ridiculous
783 (performance will be very, very poor), I have chosen to impose a limit
784 rather than extend the field.
790 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
791 leftover -K file (the existence of which was triggered by #402).
792 While we were at it, introduced process PID as part of the -K
793 filename. This should rule out race conditions when creating
796 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
797 processing considerably. Previous code took too long for large mails,
798 triggering a timeout which in turn triggers #401.
800 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
801 in the DK code in transports.c. sendfile() is not really portable,
802 hence the _LINUX specificness.
804 TF/01 In the add_headers option to the mail command in an Exim filter,
805 there was a bug that Exim would claim a syntax error in any
806 header after the first one which had an odd number of characters
809 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
810 callout verification, Exim cached a "reject" for the entire domain. This
811 is correct for most verifications, but it is not correct for a recipient
812 verification with use_sender or use_postmaster set, because in that case
813 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
814 case of MAIL FROM:<> rejection from other early rejections (e.g.
815 rejection of HELO). When verifying a recipient using a non-null MAIL
816 address, the cache is ignored if it shows MAIL FROM:<> rejection.
817 Whatever the result of the callout, the value of the domain cache is
818 left unchanged (for any other kind of callout, getting as far as trying
819 RCPT means that the domain itself is ok).
821 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
824 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
825 manifest itself as EPIPE rather than ECONNECT. When tidying away a
826 session, the daemon ignores ECONNECT errors and logs others; it now
827 ignores EPIPE as well.
829 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
830 (quoted-printable decoding).
832 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
833 later the small subsequent patch to fix an introduced bug.
835 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
837 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
839 PH/08 An error is now given if message_size_limit is specified negative.
841 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
842 to be given (somewhat) arbitrary names.
844 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
847 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
848 miscellaneous code fixes
850 PH/10 Added the log_reject_target ACL modifier to specify where to log
853 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
854 hostname. This is wrong, because it relates to the incoming message (and
855 probably the interface on which it is arriving) and not to the outgoing
856 callout (which could be using a different interface). This has been
857 changed to use the value of the helo_data option from the smtp transport
858 instead - this is what is used when a message is actually being sent. If
859 there is no remote transport (possible with a router that sets up host
860 addresses), $smtp_active_hostname is used.
862 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
863 tweaks were necessary in order to get it to work (see also 21 below):
864 (a) The code assumed that strncpy() returns a negative number on buffer
865 overflow, which isn't the case. Replaced with Exim's string_format()
867 (b) There were several signed/unsigned issues. I just did the minimum
868 hacking in of casts. There is scope for a larger refactoring.
869 (c) The code used strcasecmp() which is not a standard C function.
870 Replaced with Exim's strcmpic() function.
871 (d) The code set only $1; it now sets $auth1 as well.
872 (e) A simple test gave the error "authentication client didn't specify
873 service in request". It would seem that Dovecot has changed its
874 interface. Fortunately there's a specification; I followed it and
875 changed what the client sends and it appears to be working now.
877 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
880 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
881 address A is aliased to B and C, where B exists and C does not. Without
882 -v the output is "A verified" because verification stops after a
883 successful redirection if more than one address is generated. However,
884 with -v the child addresses are also verified. Exim was outputting "A
885 failed to verify" and then showing the successful verification for C,
886 with its parentage. It now outputs "B failed to verify", showing B's
887 parentage before showing the successful verification of C.
889 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
890 look up a TXT record in a specific list after matching in a combined
893 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
894 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
895 they consult the DNS. I had assumed they would set it the way they
896 wanted; and indeed my experiments on Linux seem to show that in some
897 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
898 To be on the safe side, however, I have now made the interface to
899 host_find_byname() similar to host_find_bydns(), with an argument
900 containing the DNS resolver options. The host_find_byname() function now
901 sets these options at its start, just as host_find_bydns() does. The smtp
902 transport options dns_qualify_single and dns_search_parents are passed to
903 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
904 of host_find_byname() use the default settings of RES_DEFNAMES
905 (qualify_single) but not RES_DNSRCH (search_parents).
907 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
908 spool_read_header() do less string testing, by means of a preliminary
909 switch on the second character of optional "-foo" lines. (This is
910 overdue, caused by the large number of possibilities that now exist.
911 Originally there were few.) While I was there, I also converted the
912 str(n)cmp tests so they don't re-test the leading "-" and the first
913 character, in the hope this might squeeze out yet more improvement.
915 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
916 flag allowing group syntax was set by the header_syntax check but not
917 turned off, possible causing trouble later; (2) The flag was not being
918 set at all for the header_verify test, causing "group"-style headers to
919 be rejected. I have now set it in this case, and also caused header_
920 verify to ignore an empty address taken from a group. While doing this, I
921 came across some other cases where the code for allowing group syntax
922 while scanning a header line wasn't quite right (mostly, not resetting
923 the flag correctly in the right place). These bugs could have caused
924 trouble for malformed header lines. I hope it is now all correct.
926 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
927 with the "reply" argument non-NULL. The code, however (which originally
928 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
929 but it didn't always do it. This confused somebody who was copying the
930 code for some other use. I have removed all the tests.
932 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
933 feature that was used to support insecure browsers during the U.S. crypto
934 embargo. It requires special client support, and Exim is probably the
935 only MTA that supported it -- and would never use it because real RSA is
936 always available. This code has been removed, because it had the bad
937 effect of slowing Exim down by computing (never used) parameters for the
938 RSA_EXPORT functionality.
940 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
941 authenticator to fail if there's a tab character in the incoming data
942 (there should never be unless someone is messing about, as it's supposed
943 to be base64-encoded). Also added, on Timo's advice, the "secured" option
944 if the connection is using TLS or if the remote IP is the same as the
945 local IP, and the "valid-client-cert option" if a client certificate has
948 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
949 authenticators. This can be used for authorization after authentication
950 succeeds. (In the case of plaintext, it servers for both authentication
953 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
954 if any retry times were supplied.
956 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
957 connection, where there is no client IP address to check. In this
958 situation, the verify now always succeeds.
960 PH/25 Applied John Jetmore's -Mset patch.
962 PH/26 Added -bem to be like -Mset, but loading a message from a file.
964 PH/27 In a string expansion for a processed (not raw) header when multiple
965 headers of the same name were present, leading whitespace was being
966 removed from all of them, but trailing whitespace was being removed only
967 from the last one. Now trailing whitespace is removed from each header
968 before concatenation. Completely empty headers in a concatenation (as
971 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
972 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
974 PH/29 [Removed. This was a change that I later backed out, and forgot to
975 correct the ChangeLog entry (that I had efficiently created) before
976 committing the later change.]
978 PH/30 Exim was sometimes attempting to deliver messages that had suffered
979 address errors (4xx response to RCPT) over the same connection as other
980 messages routed to the same hosts. Such deliveries are always "forced",
981 so retry times are not inspected. This resulted in far too many retries
982 for the affected addresses. The effect occurred only when there were more
983 hosts than the hosts_max_try setting in the smtp transport when it had
984 the 4xx errors. Those hosts that it had tried were not added to the list
985 of hosts for which the message was waiting, so if all were tried, there
986 was no problem. Two fixes have been applied:
988 (i) If there are any address or message errors in an SMTP delivery, none
989 of the hosts (tried or untried) are now added to the list of hosts
990 for which the message is waiting, so the message should not be a
991 candidate for sending over the same connection that was used for a
992 successful delivery of some other message. This seems entirely
993 reasonable: after all the message is NOT "waiting for some host".
994 This is so "obvious" that I'm not sure why it wasn't done
995 previously. Hope I haven't missed anything, but it can't do any
996 harm, as the worst effect is to miss an optimization.
998 (ii) If, despite (i), such a delivery is accidentally attempted, the
999 routing retry time is respected, so at least it doesn't keep
1000 hammering the server.
1002 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1003 in ${readsocket because some servers need this prod.
1005 PH/32 Added some extra debug output when updating a wait-xxx database.
1007 PH/33 The hint "could be header name not terminated by colon", which has been
1008 given for certain expansion errors for a long time, was not being given
1009 for the ${if def:h_colon_omitted{... case.
1011 PH/34 The spec says: "With one important exception, whenever a domain list is
1012 being scanned, $domain contains the subject domain." There was at least
1013 one case where this was not true.
1015 PH/35 The error "getsockname() failed: connection reset by peer" was being
1016 written to the panic log as well as the main log, but it isn't really
1017 panic-worthy as it just means the connection died rather early on. I have
1018 removed the panic log writing for the ECONNRESET error when getsockname()
1021 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1022 runs only) independently of the message's sender address. This meant
1023 that, if the 4xx error was in fact related to the sender, a different
1024 message to the same recipient with a different sender could confuse
1025 things. In particualar, this can happen when sending to a greylisting
1026 server, but other circumstances could also provoke similar problems.
1027 I have changed the default so that the retry time for these errors is now
1028 based a combination of the sender and recipient addresses. This change
1029 can be overridden by setting address_retry_include_sender=false in the
1032 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1033 remote server are returned as part of bounce messages. This was not
1034 happening for LMTP over a pipe (the lmtp transport), but now it is the
1035 same for both kinds of LMTP.
1037 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1038 in header lines that were in fact CNAMEs. This is no longer the case.
1040 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1041 and queue runs started by the daemon processed all messages. This has
1042 been fixed so that -R and -S can now usefully be given with -q<time>.
1044 PH/40 Import PCRE release 6.7 (fixes some bugs).
1046 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1048 PH/42 Give an error if -q is specified more than once.
1050 PH/43 Renamed the variables $interface_address and $interface_port as
1051 $received_ip_address and $received_port, to make it clear that these
1052 values apply to message reception, and not to the outgoing interface when
1053 a message is delivered. (The old names remain recognized, of course.)
1055 PH/44 There was no timeout on the connect() call when using a Unix domain
1056 socket in the ${readsocket expansion. There now is.
1058 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1059 be meaningful with "accept".
1061 SC/01 Eximstats V1.43
1062 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1064 SC/02 Eximstats V1.44
1065 Use a glob alias rather than an array ref in the generated
1066 parser. This improves both readability and performance.
1068 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1069 Collect SpamAssassin and rejection statistics.
1070 Don't display local sender or destination tables unless
1071 there is data to show.
1072 Added average volumes into the top table text output.
1074 SC/04 Eximstats V1.46
1075 Collect data on the number of addresses (recipients)
1076 as well as the number of messages.
1078 SC/05 Eximstats V1.47
1079 Added 'Message too big' to the list of mail rejection
1080 reasons (thanks to Marco Gaiarin).
1082 SC/06 Eximstats V1.48
1083 Mainlog lines which have GMT offsets and are too short to
1084 have a flag are now skipped.
1086 SC/07 Eximstats V1.49 (Alain Williams)
1087 Added the -emptyok flag.
1089 SC/08 Eximstats V1.50
1090 Fixes for obtaining the IP address from reject messages.
1092 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1093 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1094 whitesspace changes from 4.64-PH/27
1096 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1099 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1100 are found, allow negative numbers in numeric criteria)
1102 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1104 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1105 to match changes made in 4.64-PH/43
1107 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1109 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1110 "domain" is actually an IP address, and force a failure. This locks out
1111 those revolvers/nameservers that support "A-for-A" lookups, in
1112 contravention of the specifications.
1114 PH/48 When a host name was looked up from an IP address, and the subsequent
1115 forward lookup of the name timed out, the host name was left in
1116 $sender_host_name, contrary to the specification.
1118 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1119 restricted to single-key lookups, Exim was not diagnosing an error if
1120 * or *@ was used with a query-style lookup.
1122 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1124 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1125 long ago, but noone interested enough thought of it. Let's just say that
1126 the "1.1" means that there are some new functions that weren't there at
1127 some point in the past.
1129 PH/51 Error processing for expansion failure of helo_data from an smtp
1130 transport during callout processing was broken.
1132 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1133 tested/used via the -bh/-bhc/-bs options.
1135 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1136 bug, fixed in subsequent PCRE releases).
1138 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1139 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1141 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1147 SC/01 Use a glob alias rather than an array ref in eximstats generated
1148 parser. This improves both readability and performance.
1150 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1151 Don't display local sender or destination tables in eximstats unless
1152 there is data to show.
1153 Added average volumes into the eximstats top table text output.
1155 SC/03 Collect data on the number of addresses (recipients) as well
1156 as the number of messages in eximstats.
1158 TF/01 Correct an error in the documentation for the redirect router. Exim
1159 does (usually) call initgroups() when daemonizing.
1161 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1162 with consistent privilege compared to when running as a daemon.
1164 TF/03 Note in the spec that $authenticated_id is not set for local
1165 submissions from trusted users.
1167 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1168 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1170 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1171 by adding some example configuration directives to the default
1172 configuration file. A little bit of work is required to uncomment the
1173 directives and define how usernames and passwords are checked, but
1174 there is now a framework to start from.
1176 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1177 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1178 without this. I don't know how relevant this is to other LDAP libraries.
1180 PH/02 Add the verb name to the "unknown ACL verb" error.
1182 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1184 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1186 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1187 directory not expanded when it should be if an expanded home directory
1188 was set for the address (which is overridden by the transport).
1190 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1193 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1194 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1195 because it is too late at that time, and has no effect.
1197 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1198 security issue with \' (bugzilla #107). I could not use the
1199 PQescapeStringConn() function, because it needs a PGconn value as one of
1202 PH/08 When testing addresses using -bt, indicate those final addresses that
1203 are duplicates that would not cause an additional delivery. At least one
1204 person was confused, thinking that -bt output corresponded to deliveries.
1205 (Suppressing duplicates isn't a good idea as you lose the information
1206 about possibly different redirections that led to the duplicates.)
1208 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1209 systems where poll() doesn't work, in particular OS X.
1211 PH/10 Added more information to debugging output for retry time not reached.
1213 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1214 operations in malware.c.
1216 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1219 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1220 syslog_duplication set false, log lines that would normally be written
1221 both the the main log and to the reject log were not written to syslog at
1224 PH/14 In the default configuration, change the use of "message" in ACL warn
1225 statements to "add_header".
1227 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1228 not followed by a command (e.g. "seen endif").
1230 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1231 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1234 PH/17 Added extra conditions to the default value of delay_warning_condition
1238 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1239 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1240 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1243 The Auto-Submitted: and various List- headers are standardised, whereas I
1244 don't think Precedence: ever was.
1246 PH/18 Refactored debugging code in route_finduser() to show more information,
1247 in particular, the error code if getpwnam() issues one.
1249 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1250 This is apparently needed in addition to the PH/07 change above to avoid
1251 any possible encoding problems.
1253 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1254 but not after initializing Perl.
1256 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1257 output them only if debugging. By default they are written stderr,
1258 apparently, which is not desirable.
1260 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1263 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1266 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1268 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1269 authenticated or an ident call has been made. Suppress the default
1270 values for $authenticated_id and $authenticated_sender (but permit -oMai
1271 and -oMas) when testing with -bh.
1273 PH/24 Re-jigged the order of the tests in the default configuration so that the
1274 tests for valid domains and recipients precede the DNS black list and CSA
1275 tests, on the grounds that those ones are more expensive.
1277 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1278 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1279 This bug exists in every version of Exim that I still have, right back to
1282 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1283 However, an attempt to turn on case-sensitivity in a regex key by
1284 including (?-i) didn't work because the subject string was already
1285 lowercased, and the effects were non-intuitive. It turns out that a
1286 one-line patch can be used to allow (?-i) to work as expected.
1292 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1293 other effects) broke the use of negated acl sub-conditions.
1295 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1298 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1299 "Deny" causes Exim to reject the incoming connection with a 554 error.
1300 Unfortunately, if there is a major crisis, such as a disk failure,
1301 tcp-wrappers gives "deny", whereas what one would like would be some
1302 kind of temporary error. A kludge has been added to help with this.
1303 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1304 554 error is used if errno is still zero or contains ENOENT (which occurs
1305 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1308 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1310 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1311 errors. Otherwise a message that provokes a temporary error (when other
1312 messages do not) can cause a whole host to time out.
1314 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1315 addresses were routed directly to files or pipes from a redirect router.
1316 File deliveries just didn't batch; pipe deliveries might have suffered
1319 PH/06 A failure to get a lock for a hints database would erroneously always say
1320 "Failed to get write lock", even when it was really a read lock.
1322 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1323 of 0600. This has been changed to use the value of the lockfile_mode
1324 option (which defaults to 0600).
1326 PH/08 Applied small patch from the Sieve maintainer.
1328 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1329 folder from quota calculations, a direct delivery into this folder messed
1330 up the contents of the maildirsize file. This was because the regex was
1331 used only to exclude .Trash (or whatever) when the size of the mailbox
1332 was calculated. There was no check that a delivery was happening into an
1333 excluded directory. This bug has been fixed by ignoring all quota
1334 processing for deliveries into excluded directories.
1336 PH/10 Added the maildirfolder_create_regex option to appendfile.
1342 PH/01 The code for finding all the local interface addresses on a FreeBSD
1343 system running IPv6 was broken. This may well have applied to all BSD
1344 systems, as well as to others that have similar system calls. The broken
1345 code found IPv4 interfaces correctly, but gave incorrect values for the
1346 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1347 that it would not match correctly against @[] and not recognize the IPv6
1350 PH/02 The ipliteral router was not recognizing addresses of the form user@
1351 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1353 PH/03 Added disable_ipv6.
1355 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1356 decoded form, because it is most often used to construct To: headers
1357 lines in autoreplies, and the decoded form may well be syntactically
1358 invalid. However, $reply_address has leading white space removed, and all
1359 newlines turned into spaces so that the autoreply transport does not
1362 PH/05 If group was specified without a user on a router, and no group or user
1363 was specified on a transport, the group from the router was ignored.
1365 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1366 for visible compile-time settings that can be used to change these
1367 numbers, for those that want even more. Backwards compatibility with old
1368 spool files has been maintained. However, going back to a previous Exim
1369 release will lost any variables that are in spool files.
1371 PH/07 Two small changes when running in the test harness: increase delay when
1372 passing a TCP/IP connection to a new process, in case the original
1373 process has to generate a bounce, and remove special handling of
1374 127.0.0.2 (sic), which is no longer necessary.
1376 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1377 be the same on different OS.
