1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
32 PP/01 Solaris build fix for Oracle's LDAP libraries.
33 Bugzilla 1109, patch from Stephen Usher.
35 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
37 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
40 TF/02 Fix a couple more cases where we did not log the error message
41 when unlink() failed. See also change 4.74-TF/03.
43 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
44 lock up or crash if it happened to be inside a call to libc when it
45 got a SIGUSR1 from exiwhat.
47 The SIGUSR1 handler appends the current process status to the process
48 log which is later printed by exiwhat. It used to use the general
49 purpose logging code to do this, but several functions it calls are
52 The new output code in the SIGUSR1 handler is specific to the process
53 log, and simple enough that it's easy to inspect for signal safety.
54 Removing some special cases also simplifies the general logging code.
55 Removing the spurious timestamps from the process log simplifies
58 TF/04 Improved ratelimit ACL condition.
60 The /noupdate option has been deprecated in favour of /readonly which
61 has clearer semantics. The /leaky, /strict, and /readonly update modes
62 are mutually exclusive. The update mode is no longer included in the
63 database key; it just determines when the database is updated. (This
64 means that when you upgrde Exim will forget old rate measurements.)
66 Exim now checks that the per_* options are used with an update mode that
67 makes sense for the current ACL. For example, when Exim is processing a
68 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
69 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
70 must specify per_mail/readonly. If you omit the update mode it defaults to
71 /leaky where that makes sense (as before) or /readonly where required.
73 The /noupdate option is now undocumented but still supported for
74 backwards compatibility. It is equivalent to /readonly except that in
75 ACLs where /readonly is required you may specify /leaky/noupdate or
76 /strict/noupdate which are treated the same as /readonly.
78 A useful new feature is the /count= option. This is a generalization
79 of the per_byte option, so that you can measure the throughput of other
80 aggregate values. For example, the per_byte option is now equivalent
81 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
83 The per_rcpt option has been generalized using the /count= mechanism
84 (though it's more complicated than the per_byte equivalence). When it is
85 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
86 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
87 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
88 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
89 Note that using per_rcpt with a non-readonly update mode in more than
90 one ACL will cause the recipients to be double-counted. (The per_mail
91 and per_byte options don't have this problem.)
93 The handling of very low rates has changed slightly. If the computed rate
94 is less than the event's count (usually one) then this event is the first
95 after a long gap. In this case the rate is set to the same as this event's
96 count, so that the first message of a spam run is counted properly.
98 The major new feature is a mechanism for counting the rate of unique
99 events. The new per_addr option counts the number of different
100 recipients that someone has sent messages to in the last time period. It
101 behaves like per_rcpt if all the recipient addresses are different, but
102 duplicate recipient addresses do not increase the measured rate. Like
103 the /count= option this is a general mechanism, so the per_addr option
104 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
105 example, measure the rate that a client uses different sender addresses
106 with the options per_mail/unique=$sender_address. There are further
107 details in the main documentation.
109 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
111 TF/06 Removed a few PCRE remnants.
113 TF/07 Automatically extract Exim's version number from tags in the git
114 repository when doing development or release builds.
116 PP/02 Raise smtp_cmd_buffer_size to 16kB.
117 Bugzilla 879. Patch from Paul Fisher.
119 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
120 Heavily based on revision 40f9a89a from Simon Arlott's tree.
123 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
125 PP/05 Variable $av_failed, true if the AV scanner deferred.
126 Bugzilla 1078. Patch from John Horne.
128 PP/06 Stop make process more reliably on build failure.
129 Bugzilla 1087. Patch from Heiko Schlittermann.
131 PP/07 Make maildir_use_size_file an _expandable_ boolean.
132 Bugzilla 1089. Patch from Heiko Schlittermann.
134 PP/08 Handle ${run} returning more data than OS pipe buffer size.
135 Bugzilla 1131. Patch from Holger Weiß.
137 PP/09 Handle IPv6 addresses with SPF.
138 Bugzilla 860. Patch from Wolfgang Breyha.
140 PP/10 GnuTLS: support TLS 1.2 & 1.1.
142 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
145 PP/11 match_* no longer expand right-hand-side by default.
146 New compile-time build option, EXPAND_LISTMATCH_RHS.
147 New expansion conditions, "inlist", "inlisti".
149 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
151 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
153 PP/14 fix log_write() format string regression from TF/03.
154 Bugzilla 1152. Patch from Dmitry Isaikin.
160 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
162 PP/02 Harmonised TLS library version reporting; only show if debugging.
163 Layout now matches that introduced for other libraries in 4.74 PP/03.
165 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
167 PP/04 New "dns_use_edns0" global option.
169 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
172 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
173 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
175 TK/01 Updated PolarSSL code to 0.14.2.
176 Bugzilla 1097. Patch from Andreas Metzler.
178 PP/07 Catch divide-by-zero in ${eval:...}.
181 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
184 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
185 format-string attack -- SECURITY: remote arbitrary code execution.
187 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
188 time unintentionally subject to list matching rules, letting the header
189 cause arbitrary Exim lookups (of items which can occur in lists, *not*
190 arbitrary string expansion). This allowed for information disclosure.
192 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
193 INT_MIN/-1 -- value coerced to INT_MAX.
199 NM/01 Workround for PCRE version dependency in version reporting
202 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
203 This fixes portability to compilers other than gcc, notably
204 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
206 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
207 makefiles for portability to HP-UX and POSIX correctness.
209 PP/01 Permit LOOKUP_foo enabling on the make command-line.
210 Also via indented variable definition in the Makefile.
211 (Debugging by Oliver Heesakkers).
213 PP/02 Restore caching of spamd results with expanded spamd_address.
214 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
216 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
217 Improves build reliability. Fix from: Frank Elsner
219 NM/02 Fix wide character breakage in the rfc2047 coding
220 Fixes bug 1064. Patch from Andrey N. Oktyabrski
222 NM/03 Allow underscore in dnslist lookups
223 Fixes bug 1026. Patch from Graeme Fowler
225 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
226 Code patches from Adam Ciarcinski of NetBSD.
228 NM/04 Fixed exiqgrep to cope with mailq missing size issue
231 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
232 is logged, to avoid truncation. Patch from John Horne.
234 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
235 Patch from Jakob Hirsch.
237 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
238 SQL string expansion failure details.
239 Patch from Andrey Oktyabrski.
241 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
242 Patch from Simon Arlott.
244 PP/09 New lookups functionality failed to compile on old gcc which rejects
245 extern declarations in function scope.
246 Patch from Oliver Fleischmann
248 PP/10 Use sig_atomic_t for flags set from signal handlers.
249 Check getgroups() return and improve debugging.
250 Fixed developed for diagnosis in bug 927 (which turned out to be
253 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
254 Patch from Mark Zealey.
256 PP/12 Bugzilla 1056: Improved spamd server selection.
257 Patch from Mark Zealey.
259 PP/13 Bugzilla 1086: Deal with maildir quota file races.
260 Based on patch from Heiko Schlittermann.
262 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
263 Patch from Uwe Doering, sign-off by Michael Haardt.
265 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
266 variable declaration deep within a block. Bug and patch from
269 PP/15 lookups-Makefile IRIX compatibilty coercion.
271 PP/16 Make DISABLE_DKIM build knob functional.
273 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
274 Patch by Simon Arlott
276 TF/03 Fix valgrind.h portability to C89 compilers that do not support
277 variable argument macros. Our copy now differs from upstream.
283 TF/01 Failure to get a lock on a hints database can have serious
284 consequences so log it to the panic log.
286 TF/02 Log LMTP confirmation messages in the same way as SMTP,
287 controlled using the smtp_confirmation log selector.
289 TF/03 Include the error message when we fail to unlink a spool file.
291 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
292 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
293 for maintaining out-of-tree patches for some time.
295 PP/01 Bugzilla 139: Documentation and portability issues.
296 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
297 Handle per-OS dynamic-module compilation flags.
299 PP/02 Let /dev/null have normal permissions.
300 The 4.73 fixes were a little too stringent and complained about the
301 permissions on /dev/null. Exempt it from some checks.
302 Reported by Andreas M. Kirchwitz.
304 PP/03 Report version information for many libraries, including
305 Exim version information for dynamically loaded libraries. Created
306 version.h, now support a version extension string for distributors
307 who patch heavily. Dynamic module ABI change.
309 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
310 privilege escalation vulnerability whereby the Exim run-time user
311 can cause root to append content of the attacker's choosing to
314 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
317 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
318 If dropping privileges for untrusted macros, we disabled normal logging
319 on the basis that it would fail; for the Exim run-time user, this is not
320 the case, and it resulted in successful deliveries going unlogged.