1379 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1382 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1383 whether --show-vars was specified or not
1385 JJ/02 exipick: Added support for new ACL variable spool format introduced
1388 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1389 syntactically invalid From: or Reply-to: line, and a filter used this to
1390 generate an autoreply, and therefore failed to obtain an address for the
1391 autoreply, Exim could try to deliver to a non-existent relative file
1392 name, causing unrelated and misleading errors. What now happens is that
1393 it logs this as a hard delivery error, but does not attempt to create a
1396 PH/11 The exinext utility has a -C option for testing purposes, but although
1397 the given file was scanned by exinext itself; it wasn't being passed on
1398 when Exim was called.
1400 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1401 an end-of-file indication when reading a command response.
1403 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1404 compiled. In many other places in Exim, IPv6 addresses are always
1405 recognized, so I have changed this. It also means that IPv4 domain
1406 literals of the form [IPV4:n.n.n.n] are now always recognized.
1408 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1409 used if the router is not running as root, for example, when verifying at
1410 ACL time, or when using -bh. The debugging output from this situation was
1411 non-existent - all you got was a failure to exec. I have made two
1414 (a) Failures to set uid/gid, the current directory, or a process leader
1415 in a subprocess such as that created by queryprogram now generate
1416 suitable debugging ouput when -d is set.
1418 (b) The queryprogram router detects when it is not running as root,
1419 outputs suitable debugging information if -d is set, and then runs
1420 the subprocess without attempting to change uid/gid.
1422 PH/15 Minor change to Makefile for building test_host (undocumented testing
1425 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1426 additional section of a DNS packet that returns MX or SRV records.
1427 Instead, it always explicitly searches for A/AAAA records. This avoids
1428 major problems that occur when a DNS server includes only records of one
1429 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1430 fixed another bug: if SRV records were looked up and the corresponding
1431 address records were *not* found in the additional section, the port
1432 values from the SRV records were lost.
1434 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1435 using the correct key (the original address) when searching the retry
1436 rules in order to find which one to use for generating the retry hint.
1438 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1439 adding the default one. Similarly, if it contains a Reply-To: header, the
1440 errors_reply_to option, if set, is not used.
1442 PH/19 When calculating a retry time, Exim used to measure the "time since
1443 failure" by looking at the "first failed" field in the retry record. Now
1444 it does not use this if it is later than than the arrival time of the
1445 message. Instead it uses the arrival time. This makes for better
1446 behaviour in cases where some deliveries succeed, thus re-setting the
1447 "first failed" field. An example is a quota failure for a huge message
1448 when small messages continue to be delivered. Without this change, the
1449 "time since failure" will always be short, possible causing more frequent
1450 delivery attempts for the huge message than are intended.
1451 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1453 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1454 $1, $2, $3) because the numerical variables can be reset during some
1455 expansion items (e.g. "match"), thereby losing the authentication data.
1457 PH/21 Make -bV show the size of off_t variables so that the test suite can
1458 decide whether to run tests for quotas > 2G.
1460 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1461 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1462 filecount value is greater than 2G or if a quota value is greater than 2G
1463 on a system where the size of off_t is not greater than 4, a panic error
1466 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1467 never match. The debug and -bh output now contains an explicit error
1468 message indicating a malformed IPv4 address or mask.
1470 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1471 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1472 PH/23 above applies.
1474 PH/25 Do not write to syslog when running in the test harness. The only
1475 occasion when this arises is a failure to open the main or panic logs
1476 (for which there is an explicit test).
1478 PH/26 Added the /no_tell option to "control=freeze".
1480 PH/27 If a host name lookup failed very early in a connection, for example, if
1481 the IP address matched host_lookup and the reverse lookup yielded a name
1482 that did not have a forward lookup, an error message of the form "no IP
1483 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1484 could be logged. Now it outputs the IP address instead of "NULL".
1486 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1487 allows the sender and the authenticated sender to be set when
1488 submitting a message from within Exim. Since child_open_exim() is
1489 documented for local_scan(), the new function should be too.
1491 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1492 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1493 results in an empty string is now treated as unset.
1495 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1497 PH/31 Added sender_verify_fail logging option.
1499 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1500 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1501 tidied the source and removed it altogether.
1503 PH/33 When a queue run was abandoned because the load average was too high, a
1504 log line was always written; now it is written only if the queue_run log
1505 selector is set. In addition, the log line for abandonment now contains
1506 information about the queue run such as the pid. This is always present
1507 in "start" and "stop" lines but was omitted from the "abandon" line.
1509 PH/34 Omit spaces between a header name and the colon in the error message that
1510 is given when verify = headers_syntax fails (if there are lots of them,
1511 the message gets confusing).
1513 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1514 names, as there are now some PTR records that contain slashes. This check
1515 is only to protect against broken name servers that fall over on strange
1516 characters, so the fact that it applies to all lookups doesn't matter.
1518 PH/36 Now that the new test suite is complete, we can remove some of the
1519 special code in Exim that was needed for the old test suite. For example,
1520 sorting DNS records because real resolvers return them in an arbitrary
1521 order. The new test suite's fake resolver always returns records in the
1524 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1525 bounces) except when queue_only is set, to avoid logging races between
1526 the different processes.
1528 PH/38 Panic-die if .include specifies a non-absolute path.
1530 PH/39 A tweak to the "H" retry rule from its user.
1532 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1533 a label. They prevented compilation on older perls.
1535 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1536 a warning to be raised on newish perls.
1538 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1539 on queue. Changes to match documented behaviour of showing count of
1540 messages matching specified criteria.
1542 PH/40 Changed the default ident timeout from 30s to 5s.
1544 PH/41 Added support for the use of login_cap features, on those BSD systems
1545 that have them, for controlling the resources used by pipe deliveries.
1547 PH/42 The content-scanning code uses fopen() to create files in which to put
1548 message data. Previously it was not paying any attention to the mode of
1549 the files. Exim runs with umask(0) because the rest of the code creates
1550 files with open(), and sets the required mode explicitly. Thus, these
1551 files were ending up world-writeable. This was not a big issue, because,
1552 being within the spool directory, they were not world-accessible. I have
1553 created a function called modefopen, which takes an additional mode
1554 argument. It sets umask(777), creates the file, chmods it to the required
1555 mode, then resets the umask. All the relevant calls to fopen() in the
1556 content scanning code have been changed to use this function.
1558 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1559 to 24 hours. This avoids potential overflow problems when processing G
1560 and H retry rules. I suspect nobody ever tinkers with this value.
1562 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1564 PH/45 When the plaintext authenticator is running as a client, the server's
1565 challenges are checked to ensure they are valid base64 strings. By
1566 default, the authentication attempt is cancelled if an invalid string is
1567 received. Setting client_ignore_invalid_base64 true ignores these errors.
1568 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1569 they are received. Thus, the responses can be made to depend on the
1570 challenges. If an invalid string is ignored, an empty string is placed in
1573 PH/46 Messages that are created by the autoreply transport now contains a
1574 References: header, in accordance with RFCs 2822 and 3834.
1576 PH/47 Added authenticated_sender_force to the smtp transport.
1578 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1580 PH/49 Installed latest patch from the Sieve maintainer.
1582 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1583 was also set, the appendfile transport was unnecessarily scanning a
1584 directory of message files (e.g. for maildir delivery) to find the count
1585 of files (along with the size), even though it did not need this
1586 information. It now does the scan only if it needs to find either the
1587 size of the count of files.
1589 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1591 PH/52 Two bugs concerned with error handling when the smtp transport is
1594 (i) Exim was not creating retry information for temporary errors given
1595 for individual recipients after the DATA command when the smtp transport
1596 was used in LMTP mode. This meant that they could be retried too
1597 frequently, and not timed out correctly.
1599 (ii) Exim was setting the flag that allows error details to be returned
1600 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1601 recipients that were returned after the DATA command.
1603 PH/53 This is related to PH/52, but is more general: for any failing address,
1604 when detailed error information was permitted to be returned to the
1605 sender, but the error was temporary, then after the final timeout, only
1606 "retry timeout exceeded" was returned. Now it returns the full error as
1607 well as "retry timeout exceeded".
1609 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1610 do this, and (what is worse) MTAs that accept it.
1612 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1613 will now be deprecated.
1615 PH/56 New os.c-cygwin from the Cygwin maintainer.
1617 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1618 formats (previously only available in exim formats via -bpr, -bpru,
1619 and -bpra. Now also available in native and exiqgrep formats)
1621 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1622 with very large, slow to parse queues
1624 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1626 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1628 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1629 responses to authentication challenges, though it was showing the
1630 challenges; (ii) I've removed the CR characters from the debug output for
1633 PH/58 Allow for the insertion of a newline as well as a space when a string
1634 is turned into more than one encoded-word during RFC 2047 encoding. The
1635 Sieve code now uses this.
1637 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1638 data_4xx, lost_connection, tls_required.
1640 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1641 message was being sent as an SMTP response.
1643 PH/61 Add -l and -k options to exicyclog.
1645 PH/62 When verifying, if an address was redirected to one new address, so that
1646 verification continued, and the new address failed or deferred after
1647 having set something in $address_data, the value of $address_data was not
1648 passed back to the ACL. This was different to the case when no
1649 redirection occurred. The value is now passed back in both cases.
1651 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1652 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1653 use login_cap.h, so on its own it isn't the distinguishing feature. The
1654 new name refers directly to the setclassresources() function.
1656 PH/65 Added configuration files for NetBSD3.
1658 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1660 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1661 is preferred over IPv4.
1663 PH/68 The bounce_return_message and bounce_return_body options were not being
1664 honoured for bounces generated during the reception of non-SMTP messages.
1665 In particular, this applied to messages rejected by the ACL. This bug has
1666 been fixed. However, if bounce_return_message is true and bounce_return_
1667 body is false, the headers that are returned for a non-SMTP message
1668 include only those that have been read before the error was detected.
1669 (In the case of an ACL rejection, they have all been read.)
1671 PH/69 The HTML version of the specification is now built in a directory called
1672 spec_html instead of spec.html, because the latter looks like a path with
1673 a MIME-type, and this confuses some software.
1675 PH/70 Catch two compiler warnings in sieve.c.
1677 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1678 function verify_get_ident() calls ip_connect() to connect a socket, but
1679 if the "connect()" function timed out, ip_connect() used to close the
1680 socket. However, verify_get_ident() also closes the socket later, and in
1681 between Exim writes to the log, which may get opened at this point. When
1682 the socket was closed in ip_connect(), the log could get the same file
1683 descriptor number as the socket. This naturally causes chaos. The fix is
1684 not to close the socket in ip_connect(); the socket should be closed by
1685 the function that creates it. There was only one place in the code where
1686 this was missing, in the iplookup router, which I don't think anybody now
1687 uses, but I've fixed it anyway.
1689 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1690 well as to direct DNS lookups. Otherwise the handling of names in host
1691 lists is inconsistent and therefore confusing.
1697 PH/01 Two changes to the default runtime configuration:
1699 (1) Move the checks for relay_from_hosts and authenticated clients from
1700 after to before the (commented out) DNS black list checks.
1702 (2) Add control=submission to the relay_from_hosts and authenticated
1703 clients checks, on the grounds that messages accepted by these
1704 statements are most likely to be submissions.
1706 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1708 (1) Generate an error if the third argument for the ${prvs expansion is
1711 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1714 (3) Reset the variables that are obtained from the first argument of
1715 ${prvscheck and used in the second argument before leaving the code,
1716 because their memory is reclaimed, so using them afterwards may do
1719 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1720 one (it's much easier than Tom thought :-).
1722 (5) Because of (4), we can now allow for the use of $prvscheck_result
1723 inside the third argument.
1725 PH/03 For some reason, the default setting of PATH when running a command from
1726 a pipe transport was just "/usr/bin". I have changed it to
1729 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1730 anything to be listed in the output from -bV.
1732 PH/05 When a filter generated an autoreply, the entire To: header line was
1733 quoted in the delivery log line, like this:
1735 => >A.N.Other <ano@some.domain> <original@ddress> ...
1737 This has been changed so that it extracts the operative address. There
1738 may be more than one such address. If so, they are comma-separated, like
1741 => >ano@some.domain,ona@other.domain <original@ddress> ...
1743 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1744 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1745 not being looked up in the rDNS to get a host name, Exim was showing the
1746 IP address twice in Received: lines, even though the IP addresses were
1747 identical. For example:
1749 Received: from [1.2.3.4] (helo=[1.2.3.4])
1751 However, if the real host name was known, it was omitting the HELO data
1752 if it matched the actual IP address. This has been tidied up so that it
1753 doesn't show the same IP address twice.
1755 PH/07 When both +timestamp and +memory debugging was on, the value given by
1756 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1757 called by the debug printing, thereby overwriting the timestamp buffer.
1758 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1760 PH/08 When the original message was included in an autoreply transport, it
1761 always said "this is a copy of the message, including all the headers",
1762 even if body_only or headers_only was set. It now gives an appropriate
1765 PH/09 Applied a patch from the Sieve maintainer which:
1767 o fixes some comments
1768 o adds the (disabled) notify extension core
1769 o adds some debug output for the result of if/elsif tests
1770 o points to the current vacation draft in the documentation
1771 and documents the missing references header update
1775 o fixes a bug in processing the envelope test (when testing
1776 multiple envelope elements, the last element determinted the
1779 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1780 Electronic Mail") by including:
1782 Auto-submitted: auto-generated
1784 in the messages that it generates (bounce messages and others, such as
1785 warnings). In the case of bounce messages for non-SMTP mesages, there was
1786 also a typo: it was using "Auto_submitted" (underscore instead of
1787 hyphen). Since every message generated by Exim is necessarily in response
1788 to another message, thes have all been changed to:
1790 Auto-Submitted: auto-replied
1792 in accordance with these statements in the RFC:
1794 The auto-replied keyword:
1796 - SHOULD be used on messages sent in direct response to another
1797 message by an automatic process,
1799 - MUST NOT be used on manually-generated messages,
1801 - MAY be used on Delivery Status Notifications (DSNs) and Message
1802 Disposition Notifications (MDNs),
1804 - MUST NOT be used on messages generated by automatic or periodic
1805 processes, except for messages which are automatic responses to
1808 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1809 to the default Received: header definition.
1811 PH/12 Added log selector acl_warn_skipped (default on).
1813 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1814 variables because (a) they are in the wrong storage pool and (b) even if
1815 they were copied, it wouldn't work properly because of the caching.
1817 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1818 checking when decoding. Apparently there are clients that generate
1819 overlong encoded strings. Why am I not surprised?
1821 PH/15 If the first argument of "${if match_address" was not empty, but did not
1822 contain an "@" character, Exim crashed. Now it writes a panic log message
1823 and treats the condition as false.
1825 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1827 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1828 extension "envelope-auth". The code is finished and in agreement with
1829 other implementations, but there is no documentation so far and in fact,
1830 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1831 not changing the active code.
1833 Print executed "if" and "elsif" statements when debugging is used. This
1834 helps a great deal to understand what a filter does.
1836 Document more things not specified clearly in RFC3028. I had all this
1837 sorted out, when out of a sudden new issues came to my mind. Oops."
1839 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1842 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1843 canonical form (as documented). However, after a host name lookup from
1844 the IP address, check_host() was doing a simple string comparison with
1845 addresses acquired from the DNS when checking that the found name did
1846 have the original IP as one of its addresses. Since any found IPv6
1847 addresses are likely to be in abbreviated form, the comparison could
1848 fail. Luckily, there already exists a function for doing the comparison
1849 by converting both addresses to binary, so now that is used instead of
1850 the text comparison.
1852 PH/20 There was another similar case to PH/19, when a complete host name was
1853 given in a host list; looking up its IP address could give an abbreviated
1854 form, whereas the current host's name might or might not be abbreviated.
1855 The same fix has been applied.
1861 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1862 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1865 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1866 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1868 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1870 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1871 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1872 Also, for "Auto-Submitted", treat anything other than "no" as
1873 non-personal, in accordance with RFC 3834. (Previously it treated
1874 anything starting "auto-" as non-personal.)
1876 TF/01 The control=submission/name=... option had a problem with syntax
1877 errors if the name included a slash character. The /name= option
1878 now slurps the rest of the string, so it can include any characters
1879 but it must come last in the list of options (after /sender_retain
1882 PH/05 Some modifications to the interface to the fake nameserver for the new
1890 TK/01 Added the "success_on_redirect" address verification option. See
1891 NewStuff for rationale and an example.
1893 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1895 PH/02 Patch to exigrep to allow it to work on syslog lines.
1897 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1898 fread() to skip over the body file's header line, because in Cygwin the
1899 header line is locked and is inaccessible.
1901 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1902 co-exist for some time) to make it clear that it is the Exim ID that is
1903 referenced, not the Message-ID: header line.
1905 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1906 string_format() function, because snprintf() does not exist on all
1909 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1910 ${stat: expansion item.
1912 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1913 protocol synchronization error", to keep the pedants happy.
1915 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1916 well as for IRIX systems, when gcc is being used. See the host.c source
1919 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1921 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1924 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1925 transport and to the smtp transport in LMTP mode.
1927 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1929 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1930 run a filter in a subprocess. This could lead to confusion in subsequent
1931 lookups in the parent process. There should also be a search_tidyup() at
1932 the end of the subprocess.
1934 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1935 only if the host matched helo_try_verify_hosts, which caused the
1936 verification to occur when the EHLO/HELO command was issued. The ACL just
1937 tested the remembered result. Now, if a previous verification attempt has
1938 not happened, "verify = helo" does it there and then.
1940 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1942 TK/03 Fix log output including CR from clamd.
1944 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1945 exist provoked a memory error which could cause a segfault.
1947 PH/15 Installed PCRE 6.2
1949 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1951 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1952 of the problem. Specifically, suggested +O2 rather than +O1 for the
1955 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1957 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1958 record, the port was not correctly logged when the outgoing_port log
1959 selector was set (it logged the transort's default port).
1961 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1962 fallback_hosts, and "hosts" in the smtp transport.
1964 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1965 host errors such as "Connection refused".
1967 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1968 authentication with radiusclient 0.4.9:
1970 - Error returned from rc_read_config was caught wrongly
1971 - Username/password not passed on to radius server due to wrong length.
1973 The presumption is that some radiusclient API changes for 4.51/PH/17
1974 were not taken care of correctly. The code is still untested by me (my
1975 Linux distribution still has 0.3.2 of radiusclient), but it was
1976 contributed by a Radius user.
1978 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1979 expanding the "port" option of the smtp transport.
1981 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1982 while reading a MIME header. Thanks to Tom Hughes for a patch.