321 Fixed. Reported by Andreas Metzler.
327 PP/01 Date: & Message-Id: revert to normally being appended to a message,
328 only prepend for the Resent-* case. Fixes regression introduced in
329 Exim 4.70 by NM/22 for Bugzilla 607.
331 PP/02 Include check_rfc2047_length in configure.default because we're seeing
332 increasing numbers of administrators be bitten by this.
334 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
336 PP/03 Bugzilla 994: added openssl_options main configuration option.
338 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
340 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
342 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
344 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
345 without a peer certificate, leading to a segfault because of an
346 assumption that peers always have certificates. Be a little more
347 paranoid. Problem reported by Martin Tscholak.
349 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
350 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
351 NB: ClamAV planning to remove STREAM in "middle of 2010".
352 CL also introduces -bmalware, various -d+acl logging additions and
353 more caution in buffer sizes.
355 PP/09 Implemented reverse_ip expansion operator.
357 PP/10 Bugzilla 937: provide a "debug" ACL control.
359 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
361 PP/12 Bugzilla 973: Implement --version.
363 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
365 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
367 PP/15 Bugzilla 816: support multiple condition rules on Routers.
369 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
370 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
371 ignore trailing whitespace.
373 JJ/02 prevent non-panic DKIM error from being sent to paniclog
375 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
378 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
379 Notification from Dr Andrew Aitchison.
381 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
382 ExtendedDetectionInfo response format.
383 Notification from John Horne.
385 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
388 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
389 XSL and documented dependency on system catalogs, with examples of how
392 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
395 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
396 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
397 configuration file which is writeable by the Exim user or group.
399 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
400 of configuration files to cover files specified with the -C option if
401 they are going to be used with root privileges, not just the default
404 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
405 option (effectively making it always true).
407 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
408 files to be used while preserving root privileges.
410 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
411 that rogue child processes cannot use them.
413 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
414 run-time user, instead of root.
416 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
417 Exim run-time user without dropping privileges.
419 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
420 result string, instead of calling string_vformat() twice with the same
423 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
424 for other users. Others should always drop root privileges if they use
425 -C on the command line, even for a whitelisted configure file.
427 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
429 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
435 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
436 $data_path, and $header_path variables; fixed documentation bugs and
439 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
440 exipick to access non-standard spools, including the "frozen" queue
443 NM/01 Bugzilla 965: Support mysql stored procedures.
444 Patch from Alain Williams
446 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
448 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
449 Patch from Andreas Metzler
451 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
452 Patch from Kirill Miazine
454 NM/05 Bugzilla 671: Added umask to procmail example.
456 JJ/03 installed exipick 20100323.0, fixing doc bug
458 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
459 directory. Notification and patch from Dan Rosenberg.
461 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
463 TK/02 Improve log output when DKIM signing operation fails.
465 MH/01 Treat the transport option dkim_domain as a colon separated
466 list, not as a single string, and sign the message with each element,
467 omitting multiple occurences of the same signer.
469 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
470 Bugzilla 985, 986. Patch by Simon Arlott
472 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
473 Patch by Simon Arlott
475 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
476 MBX locking. Notification from Dan Rosenberg.
482 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
484 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
486 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
488 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
490 NM/04 Bugzilla 915: Flag broken perl installation during build.
496 TK/01 Added patch by Johannes Berg that expands the main option
497 "spamd_address" if it starts with a dollar sign.
499 TK/02 Write list of recipients to X-Envelope-Sender header when building
500 the mbox-format spool file for content scanning (suggested by Jakob
503 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
504 (http://www.dcc-servers.net/) support via dccifd. Activated by
505 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
507 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
508 by Mark Daniel Reidel <mr@df.eu>.
510 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
511 When building exim an external PCRE library is now needed -
512 PCRE is a system library on the majority of modern systems.
513 See entry on PCRE_LIBS in EDITME file.
515 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
516 conversation. Added nologin parameter to request.
517 Patch contributed by Kirill Miazine.
519 TF/01 Do not log submission mode rewrites if they do not change the address.
521 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
523 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
524 log files in place. Contributed by Roberto Lima.
526 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
528 TF/03 Bugzilla 615: When checking the local_parts router precondition
529 after a local_part_suffix or local_part_prefix option, Exim now
530 does not use the address's named list lookup cache, since this
531 contains cached lookups for the whole local part.
533 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
534 Robert Millan. Documentation is in experimental-spec.txt.
536 TF/04 Bugzilla 668: Fix parallel build (make -j).
538 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
540 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
541 Patch provided by Jan Srzednicki.
543 TF/05 Leading white space used to be stripped from $spam_report which
544 wrecked the formatting. Now it is preserved.
546 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
547 that they are available at delivery time.
549 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
551 TF/08 TLS error reporting now respects the incoming_interface and
552 incoming_port log selectors.
554 TF/09 Produce a more useful error message if an SMTP transport's hosts
555 setting expands to an empty string.
557 NM/06 Bugzilla 744: EXPN did not work under TLS.
558 Patch provided by Phil Pennock.
560 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
561 Patch provided by Richard Godbee.
563 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
564 acl_smtp_notquit, added index entry.
566 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
567 Patch provided by Eugene Bujak.
569 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
570 accept(). Patch provided by Maxim Dounin.
572 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
573 Patch provided by Phil Pennock.
575 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
577 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
578 Patch provided by Brad "anomie" Jorsch.
580 NM/14 Bugzilla 622: Added timeout setting to transport filter.
581 Patch provided by Dean Brooks.
583 TK/05 Add native DKIM support (does not depend on external libraries).
585 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
586 Patch provided by Graeme Fowler.
588 NM/16 Bugzilla 851: Documentation example syntax fix.
590 NM/17 Changed NOTICE file to remove references to embedded PCRE.
592 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
595 NM/19 Bugzilla 745: TLS version reporting.
596 Patch provided by Phil Pennock.
598 NM/20 Bugzilla 167: bool: condition support.
599 Patch provided by Phil Pennock.
601 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
602 clients. Patch provided by Phil Pennock.
604 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
605 Patch provided by Brad "anomie" Jorsch.
607 NM/23 Bugzilla 687: Fix misparses in eximstats.
608 Patch provided by Heiko Schlittermann.
610 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
611 Patch provided by Heiko Schlittermann.
613 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
614 plus update to original patch.
616 NM/26 Bugzilla 799: Documentation correction for ratelimit.
618 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
619 Patch provided by David Brownlee.
621 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
623 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
625 NM/30 Bugzilla 888: TLS documentation bugfixes.
627 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
629 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
630 Unlike the original bugzilla I have changed all shell scripts in src tree.
632 NM/33 Bugzilla 898: Transport filter timeout fix.
633 Patch by Todd Rinaldo.
635 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
636 Patch by Serge Demonchaux.
638 NM/35 Bugzilla 39: Base64 decode bug fixes.
639 Patch by Jakob Hirsch.
641 NM/36 Bugzilla 909: Correct connect() call in dcc code.
643 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
645 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
647 NM/39 Bugzilla 911: Fixed MakeLinks build script.
653 TK/01 Add preliminary DKIM support. Currently requires a forked version of
654 ALT-N's libdkim that I have put here:
655 http://duncanthrax.net/exim-experimental/
657 Note to Michael Haardt: I had to rename some vars in sieve.c. They
658 were called 'true' and it seems that C99 defines that as a reserved
659 keyword to be used with 'bool' variable types. That means you could
660 not include C99-style headers which use bools without triggering
661 build errors in sieve.c.
663 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
664 as mailq or other aliases. Changed the --help handling significantly
665 to do whats expected. exim_usage() emits usage/help information.
667 SC/01 Added the -bylocaldomain option to eximstats.
669 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
671 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
673 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
679 PH/01 Another patch from the Sieve maintainer.
681 PH/02 When an IPv6 address is converted to a string for single-key lookup
682 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
683 dots are used instead of colons so that keys in lsearch files need not
684 contain colons. This was done some time before quoting was made available
685 in lsearch files. However, iplsearch files do require colons in IPv6 keys
686 (notated using the quote facility) so as to distinguish them from IPv4
687 keys. This meant that lookups for IP addresses in host lists did not work
688 for iplsearch lookups.
690 This has been fixed by arranging for IPv6 addresses to be expressed with
691 colons if the lookup type is iplsearch. This is not incompatible, because
692 previously such lookups could never work.