1984 PH/24 Include config.h inside local_scan.h so that configuration settings are
1987 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1988 that in an ACL for RCPT (for example), you can examine exactly what was
1991 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1992 commands, but it was not correctly comparing the address with the actual
1993 client host address. Thus, it would show the EHLO address in Received:
1994 header lines when this was not necessary.
1996 PH/27 Added the % operator to ${eval:}.
1998 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1999 it should be ignoring failures (because with -C, for example, it has lost
2000 privilege). It wasn't ignoring creation failures other than "already
2003 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2006 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2007 because an input line was too long, either on its own, or by virtue of
2008 too many continuations, the temporary file was not being removed, and the
2009 return code was incorrect.
2011 PH/31 Missing "BOOL" in function definition in filtertest.c.
2013 PH/32 Applied Sieve patches from the maintainer.
2015 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2017 PH/33 Added "verify = not_blind".
2019 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2020 Local/Makefile (with some defaults set). These are used in built scripts
2021 such as exicyclog, but they have never been used in the exim_install
2022 script (though there are many overriding facilities there). I have
2023 arranged that the exim_install script now takes note of these two
2026 PH/35 Installed configuration files for Dragonfly.
2028 PH/36 When a locally submitted message by a trusted user did not contain a
2029 From: header, and the sender address was obtained from -f or from an SMTP
2030 MAIL command, and the trusted user did not use -F to supply a sender
2031 name, $originator_name was incorrectly used when constructing a From:
2032 header. Furthermore, $originator_name was used for submission mode
2033 messages from external hosts without From: headers in a similar way,
2034 which is clearly wrong.
2036 PH/37 Added control=suppress_local_fixups.
2038 PH/38 When log_selector = +received_sender was set, and the addition of the
2039 sender made the log line's construction buffer exactly full, or one byte
2040 less than full, an overflow happened when the terminating "\n" was
2043 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2044 when the result of a list match is failure because a DNS lookup failed.
2046 PH/40 RM_COMMAND is now used in the building process.
2048 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2049 the "build-* directories that it finds.
2051 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2052 address in a domain literal was a prefix of an interface address.
2054 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2055 when verifying a sender address, unless rewrite_headers is false.
2057 PH/44 Wrote a long comment about why errors_to addresses are verified as
2058 recipients, not senders.
2060 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2061 the ratelimit ACL was added.
2063 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2065 PH/46 Added extra information about PostgreSQL errors to the error string.
2067 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2068 avoiding the need to install special zones in a real server. This is
2069 backwards compatible; if it can't find the fake resolver, it drops back.
2070 Thus, both old and new test suites can be run.
2072 TF/02 Added util/ratelimit.pl
2074 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2075 clock is set back in time.
2077 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2078 Candler <B.Candler@pobox.com>.
2080 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2081 OK for addresses that are the result of redirections.
2083 PH/48 A number of further additions for the benefit of the new test suite,
2084 including a fake gethostbyname() that interfaces to the fake DNS resolver
2087 TF/06 The fix for widen_domains has also been applied to qualify_single and
2088 search_parents which are the other dnslookup options that can cause
2091 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2094 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2096 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2097 service" if it encountered an MX record with an empty target hostname.
2098 The message is now "an MX or SRV record indicated no SMTP service".
2100 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2101 if the DNS of the sending site is misconfigured. This is quite a
2102 common situation. This change restores the behaviour of treating a
2103 helo verification defer as a failure.
2105 PH/51 If self=fail was set on a router, the bounce message did not include the
2106 actual error message.
2112 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2114 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2115 command itself ended in error, the underlying message about the transport
2116 filter timeout was being overwritten with the pipe command error. Now the
2117 underlying error message should be appended to the second error message.
2119 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2121 PH/02 Reduce the amount of output that "make" produces by default. Full output
2122 can still be requested.
2124 PH/03 The warning log line about a condition test deferring for a "warn" verb
2125 was being output only once per connection, rather than after each
2126 occurrence (because it was using the same function as for successful
2127 "warn" verbs). This seems wrong, so I have changed it.
2129 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2130 it should not have, which might have caused a crash in the right
2131 circumstances, but probably never did.
2133 PH/04 Installed a modified version of Tony Finch's patch to make submission
2134 mode fix the return path as well as the Sender: header line, and to
2135 add a /name= option so that you can make the user's friendly name appear
2138 TF/03 Added the control = fakedefer ACL modifier.
2140 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2141 Mark Lowes for thorough testing.
2143 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2145 TK/03 Merged latest SRS patch from Miles Wilton.
2147 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2148 with the definition in sysexits.h (which is #included earlier).
2149 Fortunately, Exim does not actually use EX_OK. The code used to try to
2150 preserve the sysexits.h value, by assumimg that macro definitions were
2151 scanned for macro replacements. I have been disabused of this notion,
2152 so now the code just undefines EX_OK before #including unistd.h.
2154 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2155 in the smtp transport. When a block could not be written in a single
2156 write() function, the timeout was being re-applied to each part-write.
2157 This seems wrong - if the receiver was accepting one byte at a time it
2158 would take for ever. The timeout is now adjusted when this happens. It
2159 doesn't have to be particularly precise.
2161 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2162 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2164 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2165 without a domain if the check to <postmaster@domain> fails.
2167 SC/01 Eximstats: added -xls and the ability to specify output files
2168 (patch written by Frank Heydlauf).
2170 SC/02 Eximstats: use FileHandles for outputing results.
2172 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2174 SC/04 Eximstats: fixed display of large numbers with -nvr option
2176 SC/05 Eximstats: fixed merging of reports with empty tables.
2178 SC/06 Eximstats: added the -include_original_destination flag
2180 SC/07 Eximstats: removed tabs and trailing whitespace.
2182 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2184 TK/06 MBOX spool code: Add real "From " MBOX separator line
2185 so the .eml file is really in mbox format (even though
2186 most programs do not really care). Patch from Alex Miller.
2188 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2189 The latter is generated from $received_to and is only set if the
2190 message has one envelope recipient. SA can use these headers,
2191 obviously out-of-the-box. Patch from Alex Miller.
2193 PH/08 The ${def test on a variable was returning false if the variable's
2194 value was "0", contrary to what the specification has always said!
2195 The result should be true unless the variable is empty.
2197 PH/09 The syntax error of a character other than { following "${if
2198 def:variable_name" (after optional whitespace) was not being diagnosed.
2199 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2200 accidental colon was present, for example, could give incorrect results.
2202 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2203 result is used (not including appendfile, where other changes are about
2206 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2207 This involved changing a lot of size variables from int to off_t. It
2208 should work with maildirs and everything.
2210 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2211 spamd dying while we are connected to it.
2213 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2216 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2217 function for BDB 4.3.
2219 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2221 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2222 the sources for PCRE have been re-arranged and more files are now
2225 PH/15 The code I had for printing potentially long long variables in PH/11
2226 above was not the best (it lost precision). The length of off_t variables
2227 is now inspected at build time, and an appropriate printing format (%ld
2228 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2229 to be "long long int" or "long int". This is needed for the internal
2230 formatting function string_vformat().
2232 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2233 the configuration file to be ":syslog", then the script "guesses" where
2234 the logs files are, rather than using the compiled in default. In our
2235 case the guess is not the same as the compiled default, so the script
2236 suddenly stopped working when I started to use syslog. The patch checks
2237 to see if log_file_path is "". If so, it attempts to read it from exim
2238 with no configuration file to get the compiled in version, before it
2239 falls back to the previous guessing code."
2241 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2242 implementing BATV in an Exim configuration. See NewStuff for the gory
2245 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2246 Makefile that are specific to HP-UX.
2248 PH/18 If the "use_postmaster" option was set for a recipient callout together
2249 with the "random" option, the postmaster address was used as the MAIL
2250 FROM address for the random test, but not for the subsequent recipient
2251 test. It is now used for both.
2253 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2254 patch removes a few documentation additions to RFC 3028, because the
2255 latest draft now contains them. It adds the new en;ascii-case comparator
2256 and a new error check for 8bit text in MIME parts. Comparator and
2257 require names are now matched exactly. I enabled the subaddress
2258 extension, but it is not well tested yet (read: it works for me)."
2260 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2261 rework some of the code of TK/09 above to avoid the hardwired use of
2262 "%lld" and "long long". Replaced the call to snprintf() with a call to
2265 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2266 records point to non-existent hosts", "retry timeout exceeded", and
2267 "retry time not reached for any host after a long failure period".
2269 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2270 experimental DomainKeys support:
2272 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2273 (2) On an error such as an illegally used "control", the wrong name for
2274 the control was given.
2276 These problems did NOT occur unless DomainKeys support was compiled.
2278 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2280 PH/24 Added ${if match_ip condition.
2282 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2283 databases so that it will be absolutely obvious if a crash occurs in the
2284 DB library. This is a regular occurrence (often caused by mis-matched
2287 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2288 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2289 picked up on a user's system that detects such things. There doesn't seem
2290 to be a gcc warning option for this - only an attribute that has to be
2291 put on the function's prototype. It seems that in Fedora Core 4 they have
2292 set this on a number of new functions. No doubt there will be more in due
2295 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2296 specify a transport. However, if an address that was verified by such a
2297 router was the subject of a callout, Exim crashed because it tried to
2298 read the rcpt_include_affixes from the non-existent transport. Now it
2299 just assumes that the setting of that option is false. This bug was
2300 introduced by 4.51/PH/31.
2302 PH/28 Changed -d+all to exclude +memory, because that information is very
2303 rarely of interest, but it makes the output a lot bigger. People tend to
2304 do -d+all out of habit.
2306 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2307 code in os-type was giving problems when libc.so lives in lib64, like on
2310 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2311 aren't the modern standard, and it seems that some systems' include files
2312 don't always have them. Exim was already checking for some of the newer
2313 ones like T_AAAA, and defining it itself. I've added checks for all the
2314 record types that Exim uses.
2316 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2317 not automatically generating a new one, as it is supposed to. This
2318 prevented TLS from working. If the file did exist, but contained invalid
2319 data, a new version was generated, as expected. It was only the case of a
2320 non-existent file that was broken.
2322 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2323 with a change in libdomainkeys > 0.64.
2325 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2326 from DNS. If the selector record carries the flag, it now has
2327 precedence over the domain-wide flag.
2329 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2331 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2332 the use of a port name that isn't defined in /etc/services) occurred, the
2333 message was deferred as in a normal delivery, and thus remained on the
2334 spool, instead of being failed because of the mua_wrapper setting. This
2335 is now fixed, and I tidied up some of the mua_wrapper messages at the
2338 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2339 the messages in a hash of arrays rather than using individual hashes.
2340 This is a bit cleaner and results in dramatic memory savings, albeit
2341 at a slight CPU cost.
2343 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2344 as requested by Marc Sherman.
2346 SC/10 Eximstats: added histograms for user specified patterns as requested
2349 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2351 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2352 fopen() in the content-scanning modules that did not already have it.
2358 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2359 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2361 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2363 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2365 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2366 to test Sieve filters that use "vacation".
2368 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2369 that changes the way the GnuTLS parameters are stored in the cache file.
2370 The new format can be generated externally. For backward compatibility,
2371 if the data in the cache doesn't make sense, Exim assumes it has read an
2372 old-format file, and it generates new data and writes a new file. This
2373 means that you can't go back to an older release without removing the
2376 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2377 work if there are any delivery delays because "one_time" forces the
2378 parent to be marked "delivered", so its unseen clone is never tried
2379 again. For this reason, Exim now forbids the simultaneous setting of
2382 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2383 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2384 entry if you want to know the details. The fix, however, neglected to
2385 consider the case when local delivery batching is involved. The test for
2386 "previously delivered" was not happening when checking to see if an
2387 address could be batched with a previous (undelivered) one; under
2388 certain circumstances this could lead to multiple deliveries to the same
2391 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2392 in its include files, and this causes problems building Exim.
2394 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2395 header_syntax) but Exim was just ignoring anything given after a slash.
2396 In particular, this caused confusion with an attempt to use "verify =
2397 reverse_host_lookup/defer_ok". An error is now given when options are
2398 supplied for verify items that do not have them. (Maybe reverse_host_
2399 lookup should have a defer_ok option, but that's a different point.)
2401 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2402 defined by RFC 821) to 2048, because there were problems with some AUTH
2403 commands, and RFC 1869 says the size should be increased for extended
2404 SMTP commands that take arguments.
2406 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2409 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2410 "unknown" error; now it says that the functionality isn't in the binary.
2412 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2413 an address' error message when a string expansion fails (syntax or
2414 whatever). Otherwise the password may appear in the log. Following change
2415 PH/42 below, there is no longer a chance of it appearing in a bounce
2418 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2420 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2421 of its addresses was ever tried. (Bugzilla bug #2.)
2423 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2424 the result incorrectly in the debug output. (It correctly added a newline
2425 to what was transported.)
2427 TF/01 Added $received_time.
2429 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2430 commented out examples of how to interface to a virus scanner and to
2431 SpamAssassin. Also added commented examples of av_scanner and
2432 spamd_address settings.
2434 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2435 and controls are allowed in which ACLs. There were a couple of minor
2436 errors. Some of the entries in the conditions table (which is a table of
2437 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2438 negation of where the condition IS allowed.
2440 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2442 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2443 header file does not have a version number, so I've had to invent a new
2444 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2445 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2446 radiusclient), but it was contributed by a Radius user.
2448 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2449 files or directories, for OpenSSL.
2451 PH/19 When an Exim process that is running as root has to create an Exim log
2452 file, it does so in a subprocess that runs as exim:exim so as to get the
2453 ownership right at creation (otherwise, other Exim processes might see
2454 the file with the wrong ownership). There was no test for failure of this
2455 fork() call, which would lead to the process getting stuck as it waited
2456 for a non-existent subprocess. Forks do occasionally fail when resources
2457 run out. I reviewed all the other calls to fork(); they all seem to check
2460 PH/20 When checking for unexpected SMTP input at connect time (before writing
2461 the banner), Exim was not dealing correctly with a non-positive return
2462 from the read() function. If the client had disconnected by this time,
2463 the result was a log entry for a synchronization error with an empty
2464 string after "input=" when read() returned zero. If read() returned -1
2465 (an event I could not check), uninitialized data bytes were printed.
2466 There were reports of junk text (parts of files, etc) appearing after
2469 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2471 PH/22 Added support for macro redefinition, and (re)definition in between
2472 driver and ACL definitions.
2474 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2475 forgetting to use the resulting value; it was using the unexpanded value.
2477 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2478 hadn't been configured. The fix is from Juergen Kreileder, who
2479 understands it better than I do:
2481 "Here's what I see happening with three configured cyrus_sasl
2482 authenticators configured (plain, login, cram-md5):
2484 On startup auth_cyrus_sasl_init() gets called for each of these.
2485 This means three calls to sasl_listmech() without a specified mech_list.
2486 => SASL tests which mechs of all available mechs actually work
2487 => three warnings about OTP not working
2488 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2490 With the patch, sasl_listmech() also gets called three times. But now
2491 SASL's mech_list option is set to the server_mech specified in the the
2492 authenticator. Or in other words, the answer from sasl_listmech()
2493 gets limited to just the mech you're testing for (which is different
2495 => the return list contains just 'plain' or 'login', 'cram-md5' or
2496 nothing depending on the value of ob->server_mech.
2498 I've just tested the patch: Authentication still works fine,
2499 unavailable mechs specified in the exim configuration are still
2500 caught, and the auth.log warnings about OTP are gone."
2502 PH/25 When debugging is enabled, the contents of the command line are added
2503 to the debugging output, even when log_selector=+arguments is not
2506 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2507 answer is "GNU", and only if the return is "GNU/something" is the answer
2510 PH/27 $acl_verify_message is now set immediately after the failure of a
2511 verification in an ACL, and so is available in subsequent modifiers. In
2512 particular, the message can be preserved by coding like this:
2514 warn !verify = sender
2515 set acl_m0 = $acl_verify_message
2517 Previously, $acl_verify_message was set only while expanding "message"
2518 and "log_message" when a very denied access.
2520 PH/28 Modified OS/os.c-Linux with
2522 -#ifndef OS_LOAD_AVERAGE
2523 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2525 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2526 nomenclature these days.)
2528 PH/29 Installed patch from the Sieve maintainer that adds the options
2529 sieve_useraddress and sieve_subaddress to the redirect router.
2531 PH/30 In these circumstances:
2532 . Two addresses routed to the same list of hosts;
2533 . First host does not offer TLS;
2534 . First host accepts first address;
2535 . First host gives temporary error to second address;
2536 . Second host offers TLS and a TLS session is established;
2537 . Second host accepts second address.
2538 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2539 and peerdn, if requested) that were in fact used only for the second
2542 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2543 attention to any local part prefix or suffix that was matched by the
2544 router that accepted the address. It now behaves in the same way as it
2545 does for delivery: the affixes are removed from the local part unless
2546 rcpt_include_affixes is set on the transport.
2548 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2549 timeout during the DATA phase of an incoming message.
2551 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2552 applied a patch sanctioned by the Sieve maintainer.
2554 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2555 the uid or gid is negative. A case of a negative gid caused this to be
2556 noticed. The fix allows for either to be negative.
2558 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2559 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2562 PH/36 The change PH/12 above was broken. Fixed it.
2564 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2565 the grounds that routing the same address twice would always produce the
2566 same answer. This might have been true once, but it is certainly no
2567 longer true now. Routing a child address may depend on the previous
2568 routing that produced that child. Some complicated redirection strategies
2569 went wrong when messages had multiple recipients, and made Exim's
2570 behaviour dependent on the order in which the addresses were given.
2572 I have moved the duplicate checking until after the routing is complete.
2573 Exim scans the addresses that are assigned to local and remote
2574 transports, and removes any duplicates. This means that more work will be
2575 done, as duplicates will always all be routed, but duplicates are
2576 presumably rare, so I don't expect this is of any significance.
2578 For deliveries to pipes, files, and autoreplies, the duplicate checking
2579 still happens during the routing process, since they are not going to be
2582 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2583 It corrects a timeout issue with spamd. This is Ian's comment: "The
2584 background is that sometimes spamd either never reads data from a
2585 connection it has accepted, or it never writes response data. The exiscan
2586 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2587 blindly assumes that writes won't block so it may never time out."
2589 PH/39 Allow G after quota size as well as K and M.
2591 PH/40 The value set for $authenticated_id in an authenticator may not contain
2592 binary zeroes or newlines because the value is written to log lines and
2593 to spool files. There was no check on this. Now the value is run through
2594 the string_printing() function so that such characters are converted to
2595 printable escape sequences.