694 The situation is now rather anomolous, since one *can* have colons in
695 ordinary lsearch keys. However, making the change in all cases is
696 incompatible and would probably break a number of configurations.
698 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
701 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
702 conversion specification without a maximum field width, thereby enabling
703 a rogue spamd server to cause a buffer overflow. While nobody in their
704 right mind would setup Exim to query an untrusted spamd server, an
705 attacker that gains access to a server running spamd could potentially
706 exploit this vulnerability to run arbitrary code as the Exim user.
708 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
709 $primary_hostname instead of what libspf2 thinks the hosts name is.
711 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
712 a directory entry by the name of the lookup key. Previously, if a
713 symlink pointed to a non-existing file or a file in a directory that
714 Exim lacked permissions to read, a lookup for a key matching that
715 symlink would fail. Now it is enough that a matching directory entry
716 exists, symlink or not. (Bugzilla 503.)
718 PH/03 The body_linecount and body_zerocount variables are now exported in the
721 PH/04 Added the $dnslist_matched variable.
723 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
724 This means they are set thereafter only if the connection becomes
727 PH/06 Added the client_condition to authenticators so that some can be skipped
728 by clients under certain conditions.
730 PH/07 The error message for a badly-placed control=no_multiline_responses left
731 "_responses" off the end of the name.
733 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
735 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
736 (without spaces) instead of just copying the configuration text.
738 PH/10 Added the /noupdate option to the ratelimit ACL condition.
740 PH/11 Added $max_received_linelength.
742 PH/12 Added +ignore_defer and +include_defer to host lists.
744 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
745 way in which PCRE > 7.0 is built.
747 PH/14 Implemented queue_only_load_latch.
749 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
750 MAIL command. The effect was to mangle the value on 64-bit systems.
752 PH/16 Another patch from the Sieve maintainer.
754 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
756 PH/18 If a system quota error occurred while trying to create the file for
757 a maildir delivery, the message "Mailbox is full" was not appended to the
758 bounce if the delivery eventually timed out. Change 4.67/27 below applied
759 only to a quota excession during the actual writing of the file.
761 PH/19 It seems that peer DN values may contain newlines (and other non-printing
762 characters?) which causes problems in log lines. The DN values are now
763 passed through string_printing() before being added to log lines.
765 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
766 and InterBase are left for another time.)
768 PH/21 Added message_body_newlines option.
770 PH/22 Guard against possible overflow in moan_check_errorcopy().
772 PH/23 POSIX allows open() to be a macro; guard against that.
774 PH/24 If the recipient of an error message contained an @ in the local part
775 (suitably quoted, of course), incorrect values were put in $domain and
776 $local_part during the evaluation of errors_copy.
782 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
783 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
786 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
787 issue a MAIL command.
789 PH/02 In an ACL statement such as
791 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
793 if a client was not listed at all, or was listed with a value other than
794 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
795 the condition was not true (as it should be), so access was not denied.
796 The bug was that the ! inversion was incorrectly passed on to the second
797 item. This has been fixed.
799 PH/03 Added additional dnslists conditions == and =& which are different from
800 = and & when the dns lookup returns more than one IP address.
802 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
803 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
805 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
806 FSYNC, which compiles an option called disable_fsync that allows for
807 bypassing fsync(). The documentation is heavily laced with warnings.
809 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
811 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
812 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
813 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
814 including adding "make clean"; (3) Added -fPIC when compiling the test
815 dynamically loaded module, to get rid of a warning.
817 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
818 message fails, move_frozen_messages = true and ignore_bounce_errors_after
819 = 0s. The bug is otherwise harmless.
821 PH/07 There was a bug in the dovecot authenticator such that the value of
822 $auth1 could be overwritten, and so not correctly preserved, after a
823 successful authentication. This usually meant that the value preserved by
824 the server_setid option was incorrect.
826 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
828 PH/09 Installed PCRE release 7.0.
830 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
831 run for batched SMTP input. It is now run at the start of every message
832 in the batch. While fixing this I discovered that the process information
833 (output by running exiwhat) was not always getting set for -bs and -bS
834 input. This is fixed, and it now also says "batched" for BSMTP.
836 PH/11 Added control=no_pipelining.
838 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
839 patch, slightly modified), and move the expansion of helo_data till after
840 the connection is made in the smtp transport (so it can use these
843 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
845 PH/14 Added log_selector = +pid.
847 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
849 PH/16 Add ${if forany and ${if forall.
851 PH/17 Added dsn_from option to vary the From: line in DSNs.
853 PH/18 Flush SMTP output before performing a callout, unless control =
854 no_callout_flush is set.