2597 PH/41 $message_linecount is a new variable that contains the total number of
2598 lines in the message. Compare $body_linecount, which is the count for the
2601 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2602 bounce and delay warning messages, except in certain special cases, which
2605 (a) An SMTP error message from a remote host;
2606 (b) A message specified in a :fail: redirection;
2607 (c) A message specified in a "fail" command in a system filter;
2608 (d) A message specified in a FAIL return from the queryprogram router;
2609 (e) A message specified by the cannot_route_message router option.
2611 In these cases only, Exim does include the error details in bounce and
2612 warning messages. There are also a few cases where bland messages such
2613 as "unrouteable address" or "local delivery error" are given.
2615 PH/43 $value is now also set for the "else" part of a ${run expansion.
2617 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2618 being worked on, but at least Exim now implements the latest version to
2621 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2622 process to complete was treated as a delivery failure, a timeout while
2623 writing the message to the pipe was logged, but erroneously treated as a
2624 successful delivery. Such timeouts include transport filter timeouts. For
2625 consistency with the overall process timeout, these timeouts are now
2626 treated as errors, giving rise to delivery failures by default. However,
2627 there is now a new Boolean option for the pipe transport called
2628 timeout_defer, which, if set TRUE, converts the failures into defers for
2629 both kinds of timeout. A transport filter timeout is now identified in
2632 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2633 systems where "make" and "gmake" are different, calling "gmake" at top
2634 level broke things. I've arranged for the value of $(MAKE) to be passed
2635 from the Makefile to this script so that it can call the same version of
2639 A note about Exim versions 4.44 and 4.50
2640 ----------------------------------------
2642 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2643 changes of various kinds. As a consequence, a big documentation update was
2644 needed. This delayed the release for rather longer than seemed good, especially
2645 in the light of a couple of (minor) security issues. Therefore, the changes
2646 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2647 release. So 4.44 and 4.50 are in effect two different branches that both start
2650 I have left the 4.50 change log unchanged; it contains all the changes since
2651 4.43. The change log for 4.44 is below; many of its items are identical to
2652 those for 4.50. This seems to be the most sensible way to preserve the
2653 historical information.
2659 1. Minor wording change to the doc/README.SIEVE file.
2661 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2662 computation of the current number of files was incorrect.
2664 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2665 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2668 4. Give more explanation in the error message when the command for a transport
2669 filter fails to execute.
2671 5. There are several places where Exim runs a non-Exim command in a
2672 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2673 was being done only for the command run by the queryprogram router. It is
2674 now done for all such subprocesses. The other cases are: ${run, transport
2675 filters, and the commands run by the lmtp and pipe transports.
2677 6. Added CONFIGURE_GROUP build-time option.
2679 7. Some older OS have a limit of 256 on the maximum number of file
2680 descriptors. Exim was using setrlimit() to set 1000 as a large value
2681 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2682 systems. I've change it so that if it can't get 1000, it tries for 256.
2684 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2685 was an oversight, and furthermore, ever since the addition of extra
2686 controls (e.g. 4.43/32), the checks on when to allow different forms of
2687 "control" were broken. There should now be diagnostics for all cases when a
2688 control that does not make sense is encountered.
2690 9. Added the /retain_sender option to "control=submission".
2692 10. $recipients is now available in the predata ACL (oversight).
2694 11. Tidy the search cache before the fork to do a delivery from a message
2695 received from the command line. Otherwise the child will trigger a lookup
2696 failure and thereby defer the delivery if it tries to use (for example) a
2697 cached ldap connection that the parent has called unbind on.
2699 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2700 of $address_data from the recipient verification was clobbered by the
2701 sender verification.
2703 13. The value of address_data from a sender verification is now available in
2704 $sender_address_data in subsequent conditions in the ACL statement.
2706 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2708 15. Added a new option "connect=<time>" to callout options, to set a different
2711 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2712 was its contents. (It was OK if the option was not defined at all.)
2714 17. A "Completed" log line is now written for messages that are removed from
2715 the spool by the -Mrm option.
2717 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2718 information about exactly what failed.
2720 19. Added -dd to debug only the daemon process.
2722 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2723 handles timeouts, both on the server side and network timeouts. Renamed the
2724 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2726 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2727 It is now set to "smtps".
2729 22. $host_address is now set to the target address during the checking of
2730 ignore_target_hosts.
2732 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2733 being passed; this would have caused $sender_host_name to have been used if
2734 matching the list had actually called for a host name (not very likely,
2735 since this list is usually IP addresses). A host name is now passed as
2738 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2739 code that specifies a non-restarting handler (typically sigaction() in
2740 modern systems) in an attempt to fix a rare and obscure crash bug.
2742 25. Narrowed the window for a race in the daemon that could cause it to ignore
2743 SIGCHLD signals. This is not a major problem, because they are used only to
2744 wake it up if nothing else does.
2746 26. A malformed maildirsize file could cause Exim to calculate negative values
2747 for the mailbox size or file count. Odd effects could occur as a result.
2748 The maildirsize information is now recalculated if the size or filecount
2751 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2752 support for a long time. Removed HAVE_SYS_VFS_H.
2754 28. Installed the latest version of exipick from John Jetmore.
2756 29. In an address list, if the pattern was not a regular expression, an empty
2757 subject address (from a bounce message) matched only if the pattern was an
2758 empty string. Non-empty patterns were not even tested. This was the wrong
2759 because it is perfectly reasonable to use an empty address as part of a
2760 database query. An empty address is now tested by patterns that are
2761 lookups. However, all the other forms of pattern expect the subject to
2762 contain a local part and a domain, and therefore, for them, an empty
2763 address still always fails if the pattern is not itself empty.
2765 30. Exim went into a mad DNS loop when attempting to do a callout where the
2766 host was specified on an smtp transport, and looking it up yielded more
2767 than one IP address.
2769 31. Re-factored the code for checking spool and log partition space into a
2770 function that finds that data and another that does the check. The former
2771 is then used to implement four new variables: $spool_space, $log_space,
2772 $spool_inodes, and $log_inodes.
2774 32. The RFC2047 encoding function was originally intended for short strings
2775 such as real names; it was not keeping to the 75-character limit for
2776 encoded words that the RFC imposes. It now respects the limit, and
2777 generates multiple encoded words if necessary. To be on the safe side, I
2778 have increased the buffer size for the ${rfc2047: expansion operator from
2781 33. It is now permitted to omit both strings after an "if" condition; if the
2782 condition is true, the result is "true". As before, when the second string
2783 is omitted, a false condition yields an empty string. This makes it less
2784 cumbersome to write custom ACL and router conditions.
2786 34. Failure to deliver a bounce message always caused it to be frozen, even if
2787 there was an errors_to setting on the router. The errors_to setting is now
2790 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2791 canonical form (fully expanded) before being placed in
2792 $sender_host_address.
2794 36. The table in the code that translates DNS record types into text (T_A to
2795 "A" for instance) was missing entries for NS and CNAME. It is just possible
2796 that this could have caused confusion if both these types were looked up
2797 for the same domain, because the text type is used as part of Exim's
2798 per-process caching. But the chance of anyone hitting this buglet seems
2801 37. The dnsdb lookup has been extended in a number of ways.
2803 (1) There is a new type, "zns", which walks up the domain tree until it
2804 finds some nameserver records. It should be used with care.
2806 (2) There is a new type, "mxh", which is like "mx" except that it returns
2807 just the host names, not the priorities.
2809 (3) It is now possible to give a list of domains (or IP addresses) to be
2810 looked up. The behaviour when one of the lookups defers can be
2811 controlled by a keyword.
2813 (4) It is now possible to specify the separator character for use when
2814 multiple records are returned.
2816 38. The dnslists ACL condition has been extended: it is now possible to supply
2817 a list of IP addresses and/or domains to be looked up in a particular DNS
2820 39. Added log_selector=+queue_time_overall.
2822 40. When running the queue in the test harness, wait just a tad after forking a
2823 delivery process, to get repeatability of debugging output.
2825 41. Include certificate and key file names in error message when GnuTLS fails
2826 to set them up, because the GnuTLS error message doesn't include the name
2827 of the failing file when there is a problem reading it.
2829 42. Allow both -bf and -bF in the same test run.
2831 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2833 44. The "Exiscan patch" is now merged into the mainline Exim source.
2835 45. Sometimes the final signoff response after QUIT could fail to get
2836 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2837 < 0 before doing a fflush(). This bug looks as though it goes back to the
2838 introduction of TLS in release 3.20, but "sometimes" must have been rare
2839 because the tests only now provoked it.
2841 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2842 (this can affect the format of dates).
2844 47. exim_tidydb, when checking for the continued existence of a message for
2845 which it has found a message-specific retry record, was not finding
2846 messages that were in split spool directories. Consequently, it was
2847 deleting retry records that should have stayed in existence.
2849 48. Steve fixed some bugs in eximstats.
2851 49. The SPA authentication driver was not abandoning authentication and moving
2852 on to the next authenticator when an expansion was forced to fail,
2853 contradicting the general specification for all authenticators. Instead it
2854 was generating a temporary error. It now behaves as specified.
2856 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2857 (the order specifies the preference for clients). The order is now AES256,
2858 AES128, 3DES, ARCFOUR128.
2860 51. Small patch to Sieve code - explicitly set From: when generating an
2863 52. Exim crashed if a remote delivery caused a very long error message to be
2864 recorded - for instance if somebody sent an entire SpamAssassin report back
2865 as a large number of 550 error lines. This bug was coincidentally fixed by
2866 increasing the size of one of Exim's internal buffers (big_buffer) that
2867 happened as part of the Exiscan merge. However, to be on the safe side, I
2868 have made the code more robust (and fixed the comments that describe what
2871 53. Now that there can be additional text after "Completed" in log lines (if
2872 the queue_time_overall log selector is set), a one-byte patch to exigrep
2873 was needed to allow it to recognize "Completed" as not the last thing in
2876 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2877 patch that reportedly fixes this has been added. I am not expert enough to
2878 create a test for it. This is what the patch creator wrote:
2880 "I found a little strange behaviour of ldap code when working with
2881 Windows 2003 AD Domain, where users was placed in more than one
2882 Organization Units. When I tried to give exim partial DN, the exit code
2883 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2884 But simultaneously result of request was absolutely normal ldap result,
2885 so I produce this patch..."
2887 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2888 REFERENCE, so I have modified the code to exclude the patch when that macro
2891 55. Some experimental protocols are using DNS PTR records for new purposes. The
2892 keys for these records are domain names, not reversed IP addresses. The
2893 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2894 leaves it alone. Component reversal etc. now happens only for IP addresses.
2897 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2899 57. Double the size of the debug message buffer (to 2048) so that more of very
2900 long debug lines gets shown.
2902 58. The exicyclog utility now does better if the number of log files to keep
2903 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2905 59. Two changes related to the smtp_active_hostname option:
2907 (1) $smtp_active_hostname is now available as a variable.
2908 (2) The default for smtp_banner uses $smtp_active_hostname instead
2909 of $primary_hostname.
2911 60. The host_aton() function is supposed to be passed a string that is known
2912 to be a valid IP address. However, in the case of IPv6 addresses, it was
2913 not checking this. This is a hostage to fortune. Exim now panics and dies
2914 if the condition is not met. A case was found where this could be provoked
2915 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2916 components; fortuitously, this particular loophole had already been fixed
2917 by change 4.50/55 above.
2919 If there are any other similar loopholes, the new check in host_aton()
2920 itself should stop them being exploited. The report I received stated that
2921 data on the command line could provoke the exploit when Exim was running as
2922 exim, but did not say which command line option was involved. All I could
2923 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2924 running as the user.
2927 61. There was a buffer overflow vulnerability in the SPA authentication code
2928 (which came originally from the Samba project). I have added a test to the
2929 spa_base64_to_bits() function which I hope fixes it.
2932 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2933 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2935 63. The daemon start-up calls getloadavg() while still root for those OS that
2936 need the first call to be done as root, but it missed one case: when
2937 deliver_queue_load_max is set with deliver_drop_privilege. This is
2938 necessary for the benefit of the queue runner, because there is no re-exec
2939 when deliver_drop_privilege is set.
2941 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2942 This has been fixed.
2944 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2945 was in use, was not putting the data itself into the right store pool;
2946 consequently, it could be overwritten for a subsequent message in the same
2947 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2950 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2952 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2953 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2954 the function were treating the return as a boolean value, which happened to
2955 work because 0=false and not-0=true, but is not correct code.
2957 68. The host_aton() function was not handling scoped IPv6 addresses (those
2958 with, for example, "%eth0" on the end) correctly.
2960 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2961 negated items (that is, ~something) in unsigned ints. Some compilers
2962 apparently mutter when there is no cast.
2964 70. If an address verification called from an ACL failed, and did not produce a
2965 user-specific message (i.e. there was only a "system" message), nothing was
2966 put in $acl_verify_message. In this situation, it now puts the system
2969 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2970 session; change 4.31/43 added the unwanted input to the log line - except
2971 that it did not do this in the start of session case. It now does.
2973 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2974 This is wrong and can cause the other end to generate a synchronization
2975 error if it is another Exim or anything else that does the synchronization
2976 check. A QUIT command is no longer sent after a timeout.
2978 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2979 during host lookups.
2981 74. The defer_ok option of callout verification was not working if it was used
2982 when verifying addresses in header lines, that is, for this case:
2984 verify = header_sender/callout=defer_ok
2986 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2987 those file descriptors could be used for SMTP connections. If anything
2988 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2989 could be sent to the SMTP client, causing chaos. The daemon now opens
2990 stdin, stdout, and stderr to /dev/null when it puts itself into the
2993 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2994 log by default. The user can override this with suitable Perl magic.
2996 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2997 the log message when discard triggers, was not working for the DATA ACL or
2998 for the non-SMTP ACL.
3000 78. Error message wording change in sieve.c.
3002 79. If smtp_accept_max_per_host was set, the number of connections could be
3003 restricted to fewer than expected, because the daemon was trying to set up
3004 a new connection before checking whether the processes handling previous
3005 connections had finished. The check for completed processes is now done
3006 earlier. On busy systems, this bug wouldn't be noticed because something
3007 else would have woken the daemon, and it would have reaped the completed
3010 80. If a message was submitted locally by a user whose login name contained one
3011 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3012 It caused a spool format error. I have fixed the spool reading code. A
3013 related problem was that the "from" clause in the Received: line became
3014 illegal because of the space(s). It is now covered by ${quote_local_part.
3016 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3019 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3021 83. Patch from Sieve maintainer to fix unterminated string problem in
3022 "vacation" handling.
3024 84. Some minor changes to the Linux configuration files to help with other
3025 OS variants using glibc.
3027 85. One more patch for Sieve to update vacation handling to latest spec.
3030 ----------------------------------------------------
3031 See the note above about the 4.44 and 4.50 releases.
3032 ----------------------------------------------------
3038 1. Change 4.43/35 introduced a bug that caused file counts to be
3039 incorrectly computed when quota_filecount was set in an appendfile
3042 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3043 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3046 3. Give more explanation in the error message when the command for a transport
3047 filter fails to execute.
3049 4. There are several places where Exim runs a non-Exim command in a
3050 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3051 was being done only for the command run by the queryprogram router. It is
3052 now done for all such subprocesses. The other cases are: ${run, transport
3053 filters, and the commands run by the lmtp and pipe transports.
3055 5. Some older OS have a limit of 256 on the maximum number of file
3056 descriptors. Exim was using setrlimit() to set 1000 as a large value
3057 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3058 systems. I've change it so that if it can't get 1000, it tries for 256.
3060 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3061 was an oversight, and furthermore, ever since the addition of extra
3062 controls (e.g. 4.43/32), the checks on when to allow different forms of
3063 "control" were broken. There should now be diagnostics for all cases when a
3064 control that does not make sense is encountered.
3066 7. $recipients is now available in the predata ACL (oversight).
3068 8. Tidy the search cache before the fork to do a delivery from a message
3069 received from the command line. Otherwise the child will trigger a lookup
3070 failure and thereby defer the delivery if it tries to use (for example) a
3071 cached ldap connection that the parent has called unbind on.
3073 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3074 of $address_data from the recipient verification was clobbered by the
3075 sender verification.
3077 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3078 was its contents. (It was OK if the option was not defined at all.)
3080 11. A "Completed" log line is now written for messages that are removed from
3081 the spool by the -Mrm option.
3083 12. $host_address is now set to the target address during the checking of
3084 ignore_target_hosts.
3086 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3087 being passed; this would have caused $sender_host_name to have been used if
3088 matching the list had actually called for a host name (not very likely,
3089 since this list is usually IP addresses). A host name is now passed as
3092 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3093 code that specifies a non-restarting handler (typically sigaction() in
3094 modern systems) in an attempt to fix a rare and obscure crash bug.
3096 15. Narrowed the window for a race in the daemon that could cause it to ignore
3097 SIGCHLD signals. This is not a major problem, because they are used only to
3098 wake it up if nothing else does.
3100 16. A malformed maildirsize file could cause Exim to calculate negative values
3101 for the mailbox size or file count. Odd effects could occur as a result.
3102 The maildirsize information is now recalculated if the size or filecount
3105 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3106 support for a long time. Removed HAVE_SYS_VFS_H.
3108 18. Updated exipick to current release from John Jetmore.
3110 19. Allow an empty sender to be matched against a lookup in an address list.
3111 Previously the only cases considered were a regular expression, or an
3114 20. Exim went into a mad DNS lookup loop when doing a callout where the
3115 host was specified on the transport, if the DNS lookup yielded more than
3118 21. The RFC2047 encoding function was originally intended for short strings
3119 such as real names; it was not keeping to the 75-character limit for
3120 encoded words that the RFC imposes. It now respects the limit, and
3121 generates multiple encoded words if necessary. To be on the safe side, I
3122 have increased the buffer size for the ${rfc2047: expansion operator from
3125 22. Failure to deliver a bounce message always caused it to be frozen, even if
3126 there was an errors_to setting on the router. The errors_to setting is now
3129 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3130 canonical form (fully expanded) before being placed in
3131 $sender_host_address.
3133 24. Updated eximstats to version 1.33
3135 25. Include certificate and key file names in error message when GnuTLS fails
3136 to set them up, because the GnuTLS error message doesn't include the name
3137 of the failing file when there is a problem reading it.
3139 26. Expand error message when OpenSSL has problems setting up cert/key files.