856 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
857 was true (the default) a successful delivery failed to delete the retry
858 item, thus causing premature timeout of the address. The bug is now
861 PH/20 Added hosts_avoid_pipelining to the smtp transport.
863 PH/21 Long custom messages for fakedefer and fakereject are now split up
864 into multiline reponses in the same way that messages for "deny" and
865 other ACL rejections are.
867 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
868 with slight modification.
870 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
871 draft, changing the syntax and factoring some duplicate code".
873 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
874 for deliveries of the second and subsequent messages over the same SMTP
877 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
878 ${reduce, with only minor "tidies".
880 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
882 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
883 expansion side effects.
885 PH/27 When a message times out after an over-quota error from an Exim-imposed
886 quota, the bounce message says "mailbox is full". This message was not
887 being given when it was a system quota that was exceeded. It now should
890 MH/03 Made $recipients available in local_scan(). local_scan() already has
891 better access to the recipient list through recipients_list[], but
892 $recipients can be useful in postmaster-provided expansion strings.
894 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
895 in the case of a MAIL command with additional options following the
896 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
897 were accidentally chopped off.
899 PH/29 SMTP synchronization checks are implemented when a command is read -
900 there is a check that no more input is waiting when there shouldn't be
901 any. However, for some commands, a delay in an ACL can mean that it is
902 some time before the response is written. In this time, more input might
903 arrive, invalidly. So now there are extra checks after an ACL has run for
904 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
905 pipelining has not been advertised.
907 PH/30 MH's patch to allow iscntrl() characters to be list separators.
909 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
910 returned in the SMTP response when smtp_return_error_details was false.
913 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
914 instead of the C I/O that was originally supplied, because problems were
917 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
918 Exim which did not show up earlier: it was assuming that a call to
919 SSL_CTX_set_info_callback() might give an error value. In fact, there is
920 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
921 was a macro that became an assignment, so it seemed to work. This has
922 changed to a proper function call with a void return, hence the compile
923 error. Exim's code has been fixed.
925 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
928 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
930 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
932 PH/37 If a message is not accepted after it has had an id assigned (e.g.
933 because it turns out to be too big or there is a timeout) there is no
934 "Completed" line in the log. When some messages of this type were
935 selected by exigrep, they were listed as "not completed". Others were
936 picked up by some special patterns. I have improved the selection
937 criteria to be more general.
939 PH/38 The host_find_failed option in the manualroute router can now be set
940 to "ignore", to completely ignore a host whose IP address cannot be
941 found. If all hosts are ignored, the behaviour is controlled by the new
942 host_all_ignored option.
944 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
945 homing or because of multiple MX records with /mx) generated more than
946 one IP address, and the following item turned out to be the local host,
947 all the secondary addresses of the first item were incorrectly removed
948 from the list, along with the local host and any following hosts (which
949 is what is supposed to happen).
951 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
952 whoever called Exim into the -H file. In the case of the daemon it was
953 behaving confusingly. When first started, it used values for whoever
954 started the daemon, but after a SIGHUP it used the Exim user (because it
955 calls itself on a restart). I have changed the code so that it now always
958 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
959 message are rejected with the same error (e.g. no authentication or bad
960 sender address), and a DATA command is nevertheless sent (as can happen
961 with PIPELINING or a stupid MUA), the error message that was given to the
962 RCPT commands is included in the rejection of the DATA command. This is
963 intended to be helpful for MUAs that show only the final error to their
966 PH/42 Another patch from the Sieve maintainer.
968 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
969 Eximstats - Fixed some broken HTML links and added missing column headers
971 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
972 columns for Rejects, Temp Rejects, Ham, and Spam rows.
974 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
976 PH/43 Yet another patch from the Sieve maintainer.
978 PH/44 I found a way to check for a TCP/IP connection going away before sending
979 the response to the final '.' that terminates a message, but only in the
980 case where the client has not sent further data following the '.'
981 (unfortunately, this is allowed). However, in many cases there won't be
982 any further data because there won't be any more messages to send. A call
983 to select() can be used: if it shows that the input is "ready", there is
984 either input waiting, or the socket has been closed. An attempt to read
985 the next input character can distinguish the two cases. Previously, Exim
986 would have sent an OK response which the client would never have see.