3142 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3143 (this can affect the format of dates).
3145 28. exim_tidydb, when checking for the continued existence of a message for
3146 which it has found a message-specific retry record, was not finding
3147 messages that were in split spool directories. Consequently, it was
3148 deleting retry records that should have stayed in existence.
3150 29. eximstats updated to version 1.35
3151 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3152 1.35 - bugfix such that pie charts by volume are generated correctly
3154 30. The SPA authentication driver was not abandoning authentication and moving
3155 on to the next authenticator when an expansion was forced to fail,
3156 contradicting the general specification for all authenticators. Instead it
3157 was generating a temporary error. It now behaves as specified.
3159 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3160 (the order specifies the preference for clients). The order is now AES256,
3161 AES128, 3DES, ARCFOUR128.
3163 31. Small patch to Sieve code - explicitly set From: when generating an
3166 32. Exim crashed if a remote delivery caused a very long error message to be
3167 recorded - for instance if somebody sent an entire SpamAssassin report back
3168 as a large number of 550 error lines. This bug was coincidentally fixed by
3169 increasing the size of one of Exim's internal buffers (big_buffer) that
3170 happened as part of the Exiscan merge. However, to be on the safe side, I
3171 have made the code more robust (and fixed the comments that describe what
3174 33. Some experimental protocols are using DNS PTR records for new purposes. The
3175 keys for these records are domain names, not reversed IP addresses. The
3176 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3177 leaves it alone. Component reversal etc. now happens only for IP addresses.
3180 34. The host_aton() function is supposed to be passed a string that is known
3181 to be a valid IP address. However, in the case of IPv6 addresses, it was
3182 not checking this. This is a hostage to fortune. Exim now panics and dies
3183 if the condition is not met. A case was found where this could be provoked
3184 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3185 components; fortuitously, this particular loophole had already been fixed
3186 by change 4.50/55 or 4.44/33 above.
3188 If there are any other similar loopholes, the new check in host_aton()
3189 itself should stop them being exploited. The report I received stated that
3190 data on the command line could provoke the exploit when Exim was running as
3191 exim, but did not say which command line option was involved. All I could
3192 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3193 running as the user.
3196 35. There was a buffer overflow vulnerability in the SPA authentication code
3197 (which came originally from the Samba project). I have added a test to the
3198 spa_base64_to_bits() function which I hope fixes it.
3201 36. The daemon start-up calls getloadavg() while still root for those OS that
3202 need the first call to be done as root, but it missed one case: when
3203 deliver_queue_load_max is set with deliver_drop_privilege. This is
3204 necessary for the benefit of the queue runner, because there is no re-exec
3205 when deliver_drop_privilege is set.
3207 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3208 was in use, was not putting the data itself into the right store pool;
3209 consequently, it could be overwritten for a subsequent message in the same
3210 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3213 38. Sometimes the final signoff response after QUIT could fail to get
3214 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3215 < 0 before doing a fflush(). This bug looks as though it goes back to the
3216 introduction of TLS in release 3.20, but "sometimes" must have been rare
3217 because the tests only now provoked it.
3223 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3224 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3225 Now it returns an integer. A number of calls were still expecting a T/F
3226 return. Fortuitously, in all cases, the tests worked in OK situations,
3227 which is the norm. However, things would have gone wrong on any write
3228 failures on the smtp file descriptor. This function is used when sending
3229 messages over SMTP and also when doing verify callouts.
3231 2. When Exim is called to do synchronous delivery of a locally submitted
3232 message (the -odf or -odi options), it no longer closes stderr before doing
3235 3. Implemented the mua_wrapper option.
3237 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3239 5. Implemented the functions header_remove(), header_testname(),
3240 header_add_at_position(), and receive_remove_recipient(), and exported them
3243 6. If an ACL "warn" statement specified the addition of headers, Exim already
3244 inserted X-ACL-Warn: at the start if there was no header name. However, it
3245 was not making this test for the second and subsequent header lines if
3246 there were newlines in the string. This meant that an invalid header could
3247 be inserted if Exim was badly configured.
3249 7. Allow an ACL "warn" statement to add header lines at the start or after all
3250 the Received: headers, as well as at the end.
3252 8. Added the rcpt_4xx retry error code.
3254 9. Added postmaster_mailfrom=xxx to callout verification option.
3256 10. Added mailfrom=xxxx to the callout verification option, for verify=
3259 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3260 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3262 12. Inserted some casts to stop certain compilers warning when using pointer
3263 differences as field lengths or precisions in printf-type calls (mostly
3264 affecting debugging statements).
3266 13. Added optional readline() support for -be (dynamically loaded).
3268 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3269 same clock tick as a message's arrival, so that its received time was the
3270 same as the "first fail" time on the retry record, and that message
3271 remained on the queue past the ultimate address timeout, every queue runner
3272 would try a delivery (because it was past the ultimate address timeout) but
3273 after another failure, the ultimate address timeout, which should have then
3274 bounced the address, did not kick in. This was a "< instead of <=" error;
3275 in most cases the first failure would have been in the next clock tick
3276 after the received time, and all would be well.
3278 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3279 being recognized when the domain list was tested by the match_domain
3280 condition in an expansion string.
3282 16. Added the ${str2b64: operator.
3284 17. Exim was always calling setrlimit() to set a large limit for the number of
3285 processes, without checking whether the existing limit was already
3286 adequate. (It did check for the limit on file descriptors.) Furthermore,
3287 errors from getrlimit() and setrlimit() were being ignored. Now they are
3288 logged to the main and panic logs, but Exim does carry on, to try to do its
3289 job under whatever limits there are.
3291 18. Imported PCRE 5.0.
3293 19. Trivial typo in log message " temporarily refused connection" (the leading
3296 20. If the log selector return_path_on_delivery was set and an address was
3297 redirected to /dev/null, the delivery process crashed because it assumed
3298 that a return path would always be set for a "successful" delivery. In this
3299 case, the whole delivery is bypassed as an optimization, and therefore no
3302 21. Internal re-arrangement: the function for sending a challenge and reading
3303 a response while authentication was assuming a zero-terminated challenge
3304 string. It's now changed to take a pointer and a length, to allow for
3305 binary data in such strings.
3307 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3309 23. Exim was not respecting finduser_retries when seeking the login of the
3310 uid under which it was called; it was always trying 10 times. (The default
3311 setting of finduser_retries is zero.) Also, it was sleeping after the final
3312 failure, which is pointless.
3314 24. Implemented tls_on_connect_ports.
3316 25. Implemented acl_smtp_predata.
3318 26. If the domain in control=submission is set empty, Exim assumes that the
3319 authenticated id is a complete email address when it generates From: or
3320 Sender: header lines.
3322 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3323 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3324 chown and chgrp in /bin and hostname in /usr/bin.
3326 28. Exim was keeping the "process log" file open after each use, just as it
3327 does for the main log. This opens the possibility of it remaining open for
3328 long periods when the USR1 signal hits a daemon. Occasional processlog
3329 errors were reported, that could have been caused by this. Anyway, it seems
3330 much more sensible not to leave this file open at all, so that is what now
3333 29. The long-running daemon process does not normally write to the log once it
3334 has entered its main loop, and it closes the log before doing so. This is
3335 so that log files can straightforwardly be renamed and moved. However,
3336 there are a couple of unusual error situations where the daemon does write
3337 log entries, and I had neglected to close the log afterwards.
3339 30. The text of an SMTP error response that was received during a remote
3340 delivery was being truncated at 512 bytes. This is too short for some of
3341 the long messages that one sometimes sees. I've increased the limit to
3344 31. It is now possible to make retry rules that apply only when a message has a
3345 specific sender, in particular, an empty sender.
3347 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3348 it possible to be selective about when SMTP synchronization is enforced.
3350 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3352 32. Implemented hosts_connection_nolog.
3354 33. Added an ACL for QUIT.
3356 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3359 35. Added mailbox_size and mailbox_filecount to appendfile.
3361 36. Added control = no_multiline_responses to ACLs.
3363 37. There was a bug in the logic of the code that waits for the clock to tick
3364 in the case where the clock went backwards by a substantial amount such
3365 that the microsecond fraction of "now" was more than the microsecond
3366 fraction of "then" (but the whole seconds number was less).
3368 38. Added support for the libradius Radius client library this is found on
3369 FreeBSD (previously only the radiusclient library was supported).
3375 1. When certain lookups returned multiple values in the form name=value, the
3376 quoting of the values was not always being done properly. Specifically:
3377 (a) If the value started with a double quote, but contained no whitespace,
3379 (b) If the value contained whitespace other than a space character (i.e.
3380 tabs or newlines or carriage returns) it was not quoted.
3381 This fix has been applied to the mysql and pgsql lookups by writing a
3382 separate quoting function and calling it from the lookup code. The fix
3383 should probably also be applied to nisplus, ibase and oracle lookups, but
3384 since I cannot test any of those, I have not disturbed their existing code.
3386 2. A hit in the callout cache for a specific address caused a log line with no
3387 reason for rejecting RCPT. Now it says "Previous (cached) callout
3388 verification failure".
3390 3. There was an off-by-one bug in the queryprogram router. An over-long
3391 return line was truncated at 256 instead of 255 characters, thereby
3392 overflowing its buffer with the terminating zero. As well as fixing this, I
3393 have increased the buffer size to 1024 (and made a note to document this).
3395 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3396 when Exim is waiting for an SMTP response from a remote server, Exim
3397 restarts its select() call on the socket, thereby resetting its timeout.
3398 This is not a problem when such interrupts are rare. Somebody set up a cron
3399 job to run exiwhat every 2 minutes, which is less than the normal select()
3400 timeout (5 or 10 minutes). This meant that the select() timeout never
3401 kicked in because it was always reset. I have fixed this by comparing the
3402 time when an interrupt arrives with the time at the start of the first call
3403 to select(). If more time than the timeout has elapsed, the interrupt is
3404 treated as a timeout.
3406 5. Some internal re-factoring in preparation for the addition of Sieve
3407 extensions (by MH). In particular, the "personal" test is moved to a
3408 separate function, and given an option for scanning Cc: and Bcc: (which is
3409 not set for Exim filters).
3411 6. When Exim created an email address using the login of the caller as the
3412 local part (e.g. when creating a From: or Sender: header line), it was not
3413 quoting the local part when it contained special characters such as @.
3415 7. Installed new OpenBSD configuration files.
3417 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3418 try to make them clearer.
3420 9. Callout options, other than the timeout value, were being ignored when
3421 verifying sender addresses in header lines. For example, when using
3423 verify = header_sender/callout=no_cache
3425 the cache was (incorrectly) being used.
3427 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3428 only the Cygwin environment.
3430 11. When return_path_on_delivery was set as a log selector, if different remote
3431 addresses in the same message used different return paths and parallel
3432 remote delivery occurred, the wrong values would sometimes be logged.
3433 (Whenever a remote delivery process finished, the return path value from
3434 the most recently started remote delivery process was logged.)
3436 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3437 lines when AUTH and/or TLS are in use. This is the "received protocol"
3438 field. Exim used to use "asmtp" for authenticated SMTP, without any
3439 indication (in the protocol name) for TLS use. Now it follows the RFC and
3440 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3441 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3442 names appear in log lines as well as in Received: header lines.
3444 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3445 extensions, and comparison tests, and to fix some bugs.
3447 14. Changes to the "personal" filter test:
3449 (1) The test was buggy in that it was just doing the equivalent of
3450 "contains" tests on header lines. For example, if a user's address was
3451 anne@some.where, the "personal" test would incorrectly be true for
3453 To: susanne@some.where
3455 This test is now done by extracting each address from the header in turn,
3456 and checking the entire address. Other tests that are part of "personal"
3457 are now done using regular expressions (for example, to check local parts
3458 of addresses in From: header lines).
3460 (2) The list of non-personal local parts in From: addresses has been
3461 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3462 taken from the Sieve specification recommendations.
3464 (3) If the message contains any header line starting with "List-" it is
3465 treated as non-personal.
3467 (4) The test for "circular" in the Subject: header line has been removed
3468 because it now seems ill-conceived.
3470 15. Minor typos in src/EDITME comments corrected.
3472 16. Installed latest exipick from John Jetmore.
3474 17. If headers_add on a router specified a text string that was too long for
3475 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3476 of string_sprintf() is now avoided.
3478 18. $message_body_size was not set (it was always zero) when running the DATA
3479 ACL and the local_scan() function.
3481 19. For the "mail" command in an Exim filter, no default was being set for
3482 the once_repeat time, causing a random time value to be used if "once" was
3483 specified. (If the value happened to be <= 0, no repeat happened.) The
3484 default is now 0s, meaning "never repeat". The "vacation" command was OK
3485 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3486 (I found it when inspecting the code).
3488 20. There is now an overall timeout for performing a callout verification. It
3489 defaults to 4 times the callout timeout, which applies to individual SMTP
3490 commands during the callout. The overall timeout applies when there is more
3491 than one host that can be tried. The timeout is checked before trying the
3492 next host. This prevents very long delays if there are a large number of
3493 hosts and all are timing out (e.g. when the network connections are timing
3494 out). The value of the overall timeout can be changed by specifying an
3495 additional sub-option for "callout", called "maxwait". For example:
3497 verify = sender/callout=5s,maxwait=20s
3499 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3500 to the end before writing, but this should make it even safer).
3502 22. Exim was forgetting that it had advertised PIPELINING for the second and
3503 subsequent messages on an SMTP connection. It was also not resetting its
3504 memory on STARTTLS and an internal HELO.
3506 23. When Exim logs an SMTP synchronization error within a session, it now
3507 records whether PIPELINING has been advertised or not.
3509 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3510 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3511 rather than long int.
3513 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3515 26. Added the never_mail option to autoreply.
3521 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3522 crash if the getsockname() call failed; this can happen if a connection is
3523 closed very soon after it is established. The problem was simply in the
3524 order in which certain operations were done, causing Exim to try to write
3525 to the SMTP stream before it had set up the file descriptor. The bug has
3526 been fixed by making things happen in the correct order.
3532 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3533 before the connection was closed, thus losing the rejection response.
3535 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3536 some early Solaris releases, but causes trouble in current releases where
3537 socklen_t is defined.
3539 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3542 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3545 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3546 that releasing the top of it at the end releases what was used for sub-
3547 expansions (unless the block got too big). However, discard this block if
3548 the first thing is a variable or header, so that we can use its block when
3549 it is dynamic (useful for very large $message_headers, for example).
3551 6. Lookups now cache *every* query, not just the most recent. A new, separate
3552 store pool is used for this. It can be recovered when all lookup caches are
3553 flushed. Lookups now release memory at the end of their result strings.
3554 This has involved some general refactoring of the lookup sources.
3556 7. Some code has been added to the store_xxx() functions to reduce the amount
3557 of flapping under certain conditions.
3559 8. log_incoming_interface used to affect only the <= reception log lines. Now
3560 it causes the local interface and port to be added to several more SMTP log
3561 lines, for example "SMTP connection from", and rejection lines.
3563 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3565 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3567 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3568 could be overwritten at the end of the current message (or the start of a
3569 new message if it was set in a HELO ACL). The value is now preserved for
3570 the duration of the SMTP connection.
3572 12. If a transport had a headers_rewrite setting, and a matching header line
3573 contained an unqualified address, that address was qualified, even if it
3574 did not match any rewriting rules. The underlying bug was that the values
3575 of the flags that permit the existence of unqualified sender and recipient
3576 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3577 non-local messages, and by -bnq for local messages) were not being
3578 preserved with the message after it was received.
3580 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3581 "next input=" as part of the text comprising the host identity instead of
3582 the correct text. The code was using the same buffer for two different
3583 strings. However, depending on which order the printing function evaluated
3584 its arguments, the bug did not always show up. Under Linux, for example, my
3585 test suite worked just fine.
3587 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3588 doesn't work with some older versions of Perl. It has been changed to "my",
3589 which in any case is probably the better facility to use.
3591 15. A really picky compiler found some instances of statements for creating
3592 error messages that either had too many or two few arguments for the format
3595 16. The size of the buffer for calls to the DNS resolver has been increased
3596 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3597 for addresses that have a lot of PTR records. This alleviates a problem; it
3598 does not fully solve it.
3600 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3601 buffer now truncates the list and logs the incident, which is the same
3602 action as happens when Exim is looking up a host name and its aliases.
3603 Previously in this situation something unpredictable would happen;
3604 sometimes it was "internal error: store_reset failed".
3606 18. If a server dropped the connection unexpectedly when an Exim client was
3607 using GnuTLS and trying to read a response, the client delivery process
3608 crashed while trying to generate an error log message.
3610 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3611 string, for example:
3613 warn message = H1: something\nH2: something
3615 the text was added as a single header line from Exim's point of view
3616 though it ended up OK in the delivered message. However, searching for the
3617 second and subsequent header lines using $h_h2: did not work. This has been
3618 fixed. Similarly, if a system filter added multiple headers in this way,
3619 the routers could not see them.
3621 20. Expanded the error message when iplsearch is called with an invalid key to
3622 suggest using net-iplsearch in a host list.
3624 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3625 ACLs are no longer actually imposed (and a message to that effect is
3628 22. If a "gecos" field in a passwd entry contained escaped characters, in
3629 particular, if it contained a \" sequence, Exim got it wrong when building
3630 a From: or a Sender: header from that name. A second bug also caused
3631 incorrect handling when an unquoted " was present following a character
3632 that needed quoting.
3634 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3635 was not being matched caselessly.
3637 24. Arranged for all hyphens in the exim.8 source to be escaped with
3640 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3641 properly. Recipient callout cache records were still being keyed to include
3642 the sender, even when use_sender was set false. This led to far more
3643 callouts that were necessary. The sender is no longer included in the key
3644 when use_sender is false.
3646 26. Added "control = submission" modifier to ACLs.
3648 27. Added the ${base62d: operator to decode base 62 numbers.
3650 28. dnsdb lookups can now access SRV records.
3652 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3653 the configuration file.
3655 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3656 (-v) mode. This makes the output for a verbose queue run more intelligible.
3658 31. Added a use_postmaster feature to recipient callouts.
3660 32. Added the $body_zerocount variable, containing the number of binary zero
3661 bytes in the message body.
3663 33. The time of last modification of the "new" subdirectory is now used as the
3664 "mailbox time last read" when there is a quota error for a maildir
3667 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3669 35. Added +ignore_unknown as a special item in host lists.