987 This could lead to message repetition. This fix should cure that, at
988 least in a lot of common cases.
990 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
991 advertised in response to EHLO.
997 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
998 fixed by 4.65/MH/01 (is this a record?) are fixed:
1000 (i) An empty string was always treated as zero by the numeric comparison
1001 operators. This behaviour has been restored.
1003 (ii) It is documented that the numeric comparison operators always treat
1004 their arguments as decimal numbers. This was broken in that numbers
1005 starting with 0 were being interpreted as octal.
1007 While fixing these problems I realized that there was another issue that
1008 hadn't been noticed. Values of message_size_limit (both the global option
1009 and the transport option) were treated as octal if they started with 0.
1010 The documentation was vague. These values are now always treated as
1011 decimal, and I will make that clear in the documentation.
1017 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1018 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1021 MH/01 Don't check that the operands of numeric comparison operators are
1022 integers when their expansion is in "skipping" mode (fixes bug
1023 introduced by 4.64-PH/07).
1025 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1026 child addresses, Exim now panics and dies. Previously, because the count
1027 is held in a short int, deliveries were likely to be lost. As such a
1028 large number of recipients for a single message is ridiculous
1029 (performance will be very, very poor), I have chosen to impose a limit
1030 rather than extend the field.
1036 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1037 leftover -K file (the existence of which was triggered by #402).
1038 While we were at it, introduced process PID as part of the -K
1039 filename. This should rule out race conditions when creating
1042 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1043 processing considerably. Previous code took too long for large mails,
1044 triggering a timeout which in turn triggers #401.
1046 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1047 in the DK code in transports.c. sendfile() is not really portable,
1048 hence the _LINUX specificness.
1050 TF/01 In the add_headers option to the mail command in an Exim filter,
1051 there was a bug that Exim would claim a syntax error in any
1052 header after the first one which had an odd number of characters
1055 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1056 callout verification, Exim cached a "reject" for the entire domain. This
1057 is correct for most verifications, but it is not correct for a recipient
1058 verification with use_sender or use_postmaster set, because in that case
1059 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1060 case of MAIL FROM:<> rejection from other early rejections (e.g.
1061 rejection of HELO). When verifying a recipient using a non-null MAIL
1062 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1063 Whatever the result of the callout, the value of the domain cache is
1064 left unchanged (for any other kind of callout, getting as far as trying
1065 RCPT means that the domain itself is ok).
1067 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1070 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1071 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1072 session, the daemon ignores ECONNECT errors and logs others; it now
1073 ignores EPIPE as well.
1075 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1076 (quoted-printable decoding).
1078 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1079 later the small subsequent patch to fix an introduced bug.
1081 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1083 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1085 PH/08 An error is now given if message_size_limit is specified negative.
1087 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1088 to be given (somewhat) arbitrary names.
1090 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1093 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1094 miscellaneous code fixes
1096 PH/10 Added the log_reject_target ACL modifier to specify where to log
1099 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1100 hostname. This is wrong, because it relates to the incoming message (and
1101 probably the interface on which it is arriving) and not to the outgoing
1102 callout (which could be using a different interface). This has been
1103 changed to use the value of the helo_data option from the smtp transport
1104 instead - this is what is used when a message is actually being sent. If
1105 there is no remote transport (possible with a router that sets up host
1106 addresses), $smtp_active_hostname is used.
1108 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1109 tweaks were necessary in order to get it to work (see also 21 below):
1110 (a) The code assumed that strncpy() returns a negative number on buffer
1111 overflow, which isn't the case. Replaced with Exim's string_format()
1113 (b) There were several signed/unsigned issues. I just did the minimum
1114 hacking in of casts. There is scope for a larger refactoring.
1115 (c) The code used strcasecmp() which is not a standard C function.
1116 Replaced with Exim's strcmpic() function.
1117 (d) The code set only $1; it now sets $auth1 as well.
1118 (e) A simple test gave the error "authentication client didn't specify
1119 service in request". It would seem that Dovecot has changed its
1120 interface. Fortunately there's a specification; I followed it and
1121 changed what the client sends and it appears to be working now.
1123 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1126 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1127 address A is aliased to B and C, where B exists and C does not. Without
1128 -v the output is "A verified" because verification stops after a
1129 successful redirection if more than one address is generated. However,
1130 with -v the child addresses are also verified. Exim was outputting "A
1131 failed to verify" and then showing the successful verification for C,
1132 with its parentage. It now outputs "B failed to verify", showing B's
1133 parentage before showing the successful verification of C.
1135 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1136 look up a TXT record in a specific list after matching in a combined
1139 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1140 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1141 they consult the DNS. I had assumed they would set it the way they
1142 wanted; and indeed my experiments on Linux seem to show that in some
1143 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1144 To be on the safe side, however, I have now made the interface to
1145 host_find_byname() similar to host_find_bydns(), with an argument
1146 containing the DNS resolver options. The host_find_byname() function now
1147 sets these options at its start, just as host_find_bydns() does. The smtp
1148 transport options dns_qualify_single and dns_search_parents are passed to
1149 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1150 of host_find_byname() use the default settings of RES_DEFNAMES
1151 (qualify_single) but not RES_DNSRCH (search_parents).
1153 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1154 spool_read_header() do less string testing, by means of a preliminary
1155 switch on the second character of optional "-foo" lines. (This is
1156 overdue, caused by the large number of possibilities that now exist.
1157 Originally there were few.) While I was there, I also converted the
1158 str(n)cmp tests so they don't re-test the leading "-" and the first
1159 character, in the hope this might squeeze out yet more improvement.
1161 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1162 flag allowing group syntax was set by the header_syntax check but not
1163 turned off, possible causing trouble later; (2) The flag was not being
1164 set at all for the header_verify test, causing "group"-style headers to
1165 be rejected. I have now set it in this case, and also caused header_
1166 verify to ignore an empty address taken from a group. While doing this, I
1167 came across some other cases where the code for allowing group syntax
1168 while scanning a header line wasn't quite right (mostly, not resetting
1169 the flag correctly in the right place). These bugs could have caused
1170 trouble for malformed header lines. I hope it is now all correct.
1172 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1173 with the "reply" argument non-NULL. The code, however (which originally
1174 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1175 but it didn't always do it. This confused somebody who was copying the
1176 code for some other use. I have removed all the tests.
1178 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1179 feature that was used to support insecure browsers during the U.S. crypto
1180 embargo. It requires special client support, and Exim is probably the
1181 only MTA that supported it -- and would never use it because real RSA is
1182 always available. This code has been removed, because it had the bad
1183 effect of slowing Exim down by computing (never used) parameters for the
1184 RSA_EXPORT functionality.
1186 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1187 authenticator to fail if there's a tab character in the incoming data
1188 (there should never be unless someone is messing about, as it's supposed
1189 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1190 if the connection is using TLS or if the remote IP is the same as the
1191 local IP, and the "valid-client-cert option" if a client certificate has
1194 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1195 authenticators. This can be used for authorization after authentication
1196 succeeds. (In the case of plaintext, it servers for both authentication
1199 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1200 if any retry times were supplied.
1202 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1203 connection, where there is no client IP address to check. In this
1204 situation, the verify now always succeeds.
1206 PH/25 Applied John Jetmore's -Mset patch.
1208 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1210 PH/27 In a string expansion for a processed (not raw) header when multiple
1211 headers of the same name were present, leading whitespace was being
1212 removed from all of them, but trailing whitespace was being removed only
1213 from the last one. Now trailing whitespace is removed from each header
1214 before concatenation. Completely empty headers in a concatenation (as
1215 before) are ignored.
1217 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1218 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1220 PH/29 [Removed. This was a change that I later backed out, and forgot to
1221 correct the ChangeLog entry (that I had efficiently created) before
1222 committing the later change.]
1224 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1225 address errors (4xx response to RCPT) over the same connection as other
1226 messages routed to the same hosts. Such deliveries are always "forced",
1227 so retry times are not inspected. This resulted in far too many retries
1228 for the affected addresses. The effect occurred only when there were more
1229 hosts than the hosts_max_try setting in the smtp transport when it had
1230 the 4xx errors. Those hosts that it had tried were not added to the list
1231 of hosts for which the message was waiting, so if all were tried, there
1232 was no problem. Two fixes have been applied:
1234 (i) If there are any address or message errors in an SMTP delivery, none
1235 of the hosts (tried or untried) are now added to the list of hosts
1236 for which the message is waiting, so the message should not