3671 36. Code for decoding IPv6 addresses in host lists is now included, even if
3672 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3673 address was recognized as an IP address, but was then not correctly decoded
3674 into binary, causing unexpected and incorrect effects when compared with
3681 1. Very minor rewording of debugging text in manualroute to say "list of
3682 hosts" instead of "hostlist".
3684 2. If verify=header_syntax was set, and a header line with an unqualified
3685 address (no domain) and a large number of spaces between the end of the
3686 name and the colon was received, the reception process suffered a buffer
3687 overflow, and (when I tested it) crashed. This was caused by some obsolete
3688 code that should have been removed. The fix is to remove it!
3690 3. When running in the test harness, delay a bit after writing a bounce
3691 message to get a bit more predictability in the log output.
3693 4. Added a call to search_tidyup() just before forking a reception process. In
3694 theory, someone could use a lookup in the expansion of smtp_accept_max_
3695 per_host which, without the tidyup, could leave open a database connection.
3697 5. Added the variables $recipient_data and $sender_data which get set from a
3698 lookup success in an ACL "recipients" or "senders" condition, or a router
3699 "senders" option, similar to $domain_data and $local_part_data.
3701 6. Moved the writing of debug_print from before to after the "senders" test
3704 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3705 problems for message scanning, either using a data ACL, or using
3706 local_scan() because the Received: header was not generated till after they
3707 were called (in order to set the time as the time of reception completion).
3708 I have revised the way this works. The header is now generated after the
3709 body is received, but before the ACL or local_scan() are called. After they
3710 are run, the timestamp in the header is updated.
3716 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3717 before starting a queue runner without re-exec. This happened only when
3718 deliver_drop_privilege was set or when the Exim user was set to root. The
3719 effect of the bug was that timeouts during subsequent deliveries caused
3720 crashes instead of being properly handled. The handler is now left at its
3721 default (and expected) setting.
3723 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3724 message, again when deliver_drop_privilege is set or Exim is run as root.
3725 The bug described in (1) was not present in this case, but the tidying up
3726 of the other signals was missing. I have made the two cases consistent.
3728 3. The ignore_target_hosts setting on a manualroute router was being ignored
3729 for hosts that were looked up using the /MX notation.
3731 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3734 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3735 operated on the sender address. After changing the $sender_address to <>
3736 for the sender address verify, Exim was re-instated it as the original
3737 (before rewriting) address, but remembering that it had rewritten it, so it
3738 wasn't rewriting it again. This bug also had the effect of breaking the
3739 sender address verification caching when the sender address was rewritten.
3741 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3742 This has been changed so that if the ip literal address matches
3743 ignore_target_hosts, the router declines.
3745 7. Added expansion conditions match_domain, match_address, and match_local_
3746 part (NOT match_host).
3748 8. The placeholder for the Received: header didn't have a length field set.
3750 9. Added code to Exim itself and to exim_lock to test for a specific race
3751 condition that could lead to file corruption when using MBX delivery. The
3752 issue is with the lockfile that is created in /tmp. If this file is removed
3753 after a process has opened it but before that process has acquired a lock,
3754 there is the potential for a second process to recreate the file and also
3755 acquire a lock. This could lead to two Exim processes writing to the file
3756 at the same time. The added code performs the same test as UW imapd; it
3757 checks after acquiring the lock that its file descriptor still refers to
3758 the same named file.
3760 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3761 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3764 11. Added the smtp_active_hostname option. If used, this will typically be made
3765 to depend on the incoming interface address. Because $interface_address is
3766 not set up until the daemon has forked a reception process, error responses
3767 that can happen earlier (such as "too many connections") no longer contain
3770 12. If an expansion in a condition on a "warn" statement fails because a lookup
3771 defers, the "warn" statement is abandoned, and the next ACL statement is
3772 processed. Previously this caused the whole ACL to be aborted.
3774 13. Added the iplsearch lookup type.
3776 14. Added ident_timeout as a log selector.
3778 15. Added tls_certificate_verified as a log selector.
3780 16. Added a global option tls_require_ciphers (compare the smtp transport
3781 option of the same name). This controls incoming TLS connections.
3783 17. I finally figured out how to make tls_require_ciphers do a similar thing
3784 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3785 before starting the TLS session.
3787 18. Tabs are now shown as \t in -bP output.
3789 19. If the log selector return_path_on_delivery was set, Exim crashed when
3790 bouncing a message because it had too many Received: header lines.
3792 20. If two routers both had headers_remove settings, and the first one included
3793 a superfluous trailing colon, the final name in the first list and the
3794 first name in the second list were incorrectly joined into one item (with a
3795 colon in the middle).
3801 1. Added -C and -D options to the exinext utility, mainly to make it easier
3802 to include in the automated testing, but these could be helpful when
3803 multiple configurations are in use.
3805 2. The exinext utility was not formatting the output nicely when there was
3806 an alternate port involved in the retry record key, nor when there was a
3807 message id as well (for retries that were specific to a specific message
3808 and a specific host). It was also confused by IPv6 addresses, because of
3809 the additional colons they contain. I have fixed the IPv4 problem, and
3810 patched it up to do a reasonable job for IPv6.
3812 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3813 delivery, the log line now contains "pipelined" if PIPELINING was used.
3815 4. An SMTP transport process used to panic and die if the bind() call to set
3816 an explicit outgoing interface failed. This has been changed; it is now
3817 treated in the same way as a connect() failure.
3819 5. A reference to $sender_host_name in the part of a conditional expansion
3820 that was being skipped was still causing a DNS lookup. This no longer
3823 6. The def: expansion condition was not recognizing references to header lines
3824 that used bh_ and bheader_.
3826 7. Added the _cache feature to named lists.
3828 8. The code for checking quota_filecount in the appendfile transport was
3829 allowing one more file than it should have been.
3831 9. For compatibility with Sendmail, the command line option
3839 and sets the incoming protocol and host name (for trusted callers). The
3840 host name and its colon can be omitted when only the protocol is to be set.
3841 Note the Exim already has two private options, -pd and -ps, that refer to
3842 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3843 "s", but I don't think that's a major issue.
3845 10. A number of refactoring changes to the code, none of which should affect
3848 (a) The number of logging options was getting close to filling up the
3849 32-bit word that was used as a bit map. I have split them into two classes:
3850 those that are passed in the argument to log_write(), and those that are
3851 only ever tested independently outside of that function. These are now in
3852 separate 32-bit words, so there is plenty of room for expansion again.
3853 There is no change in the user interface or the logging behaviour.
3855 (b) When building, for example, log lines, the code previously used a
3856 macro that called string_cat() twice, in order to add two strings. This is
3857 not really sufficiently general. Furthermore, there was one instance where
3858 it was actually wrong because one of the argument was used twice, and in
3859 one call a function was used. (As it happened, calling the function twice
3860 did not affect the overall behaviour.) The macro has been replaced by a
3861 function that can join an arbitrary number of extra strings onto a growing
3864 (c) The code for expansion conditions now uses a table and a binary chop
3865 instead of a serial search (which was left over from when there were very
3866 few conditions). Also, it now recognizes conditions like "pam" even when
3867 the relevant support is not compiled in: a suitably worded error message is
3868 given if an attempt is made to use such a condition.
3870 11. Added ${time_interval:xxxxx}.
3872 12. A bug was causing one of the ddress fields not to be passed back correctly
3873 from remote delivery subprocesses. The field in question was not being
3874 subsequently used, so this caused to problems in practice.
3876 13. Added new log selectors queue_time and deliver_time.
3878 14. Might have fixed a bug in maildirsizefile handling that threw up
3879 "unexpected character" debug warnings, and recalculated the data
3880 unnecessarily. In any case, I expanded the warning message to give more
3883 15. Added the message "Restricted characters in address" to the statements in
3884 the default ACL that block characters like @ and % in local parts.
3886 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3887 Three changes have been made:
3889 (a) There was a serious bug; a negative response to MAIL caused the whole
3890 recipient domain to be cached as invalid, thereby blocking all messages
3891 to all local parts at the same domain, from all senders. This bug has
3892 been fixed. The domain is no longer cached after a negative response to
3893 MAIL if the sender used is not empty.
3895 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3898 (c) A new callout option, "use_sender" has been added for people who want
3899 the modified behaviour.
3905 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3908 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3909 indeed breaks things for older releases.
3911 3. Added additional logging to the case where there is a problem reading data
3912 from a filter that is running in a subprocess using a pipe, in order to
3913 try to track down a specific problem.
3915 4. Testing facility fudge: when running in the test harness and attempting
3916 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3917 getting "No route to host". Convert this to a timeout.
3919 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3922 6. Some OS don't have socklen_t but use size_t instead. This affects the
3923 fifth argument of getsockopt() amongst other things. This is now
3924 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3925 can be set for individual OS. I have set it for SunOS5, OSF1, and
3926 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3927 some earlier ones do not.
3929 7. Change 4.30/15 was not doing the test caselessly.
3931 8. The standard form for an IPv6 address literal was being rejected by address
3932 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3933 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3934 this, as well as the form without the "IPv6" on the front (but only when
3935 address literals are enabled, of course).
3937 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3939 10. Exim crashed if a message with an empty sender address specified by -f
3940 encountered a router with an errors_to setting. This could be provoked only
3941 by a command such as
3945 where an empty string was supplied; "<>" did not hit this bug.
3947 11. Installed PCRE release 4.5.
3949 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3950 remained set. It is now erased.
3952 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3953 times from message ids (which are base 36 rather than the normal 62).
3955 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3956 were being counted as actual protocol errors, and logged if the log
3957 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3958 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3959 and DATA following a set of rejected RCPTs do not count as protocol errors.
3960 In other words, Exim assumes they were pipelined, though this may not
3961 actually be the case. Of course, in all cases the client gets an
3962 appropriate error code.
3964 15. If a lookup fails in an ACL condition, a message about the failure may
3965 be available; it is used if testing the ACL cannot continue, because most
3966 such messages specify what the cause of the deferral is. However, some
3967 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3968 that caused an old message to be retained and used if a later statement
3969 caused a defer, replacing the real cause of the deferral.
3971 16. If an IP address had so many PTR records that the DNS lookup buffer
3972 was not large enough to hold them, Exim could crash while trying to process
3973 the truncated data. It now detects and logs this case.
3975 17. Further to 4.21/58, another change has been made: if (and only if) the
3976 first line of a message (the first header line) ends with CRLF, a bare LF
3977 in a subsequent header line has a space inserted after it, so as not to
3978 terminate the header.
3980 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3981 unnecessarily, used atoi() instead of strtol(), and didn't check the
3982 termination when getting file sizes from file names by regex.
3984 19. Completely re-implemented the support for maildirsize files, in the light
3985 of a number of problems with the previous contributed implementation
3986 (4.30/29). In particular:
3988 . If the quota is zero, the maildirsize file is maintained, but no quota is
3991 . If the maildir directory does not exist, it is created before any attempt
3992 to write a maildirsize file.
3994 . The quota value in the file is just a cache; if the quota is changed in
3995 the transport, the new value overrides.
3997 . A regular expression is available for excluding directories from the
4000 20. The autoreply transport checks the characters in options that define the
4001 message's headers; it allows continued headers, but it was checking with
4002 isspace() after an embedded newline instead of explicitly looking for a
4005 21. If all the "regular" hosts to which an address was routed had passed their
4006 expiry times, and had not reached their retry times, the address was
4007 bounced, even if fallback hosts were defined. Now Exim should go on to try
4010 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4011 equivalent code in the SMTP transport. Some hosts send humungous responses
4012 to HELO/EHLO, more than 1024 it seems.
4014 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4015 gives compiler warnings in some environments. I've now done it "properly",
4018 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4019 (because of problems with the built-in one) was declared to return uschar *
4020 instead of char *, causing compiler failure.
4022 25. Fixed a file descriptor leak when processing alias/forward files.
4024 26. Fixed a minor format string issue in dbfn.c.
4026 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4028 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4029 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4032 29. When a sender address is verified, it is cached, to save repeating the test
4033 when there is more than one recipient in a message. However, when the
4034 verification involves a callout, it is possible for different callout
4035 options to be set for different recipients. It is too complicated to keep
4036 track of this in the cache, so now Exim always runs a verification when a
4037 callout is required, relying on the callout cache for the optimization.
4038 The overhead is duplication of the address routing, but this should not be
4041 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4042 to be verified with callout=postmaster, and the main callout worked but the
4043 postmaster check failed, the verification correctly failed. However, if a
4044 subsequent RCPT command asked for sender verification *without* the
4045 postmaster check, incorrect caching caused this verification also to fail,
4048 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4049 it was not caching the DNS options (qualify_single, search_parents) that
4050 were used when the lookup failed. A subsequent lookup with different
4051 options therefore always gave the same answer, though there were cases
4052 where it should not have. (Example: a "domains = !$mx_any" option on a
4053 dnslookup router: the "domains" option is always processed without any
4054 widening, but the router might have qualify_single set.) Now Exim uses the
4055 cached value only when the same options are set.
4057 32. Added John Jetmore's "exipick" utility to the distribution.
4059 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4060 than a timeout (e.g. a certificate is required, and is not provided), an
4061 Exim server now closes the connection immediately. Previously it waited for
4062 the client to close - but if the client is SSL, it seems that they each
4063 wait for each other, leading to a delay before one of them times out.
4065 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4066 maintained 0.8.x compatibility because I don't think many are using it, and
4067 it is clearly obsolete.
4069 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4072 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4073 client certificate was expired. A simple patch fixes this, though I don't
4074 understand the full logic of why the verify callback is called multiple
4077 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4078 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4079 which causes problems with some clients (such as the Certicom SSL Plus
4080 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4081 disables the coutermeasure allowing Eudora to connect."
4083 38. Exim was not checking that a write() to a log file succeeded. This could
4084 lead to Bad Things if a log got too big, in particular if it hit a file
4085 size limit. Exim now panics and dies if it cannot write to a log file, just
4086 as it does if it cannot open a log file.
4088 39. Modified OS/Makefile-Linux so that it now contains
4090 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4092 The two -D definitions ensure that Exim is compiled with large file
4093 support, which makes it possible to handle log files that are bigger than
4096 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4097 instance) a domain was checked against a named list that involved a lookup,
4098 causing $domain_data to be set, then another domain was checked against the
4099 same list, then the first domain was re-checked, the value of $domain_data
4100 after the final check could be wrong. In particular, if the second check
4101 failed, it could be set empty. This bug probably also applied to
4104 41. The strip_trailing_dot option was not being applied to the address given
4105 with the -f command-line option.
4107 42. The code for reading a message's header from the spool was incrementing
4108 $received_count, but never initializing it. This meant that the value was
4109 incorrect (doubled) while delivering a message in the same process in which
4110 it was received. In the most common configuration of Exim, this never
4111 happens - a fresh exec is done - but it can happen when
4112 deliver_drop_privilege is set.
4114 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4115 it now includes up to 150 characters of the unexpected data in the log
4118 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4119 and building data strings. The size of both of these buffers was 10 000
4120 bytes - far larger than anybody would *ever* want, thought I. Needless to
4121 say, somebody hit the limit. I have increased the maximum line length to
4122 20 000 and the maximum data length of concatenated lines to 100 000. I have
4123 also fixed two bugs, because there was no checking on these buffers. Tsk,
4124 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4125 buffer is too small.
4127 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4128 lsearch lookups. Now it does.
4130 46. When parsing a route_list item in a manualroute router, a fixed-length
4131 buffer was used for the list of hosts. I made this 1024 bytes long,
4132 thinking that nobody would ever have a list of hosts that long. Wrong.
4133 Somebody had a whole pile of complicated expansion conditions, and the
4134 string was silently truncated, leading to an expansion error. It turns out
4135 that it is easier to change to an unlimited length (owing to other changes
4136 that have happened since this code was originally written) than to build
4137 structure for giving a limitation error. The length of the item that
4138 expands into the list of hosts is now unlimited.
4140 47. The lsearch lookup could not handle data where the length of text line was
4141 more than 4095 characters. Such lines were truncated, leading to shortened
4142 data being returned. It should now handle lines of any length.
4144 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4145 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4148 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4149 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4150 no longer happens in comment lines. A list of replacements is now placed
4151 at the head of all of the source files, except those whose only change is
4152 to replace PERL_COMMAND in the very first #! line.
4154 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4155 messages on the queue, with a bottom-up merge sort, using code contributed
4156 by Michael Haardt. This should make operations like -bp somewhat faster on
4157 large queues. It won't affect queue runners, except when queue_run_in_order
4160 51. Installed eximstats 1.31 in the distribution.
4162 52. Added support for SRV lookups to the dnslookup router.
4164 53. If an ACL referred to $message_body or $message_body_end, the value was not
4165 reset for any messages that followed in the same SMTP session.
4167 54. The store-handling optimization for building very long strings was not
4168 differentiating between the different store pools. I don't think this
4169 actually made any difference in practice, but I've tidied it.
4171 55. While running the routers to verify a sender address, $sender_address
4172 was still set to the sender address. This is wrong, because when routing to
4173 send a bounce to the sender, it would be empty. Therefore, I have changed
4174 it so that, while verifying a sender address, $sender_address is set to <>.
4175 (There is no change to what happens when verifying a recipient address.)
4177 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4178 target A or AAAA records (if not already returned) without resetting the
4179 qualify_single or search_parents options of the DNS resolver. These are
4180 inappropriate in this case because the targets of MX and SRV records must
4181 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4182 target that, when qualified, matched something in the local domain. These
4183 two options are now turned off when doing these lookups.
4185 57. It seems that at least some releases of Reiserfs (which does not have the
4186 concept of a fixed number of inodes) returns zero and not -1 for the
4187 number of available inodes. This interacted badly with check_spool_inodes,
4188 which assumed that -1 was the "no such thing" setting. What I have done is
4189 to check that the total number of inodes is greater than zero before doing
4190 the test of how many are available.
4192 58. When a "warn" ACL statement has a log_message modifier, the message is
4193 remembered, and not repeated. This is to avoid a lot of repetition when a
4194 message has many recipients that cause the same warning to be written.
4195 Howewer, Exim was preserving the list of already written lines for an
4196 entire SMTP session, which doesn't seem right. The memory is now reset if a
4197 new message is started.
4199 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4200 debugging output unless log_rewrite was also set.
4202 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4203 of (int)(handle) when we know that handle contains (void *)(-1).
4205 61. The Exim daemon panic-logs an error return when it closes the incoming
4206 connection. However "connection reset by peer" seems to be common, and
4207 isn't really an error worthy of noting specially, so that particular error
4210 62. When Exim is trying to find all the local interfaces, it used to panic and
4211 die if the ioctl to get the interface flags failed. However, it seems that
4212 on at least one OS (Solaris 9) it is possible to have an interface that is
4213 included in the list of interfaces, but for which you get a failure error
4214 for this call. This happens when the interface is not "plumbed" into a
4215 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4216 failure of the "get flags" call assumes that the interface is down.
4218 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4219 makes life easier for people who are doing arithmetic on fields extracted
4220 from dates, where you often get leading zeros that should not be
4221 interpreted as octal.
4223 64. Added qualify_domain to the redirect router, to override the global
4226 65. If a pathologically long header line contained very many addresses (the
4227 report of this problem mentioned 10 000) and each of them was rewritten,
4228 Exim could use up a very large amount of memory. (It kept on making new
4229 copies of the header line as it rewrote, and never released the old ones.)
4230 At the expense of a bit more processing, the header rewriting function has
4231 been changed so that it no longer eats memory in this way.
4233 66. The generation of the Received: header has been moved from the time that a
4234 message starts to be received, to the time that it finishes. The timestamp
4235 in the Received: header should now be very close to that of the <= log
4236 line. There are two side-effects of this change:
4238 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4239 logged header lines no longer include the local Received: line, because
4240 it has not yet been created. The same applies to a copy of the message
4241 that is returned to a non-SMTP sender when a message is rejected.
4243 (b) When a filter file is tested using -bf, no additional Received: header
4244 is added to the test message. After some thought, I decided that this
4247 This change does not affect the value of $received_for. It is still set
4248 after address rewriting, but before local_scan() is called.
4250 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4252 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4253 gave an unhelpful panic error message, and a defer error. I have managed to
4254 change this behaviour so that it now rejects any supplied certificate,
4255 which seems right, as the list of acceptable certificates is empty.
4257 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4258 gave an unhelpful defer error. I have not managed to make this reject any
4259 supplied certificates, but the error message it gives is "no certificate
4260 supplied", which is not helpful.
4262 70. exigrep's output now also includes lines that are not associated with any
4263 message, but which match the given pattern. Implemented by a patch from
4264 Martin Sluka, which also tidied up the Perl a bit.
4266 71. Recipient callout verification, like sender verification, was using <> in
4267 the MAIL FROM command. This isn't really the right thing, since the actual
4268 sender may affect whether the remote host accepts the recipient or not. I
4269 have changed it to use the actual sender in the callout; this means that
4270 the cache record is now keyed on a recipient/sender pair, not just the
4271 recipient address. There doesn't seem to be a real danger of callout loops,
4272 since a callout by the remote host to check the sender would use <>.
4273 [SEE ABOVE: changed after hitting problems.]
4275 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4276 temporary errors. However, in the case of such a code being given after
4277 the end of a data transmission (i.e. after ".") Exim was failing to write
4278 a retry record for the message. (Yes, there was some broken host that was
4279 actually sending 8xx at this point.)
4281 73. An unknown lookup type in a host list could cause Exim to panic-die when
4282 the list was checked. (An example that provoked this was putting <; in the
4283 middle of a list instead of at the start.) If this happened during a DATA
4284 ACL check, a -D file could be left lying around. This kind of configuration
4285 error no longer causes Exim to die; instead it causes a defer errror. The
4286 incident is still logged to the main and panic logs.
4288 74. Buglet left over from Exim 3 conversion. The message "too many messages
4289 in one connection" was written to the rejectlog but not the mainlog, except
4290 when address rewriting (yes!) was being logged.
4292 75. Added write_rejectlog option.
4294 76. When a system filter was run not as root (that is, when system_filter_user
4295 was set), the values of the $n variables were not being returned to the
4296 main process; thus, they were not subsequently available in the $sn
4299 77. Added +return_path_on_delivery log selector.
4301 78. A connection timeout was being treated differently from recipients deferred
4302 when testing hosts_max_try with a message that was older than the host's
4303 retry timeout. (The host should not be counted, thus allowing all hosts to
4304 be tried at least once before bouncing.) This may have been the cause of an
4305 occasionally reported bug whereby a message would remain on the queue
4306 longer than the retry timeout, but would be bounced if a delivery was
4307 forced. I say "may" because I never totally pinned down the problem;
4308 setting up timeout/retry tests is difficult. See also the next item.
4310 79. The ultimate address timeout was not being applied to errors that involved
4311 a combination of host plus message (for example, a timeout on a MAIL
4312 command). When an address resolved to a number of possible hosts, and they
4313 were not all tried for each delivery (e.g. because of hosts_max_try), a
4314 message could remain on the queue longer than the retry timeout.
4316 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4319 81. Fixed an obscure SMTP outgoing bug which required at least the following
4320 conditions: (a) there was another message waiting for the same server;
4321 (b) the server returned 5xx to all RCPT commands in the first message so
4322 that the message was not completed; (c) the server dropped the connection
4323 or gave a negative response to the RSET that Exim sends to abort the
4324 transaction. The observed case was a dropped connection after DATA that had
4325 been sent in pipelining mode. That is, the server had advertised PIPELINING
4326 but was not implementing it correctly. The effect of the bug was incorrect
4327 behaviour, such as trying another host, and this could lead to a crash.
4333 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4334 and daemon.c were passed as pointers to ints; they should have been
4335 pointers to socklen_t variables (which are typically unsigned ints).
4337 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4340 3. Fixed a really odd bug that affected only the testing scheme; patching a
4341 certain fixed string in the binary changed the value of another string that
4342 happened to be identical to the end of the original first string.
4344 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4345 name", it returns that address as the IP address. On some operating
4346 systems (e.g. Solaris), it also passes back the IP address string as the
4347 "host name". However, on others (e.g. Linux), it passes back an empty
4348 string. Exim wasn't checking for this, and was changing the host name to an
4349 empty string, assuming it had been canonicized.
4351 5. Although rare, it is permitted to have more than one PTR record for a given
4352 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4353 all the names associated with an address, because they do in Solaris.
4354 However, it seems that they do not in Linux for data that comes from the
4355 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4356 I found this out when I moved to a new Linux workstation and tried to run
4357 the Exim test suite.
4359 To get round this problem I have changed the code so that it now does its
4360 own call to the DNS to look up PTR records when searching for a host name.
4361 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4362 addresses that are only in /etc/hosts are still found.
4364 This behaviour is, however, controlled by an option called host_lookup_
4365 order, which defaults to "bydns:byaddr". If people want to use the other
4366 order, or indeed, just use one or the other means of lookup, they can
4367 specify it in this variable.
4369 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4370 some operating systems, this comes back from gethostbyaddr() as an empty
4371 string, and this is what Exim used to test for. However, it seems that in
4372 other systems, "." is yielded. Exim now tests for this case too.
4374 7. The values of check_spool_space and check_log_space are now held internally
4375 as a number of kilobytes instead of an absolute number of bytes. If a
4376 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4377 kilobyte. This means that much larger values can be stored.
4379 8. Exim monitor: an attempt to get the action menu when not actually pointing
4380 at a message produces an empty menu entitled "No message selected". This
4381 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4382 no entries in it ("Shell widget menu has zero width and/or height"). So I
4383 have added a single, blank menu entry in this case.
4385 9. Added ${quote_local_part.
4387 10. MIME decoding is now applied to the contents of Subject: header lines when
4390 11. Now that a reference to $sender_host_address automatically causes a reverse
4391 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4392 host lookup before query-style lookups in lists that might use this
4393 variable. This has therefore been abolished, and the "net-" prefix is no
4394 longer necessary for query-style lookups.
4396 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4397 have been a typo for LFLAGS, so it has been changed.
4399 13. The install script calls Exim with "-C /dev/null" in order to find the
4400 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4401 to be output. Howeve, since Exim outputs its version number before the
4402 error, it didn't break the script. It just looked ugly. I fixed this by
4403 always allowing "-C /dev/null" if the caller is root.
4405 14. Ignore overlarge ACL variable number when reading spool file - insurance
4406 against a later release with more variables having written the file.
4408 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4409 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4410 this, as well as the form without the "IPv6" on the front.
4412 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4413 OS/Makefile-Darwin file.
4415 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4416 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4418 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4419 variable. However, some OS use ints for inodes. I've added cast to long int
4420 to get rid of the compiler warning.
4422 19. I had forgotten to lock out "/../" in configuration file names when
4423 ALT_CONFIG_PREFIX was set.
4425 20. Routers used for verification do not need to specify transports. However,
4426 if such a router generated a host list, and callout was configured, Exim
4427 crashed, because it could not find a port number from the (non-existent)
4428 transport. It now assumes port 25 in this circumstance.
4430 21. Added the -t option to exigrep.
4432 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4433 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4434 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4435 them is set, LOOKUP_LSEARCH is forced.
4437 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4439 24. Added sender and host information to the "rejected by local_scan()" log
4440 line; previously there was no indication of these.
4442 25. Added .include_if_exists.
4444 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4445 Linux. It turns out that not all file systems support this. Apparently some
4446 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4447 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4448 not supported on the file descriptor, is now ignored when Exim is trying to
4449 sync a directory. This applies only to Linux.
4451 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4453 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4454 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4457 29. Added support for maildirsize files from supplied patch (modified a bit).
4459 30. The use of :fail: followed by an empty string could lead Exim to respond to
4460 sender verification failures with (e.g.):
4462 550 Verification failed for <xxx>
4463 550 Sender verify failed
4465 where the first response line was missing the '-' that indicates it is not
4466 the final line of the response.
4468 31. The loop for finding the name of the user that called Exim had a hardwired
4469 limit of 10; it now uses the value of finduser_retries, which is used for
4470 all other user lookups.
4472 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4475 33. Exim was neglecting to zero errno before one call of strtol() when
4476 expanding a string and expecting an integer value. On some systems this
4477 resulted in spurious "integer overflow" errors. Also, it was casting the
4478 result into an int without checking.
4480 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4481 did not work. The code looks as if it has *never* worked, though it appears
4482 to have been documented since at least releast 1.62. I have made it work.
4484 35. The "timeout_DNS" error in retry rules, also documented since at least
4485 1.62, also never worked. As it isn't clear exactly what this means, and
4486 clearly it isn't a major issue, I have abolished the feature by treating it
4487 as "timeout", and writing a warning to the main and panic logs.
4489 36. The display of retry rules for -brt wasn't always showing the error code
4492 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4493 timeout_connect_A, timeout_connect_MX.
4495 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4496 to the empty sender.
4498 39. The daemon was not analysing the content of -oX till after it had closed
4499 stderr and disconnected from the controlling terminal. This meant that any
4500 syntax errors were only noted on the panic log, and the return code from
4501 the command was 0. By re-arranging the code a little, I've made the
4502 decoding happen first, so such errors now appear on stderr, and the return
4503 code is 1. However, the actual setting up of the sockets still happens in
4504 the disconnected process, so errors there are still only recorded on the
4507 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4508 connections (as happens on some IP stacks) was logged at start up time as
4509 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4510 it from "IPv6 and IPv4", which means that two separate sockets are being
4513 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4514 says whether AF_INET or AF_INET6 was passed as an argument.
4516 42. Exiwhat output was messed up when time zones were included in log
4519 43. Exiwhat now gives more information about the daemon's listening ports,
4520 and whether -tls-on-connect was used.
4522 44. The "port" option of the smtp transport is now expanded.
4524 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4525 silently ignored. Now an error message is written to the main and panic
4528 46. There's a new ACL modifier called "logwrite" which writes to a log file
4529 as soon as it is encountered.
4531 47. Added $local_user_uid and $local_user_gid at routing time.
4533 48. Exim crashed when trying to verify a sender address that was being
4536 49. Exim was recognizing only a space character after ".include". It now also
4537 recognizes a tab character.
4539 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4540 extracting the relevant information from the specification. The man page no
4541 longer contains scrambled data for the -d option, and I've added a section
4542 at the front about calling Exim under different names.
4544 51. Added "extra_headers" argument to the "mail" command in filter files.
4546 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4549 53. Installed eximstats 1.29.
4551 54. Added transport_filter_timeout as a generic transport option.
4553 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4554 Cc: header lines. This was required by RFC 822, but it not required by RFC
4557 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4558 incoming messages that did not have them. Now it does so only if the
4559 message originates locally, that is, if there is no associated remote host
4560 address. When Resent- header lines are present, this applies to the Resent-
4561 lines rather than the non-Resent- lines.
4563 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4564 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4566 58. Messages for configuration errors now include the name of the main
4567 configuration file - useful now that there may be more than one file in a
4568 list (.included file names were always shown).
4570 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4571 for those rare installations that do not start the daemon as root or run it
4572 setuid root. I've cut out the call to initgroups() if the daemon is not
4575 60. The Exim user and group can now be bound into the binary as text strings
4576 that are looked up at the start of Exim's processing.
4578 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4580 62. Added $mailstore_basename variable.
4582 63. Installed patch to sieve.c from Michael Haardt.
4584 64. When Exim failed to open the panic log after failing to open the main log,
4585 the original message it was trying to log was written to stderr and debug
4586 output, but if they were not available (the usual case in production), it
4587 was lost. Now it is written to syslog before the two lines that record the
4588 failures to open the logs.
4590 65. Users' Exim filters run in subprocesses under the user's uid. It is
4591 possible for a "deliver" command or an alias in a "personal" command to
4592 provoke an address rewrite. If logging of address rewriting is configured,
4593 this fails because the process is not running as root or exim. There may be
4594 a better way of dealing with this, but for the moment (because 4.30 needs
4595 to be released), I have disabled address rewrite logging when running a
4596 filter in a non-root, non-exim process.
4602 1. The buildconfig auxiliary program wasn't quoting the value set for
4603 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4604 not defined. This bug was masked in 4.22 by the effect that was fixed in
4607 2. Some messages that were rejected after a message id was allocated were
4608 shown as "incomplete" by exigrep. It no longer does this for messages that
4609 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4611 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4612 have allow_domain_literals set, the ID did not get logged in the <= line.
4613 Domain literals are now always recognized in Message-ID: header lines.
4615 4. The first argument for a ${extract expansion item is the key name or field
4616 number. Leading and trailing spaces in this item were not being ignored,
4617 causing some misleading effects.
4619 5. When deliver_drop_privilege was set, single queue runner processes started
4620 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4621 same command in the process it spins off) were not dropping privilege.
4623 6. When the daemon running as "exim" started a queue runner, it always
4624 re-executed Exim in the spun-off process. This is a waste of effort when
4625 deliver_drop_privilege is set. The new process now just calls the
4626 queue-runner function directly.
4632 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4635 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4636 ignored. Though the use of -oP was forcing the writing of a pid file, it
4637 was always written to the default place.
4639 3. If the message "no IP address found for host xxxx" is generated during
4640 incoming verification, it is now followed by identification of the incoming
4641 connection (so you can more easily find what provoked it).
4643 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4645 5. Added some features to "harden" Exim a bit more against certain attacks:
4647 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4648 be put in Local/Makefile. This is like the never_users runtime option,
4649 but it cannot be overridden. The default setting is "root".
4651 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4652 prefix string with which any file named in a -C command line option
4655 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4656 is retained for -C and -D only if the caller of Exim is root. Without
4657 it, the exim user may also use -C and -D and retain privilege.
4659 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4660 command line option is disabled.
4662 6. Macro names set by the -D option must start with an upper case letter, just
4663 like macro names defined in the configuration file.
4665 7. Added "dereference=" facility to LDAP.
4667 8. Two instances of the typo "uknown" in the source files are fixed.
4669 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4670 the Configure-Makefile script screwed up while processing it.
4672 10. Incorporated PCRE 4.4.
4674 11. The SMTP synchronization check was not operating right at the start of an
4675 SMTP session. For example, it could not catch a HELO sent before the client
4676 waited for the greeting. There is now a check for outstanding input at the
4677 point when the greeting is written. Because of the duplex, asynchronous
4678 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4679 way, but not yet received, when the check is performed.
4681 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4682 on TCP/IP sockets, because this apparently causes some broken clients to
4685 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4686 unchanged) from the Cygwin maintainer.
4688 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4689 was supported instead of testing for mbx. Effectively a typo.
4691 15. The spa authenticator server code was not checking that the input it
4692 received was valid base64.
4694 16. The debug output line for the "set" modifier in ACLs was not showing the
4695 name of the variable that was being set.
4697 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4699 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4700 lookup on its own. Something else was needed to trigger the lookup. For
4701 example, a match in host_lookup or the need for a host name in a host list.
4702 Now, if $sender_host_name is referenced and the host name has not yet been
4703 looked up, a lookup is performed. If the lookup fails, the variable remains
4704 empty, and $host_lookup_failed is set to "1".
4706 19. Added "eqi" as a case-independent comparison operator.
4708 20. The saslauthd authentication condition could segfault if neither service
4709 nor realm was specified.
4711 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4712 error message that was logged was misleading, and incoming SMTP
4713 connections were dropped. The message is now more accurate, and temporary
4714 errors are given to SMTP connections.
4716 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4717 (except RSET) with a 421 error, until QUIT is received. However, it was
4718 failing to send a response to QUIT.
4720 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4721 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4722 ACL is now run after the helo_try_verify_hosts code.
4724 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4725 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4726 case-independent, so other case variants are also recognized). Apparently
4727 some systems use these upper case variants.
4729 25. If more than two messages were waiting for the same host, and a transport
4730 filter was specified for the transport, Exim sent two messages over the
4731 same TCP/IP connection, and then failed with "socket operation on non-
4732 socket" when it tried to send the third.
4734 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4736 27. The extern definition of crypt16() in expand.c was not being excluded when
4737 the OS had its own crypt16() function.
4739 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4740 as a preferred synonym for return_size_limit, both as an option and as an
4743 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4745 30. Changed the default configuration ACL to relax the local part checking rule
4746 for addresses that are not in any local domains. For these addresses,
4747 slashes and pipe symbols are allowed within local parts, but the sequence
4748 /../ is explicitly forbidden.
4750 31. SPA server authentication was not clearing the challenge buffer before
4753 32. log_message in a "warn" ACL statement was writing to the reject log as
4754 well as to the main log, which contradicts the documentation and doesn't
4755 seem right (because no rejection is happening). So I have stopped it.
4757 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4758 However, I am unable to do any testing of this.
4760 34. Fixed an infelicity in the appendfile transport. When checking directories
4761 for a mailbox, to see if any needed to be created, it was accidentally
4762 using path names with one or more superfluous leading slashes; tracing
4763 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4765 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4766 contents are added to the log line that is written for every discarded
4767 recipient. (Previously a log_message setting was ignored.)
4769 36. The ${quote: operator now quotes the string if it is empty.
4771 37. The install script runs exim in order to find its version number. If for
4772 some reason other than non-existence or emptiness, which it checks, it
4773 could not run './exim', it was installing it with an empty version number,
4774 i.e. as "exim-". This error state is now caught, and the installation is
4777 38. An argument was missing from the function that creates an error message
4778 when Exim fails to connect to the socket for saslauthd authentication.
4779 This could cause Exim to crash, or give a corrupted message.
4781 39. Added isip, isip4, and isip6 to ${if conditions.
4783 40. The ACL variables $acl_xx are now saved with the message, and can be
4784 accessed later in routers, transports, and filters.
4786 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4787 strings in the file are not string-expanded.
4789 42. If a MAIL command specified a SIZE value that was too large to fit into an
4790 int variable, the check against message_size_limit failed. Such values are
4791 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4792 day this will have to be increased, but I don't think I want to be around
4793 when emails are that large.
4800 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4801 iconv() is not standard in FreeBSD.
4803 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4804 IPv6 enabled. The observed symptom was a segmentation fault on return from
4805 the function os_common_find_running_interfaces() in src/os.c.
4807 3. In the check_special_case() function in daemon.c I had used "errno" as an
4808 argument name, which causes warnings on some systems. This was basically a
4809 typo, since it was named "eno" in the comments!
4811 4. The code that waits for the clock to tick (at a resolution of some fraction
4812 of a second) so as to ensure message-id uniqueness was always waiting for
4813 at least one whole tick, when it could have waited for less. [This is
4814 almost certainly not relevant at current processor speeds, where it is
4815 unlikely to ever wait at all. But we try to future-proof.]
4817 5. The function that sleeps for a time interval that includes fractions of a
4818 second contained a race. It did not block SIGALRM between setting the
4819 timer, and suspending (a couple of lines later). If the interval was short
4820 and the sigsuspend() was delayed until after it had expired, the suspension
4821 never ended. On busy systems this could lead to processes getting stuck for
4824 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4825 process, before it forks any delivery processes. The open lookup caching
4826 mechanism meant that the open file or database connection was passed into
4827 the delivery process. The problem was that delivery processes always tidy
4828 up cached lookup data. This could cause a problem for the next delivery
4829 process started by the queue runner, because the external queue runner
4830 process does not know about the closure. So the next delivery process
4831 still has data in the lookup cache. In the case of a file lookup, there was
4832 no problem because closing a file descriptor in a subprocess doesn't affect
4833 the parent. However, if the lookup was caching a connection to a database,
4834 the connection was closed, and the second delivery process was likely to
4835 see errors such as "PGSQL: query failed: server closed the connection
4836 unexpectedly". The problem has been fixed by closing all cached lookups
4837 in a queue runner before running a delivery process.
4839 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4840 seem to have the "const" qualifier which it has on other OS. I've
4843 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4844 *of the same type* (client or server) with the same public name that an
4845 error should be diagnosed.
4847 9. When Exim looked up a host name for an IP address, but failed to find the
4848 original IP address when looking up the host name (a safety check), it
4849 output the message "<ip address> does not match any IP for NULL", which was
4850 confusing, to say the least. The bug was that the host name should have
4851 appeared instead of "NULL".
4853 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4854 user that is built into the binary, and the -C or -D options is used, root
4855 privilege is dropped before the configuration file is read. In addition,
4856 logging is switched to stderr instead of the normal log files. If the
4857 configuration then re-defines the Exim user, the unprivileged environment
4858 is probably not what is expected, so Exim logs a panic warning message (but
4861 However, if deliver_drop_privilege is set, the unprivileged state may well
4862 be exactly what is intended, so the warning has been cut out in that case,
4863 and Exim is allowed to try to write to its normal log files.
4869 1. smtp_return_error_details was not giving details for temporary sender
4870 or receiver verification errors.
4872 2. Diagnose a configuration error if two authenticators have the same public
4875 3. Exim used not to create the message log file for a message until the first
4876 delivery attempt. This could be confusing when incoming messages were held
4877 for policy or load reasons. The message log file is now created at the time
4878 the message is received, and an initial "Received" line is written to it.
4880 4. The automatically generated man page for command line options had a minor
4881 bug that caused no ill effects; however, a more serious problem was that
4882 the procedure for building the man page automatically didn't always
4883 operate. Consequently, release 4.20 contains an out-of-date version. This
4884 shouldn't happen again.
4886 5. When building Exim with embedded Perl support, the script that builds the
4887 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4888 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4890 6. The freeze_tell option was not being used for messages that were frozen on
4891 arrival, either by an ACL or by local_scan().
4893 7. Added the smtp_incomplete_transaction log selector.
4895 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4896 was accepting AUTH without a new EHLO.
4898 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4899 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4902 10. Logging of TCP/IP connections (when configured) now happens in the main
4903 daemon process instead of the child process, so that the TCP/IP connection
4904 count is more accurate (but it can never be perfect).
4906 11. The use of "drop" in a nested ACL was not being handled correctly in the
4907 outer ACL. Now, if condition failure induced by the nested "drop" causes
4908 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4909 or "require"), the connection is dropped.
4911 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4912 that yield "discard" can now be used with an "accept" or a "discard" verb,
4913 but an error is generated for any others (because I can't see a useful way
4914 to define what should happen).
4916 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4917 are now processed in the same way as lines in the Exim configuration file.
4918 In particular, continuation lines are supported.
4920 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4922 15. Added -ti meaning -t -i.
4924 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4925 domains, and warn by logging if others are found.
4927 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4928 returned from GIFCONF when Exim is trying to find the list of interfaces on
4929 a host. The code in os.c has been modified to copy each ifreq to an aligned
4930 structure in all cases.
4932 Also, in some cases, the returned ifreq's were being copied to a 'struct
4933 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4934 means the last couple of bytes of an IPv6 address could be chopped if the
4935 ifreq contained only a normal sockaddr (14 bytes storage).
4937 18. Named domain lists were not supported in the hosts_treat_as_local option.
4938 An entry such as +xxxx was not recognized, and was treated as a literal
4941 19. Ensure that header lines added by a DATA ACL are included in the reject log
4942 if the ACL subsequently rejects the message.
4944 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4945 MD5 (which is deprecated).
4947 21. When testing a filter file using -bf, Exim was writing a message when it
4948 took the sender from a "From " line in the message, but it was not doing so
4949 when it took $return_path from a Return-Path: header line. It now does.
4951 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4952 with a valid header line field name (a series of printing characters
4953 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4955 23. Changed "disc" in the source to "disk" to conform to the documentation and
4956 the book and for uniformity.
4958 24. Ignore Sendmail's -Ooption=value command line item.
4960 25. When execve() failed while trying to run a command in a pipe transport,
4961 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4962 could be confused with a return value of 69 from the command itself. This
4963 has been changed to 127, the value the shell returns if it is asked to run
4964 a non-existent command. The wording for the related log line suggests a
4965 non-existent command as the problem.
4967 26. If received_header_text expands to an empty string, do not add a Received:
4968 header line to the message. (Well, it adds a token one on the spool, but
4969 marks it "old" so that it doesn't get used or transmitted.)
4971 27. Installed eximstats 1.28 (addition of -nt option).
4973 28. There was no check for failure on the call to getsockname() in the daemon
4974 code. This can fail if there is a shortage of resources on the system, with
4975 ENOMEM, for example. A temporary error is now given on failure.
4977 29. Contrary to the C standard, it seems that in some environments, the
4978 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4979 program. Exim now does this explicitly; it affects the formatting of
4980 timestamps using strftime().
4982 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4983 complaints. I've now initialized all the variables, to avoid this.
4985 32. Header lines added by a system filter were not being "seen" during
4986 transport-time rewrites.
4988 33. The info_callback() function passed to OpenSSL is set up with type void
4989 (*)(SSL *, int, int), as described somewhere. However, when calling the
4990 function (actually a macro) that sets it up, the type void(*)() is
4991 expected. I've put in a cast to prevent warnings from picky compilers.
4993 34. If a DNS black list lookup found a CNAME record, but there were no A
4994 records associated with the domain it pointed at, Exim crashed.
4996 35. If a DNS black list lookup returned more than one A record, Exim ignored
4997 all but the first. It now scans all returned addresses if a particular IP
4998 value is being sought. In this situation, the contents of the
4999 $dnslist_value variable are a list of all the addresses, separated by a
5002 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5003 to accept a host name and all its aliases if the forward lookup for any of
5004 them yielded the IP address of the incoming connection. Now it accepts only
5005 those names whose forward lookup yields the correct IP address. Any other
5006 names are discarded. This closes a loophole whereby a rogue DNS
5007 administrator could create reverse DNS records to break through a
5008 wildcarded host restriction in an ACL.
5010 37. If a user filter or a system filter that ran in a subprocess used any of
5011 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5012 the wrong values were passed to the pipe command ($thisaddress had the
5013 value of $0, $0 had the value of $1, etc). This bug was introduced by
5014 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5016 38. Improved the line breaking for long SMTP error messages from ACLs.
5017 Previously, if there was no break point between 40 and 75 characters, Exim
5018 left the rest of the message alone. Two changes have been made: (a) I've
5019 reduced the minimum length to 35 characters; (b) if it can't find a break
5020 point between 35 and 75 characters, it looks ahead and uses the first one
5021 that it finds. This may give the occasional overlong line, but at least the
5022 remaining text gets split now.
5024 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5025 file descriptors might be low, and that setting 1000 would always raise it.
5026 It turns out that in some environments, the limit is already over 1000 and
5027 that lowering it causes trouble. So now Exim takes care not to decrease it.
5029 40. When delivering a message, the value of $return_path is set to $sender_
5030 address at the start of routing (routers may change the value). By an
5031 oversight, this default was not being set up when an address was tested by
5032 -bt or -bv, which affected the outcome if any router or filter referred to
5035 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5036 log only when "message" or "log_message" are set. However, if one of the
5037 conditions was an address verification, or a call to a nested ACL, the
5038 messages generated by the underlying test were being passed through. This
5039 no longer happens. The underlying message is available in $acl_verify_
5040 message for both "message" and "log_message" expansions, so it can be
5041 passed through if needed.
5043 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5044 new expansion $bh_ to give the encoded byte string without charset
5045 translation. Translation happens only if iconv() is available; HAVE_ICONV
5046 indicates this at build time. HEADERS_CHARSET gives the charset to
5047 translate to; headers_charset can change it in the configuration, and
5048 "headers charset" can change it in an individual filter file.
5050 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5051 that creates RFC 2047 encoded "words" labels them as that charset instead
5052 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5053 expansion operator; (ii) when Exim creates a From: line for a local
5054 message; (iii) when a header line is rewritten to include a "phrase" part.
5056 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5057 buggy, causing it to skip the first lines of messages whose message ID
5058 ended in 'D'. This would not have bitten before Exim release 4.14, because
5059 message IDs were unlikely to end in 'D' before then. The effect was to have
5060 incorrect size information for certain domains.
5062 45. #include "config.h" was missing at the start of the crypt16.c module. This
5063 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5066 46. If there was a timeout during a "random" callout check, Exim treated it as
5067 a failure of the random address, and carried on sending RSET and the real
5068 address. If the delay was just some slowness somewhere, the response to the
5069 original RCPT would be taken as a response to RSET and so on, causing
5070 mayhem of various kinds.
5072 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5073 when I implemented it. It didn't allow for the fact that some option values
5074 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5075 the right test for positive values.
5077 48. Domain names in DNS records are case-independent. Exim always looks them up
5078 in lower case. Some resolvers return domain names in exactly the case they
5079 appear in the zone file, that is, they may contain uppercase letters. Not
5080 all resolvers do this - some return always lower case. Exim was treating a
5081 change of case by a resolver as a change of domain, similar to a widening
5082 of a domain abbreviation. This triggered its re-routing code and so it was
5083 trying to route what was effectively the same domain again. This normally
5084 caused routing to fail (because the router wouldn't handle the domain
5085 twice). Now Exim checks for this case specially, and just changes the
5086 casing of the domain that it ultimately uses when it transmits the message
5089 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5092 50. If a filter generated a file delivery with a non-absolute name (possible if
5093 no home directory exists for the router), the forbid_file option was not
5096 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5097 the existing equality matching.
5099 52. Exim was using ints instead of ino_t variables in some places where it was
5100 dealing with inode numbers.
5102 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5103 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5104 called TMPDIR, and if it finds it is different, it changes its value.
5106 54. The smtp_printf() function is now made available to local_scan() so
5107 additional output lines can be written before returning. There is also an
5108 smtp_fflush() function to enable the detection of a dropped connection.
5109 The variables smtp_input and smtp_batched_input are exported to
5112 55. Changed the default runtime configuration: the message "Unknown user"
5113 has been removed from the ACL, and instead placed on the localuser router,
5114 using the cannot_route_message feature. This means that any verification
5115 failures that generate their own messages won't get overridden. Similarly,
5116 the "Unrouteable address" message that was in the ACL for unverifiable
5117 relay addresses has also been removed.
5119 56. Added hosts_avoid_esmtp to the smtp transport.
5121 57. The exicyclog script was not checking for the esoteric option
5122 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5123 will work only if exicyclog is run under the appropriate euid.
5125 58. Following a discussion on the list, the rules by which Exim recognises line
5126 endings on incoming messages have been changed. The -dropcr and drop_cr
5127 options are now no-ops, retained only for backwards compatibility. The
5128 following line terminators are recognized: LF CRLF CR. However, special
5129 processing applies to CR:
5131 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5132 nor a local message in the state where . is a terminator.
5134 (ii) If a bare CR is encountered in a header line, an extra space is added
5135 after the line terminator so as not to end the header. The reasoning
5136 behind this is that bare CRs in header lines are most likely either
5137 to be mistakes, or people trying to play silly games.
5139 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5140 was being incorrectly given as 18 bytes larger than it should have been.
5141 This is a VOB (very old bug).
5143 60. This may never have affected anything current, but just in case it has:
5144 When the local host is found other than at the start of a list of hosts,
5145 the local host, those with the same MX, and any that follow, are discarded.
5146 When the list in question was part of a longer list of hosts, the following
5147 hosts (not currently being processed) were also being discarded. This no
5148 longer happens. I'm not sure if this situation could ever has previously
5151 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5154 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5155 header. This is something that is recommended in a new Internet Draft, and
5156 is something that is documented as being done by Sendmail. There are two
5157 possible values. For messages generated by the autoreply transport, Exim
5160 Auto-Submitted: auto-replied
5162 whereas for all other generated messages (e.g. bounces) it adds
5164 Auto-Submitted: auto-generated
5166 63. The "personal" condition in filters now includes a test for the
5167 Auto-Submitted: header. If it contains the string "auto-" the message it
5168 not considered personal.
5170 64. Added rcpt_include_affixes as a generic transport option.
5172 65. Added queue_only_override (default true).
5174 66. Added the syslog_duplication option.
5176 67. If what should have been the first header line of a message consisted of
5177 a space followed by a colon, Exim was mis-interpreting it as a header line.
5178 It isn't of course - it is syntactically invalid and should therefore be
5179 treated as the start of the message body. The misbehaviour could have
5180 caused a number of strange effects, including loss of data in subsequent
5181 header lines, and spool format errors.
5183 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5184 client host had authenticated. This control can now be exercised by an ACL
5185 for more flexibility.
5187 69. By default, callouts do not happen when testing with -bh. There is now a
5188 variant, -bhc, which does actually run the callout code, including
5189 consulting and updating the callout cache.
5191 70. Added support for saslauthd authentication, courtesy of Alexander
5194 71. If statvfs() failed on the spool or log directories while checking their
5195 size for availability, Exim confusingly gave the error "space shortage".
5196 Furthermore, in debugging mode it crashed with a floating point exception.
5197 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5198 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5199 serious problem, Exim now writes to the main and panic logs when this
5200 happens, with details of the failure. It then refuses to accept the
5201 incoming message, giving the message "spool directory problem" or "log
5202 directory problem" with a 421 code for SMTP messages.
5204 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5205 0, 1, and 2 exist, because some OS complain for execs without them (see
5206 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5207 descriptors. However, the code omitted to check that the open succeeded,
5208 causing mysterious errors if for some reason the permissions on /dev/null
5209 got screwed. Now Exim writes a message to the main and panic logs, and
5210 bombs out if it can't open /dev/null.
5212 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5213 interact so that it is all more flexible. It is supposed to remain
5214 backwards compatible. Also added extra_local_interfaces.
5216 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5217 to bomb out with an assertion failure - to the client this appears as a
5218 connection drop. This problem occurs in the part of the code that was taken
5219 from the Samba project. Fortunately, the assertion is in a very simple
5220 function, so I have fixed this by reproducing the function inline in the
5221 one place where it is called, and arranging for authentication to fail
5222 instead of killing the process with assert().
5224 75. The SPA client code was not working when the server requested OEM rather
5225 than Unicode encoding.
5227 76. Added code to make require_files with a specific uid setting more usable in
5228 the case where statting the file as root fails - usually a non-root-mounted
5229 NFS file system. When this happens and the failure is EACCES, Exim now
5230 forks a subprocess and does the per-uid checking as the relevant uid.
5232 77. Added process_log_path.
5234 78. If log_file_path was not explicitly set, a setting of check_log_space or
5235 check_log_inodes was ignored.
5237 79. If a space check for the spool or log partitions fails, the incident is now
5238 logged. Of course, in the latter case the data may get lost...
5240 80. Added the %p formatting code to string_format() so that it can be used to
5241 print addresses in debug_print(). Adjusted all the address printing in the
5242 debugging in store.c to use %p rather than %d.
5244 81. There was a concern that a line of code in smtp_in.c could overflow a
5245 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5246 initially expressed, the concern was not well-founded, because trailing
5247 spaces are removed early. However, if the trailing spaces were followed by
5248 a NULL, they did not get removed, so the overflow was possible. Two fixes
5251 (a) I re-wrote the offending code in a cleaner fashion.
5252 (b) If an incoming SMTP command contains a NULL character, it is rejected
5255 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5256 runs initgroups(), so that if the Exim user is in any additional groups,
5257 they will be used during message reception.
5263 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob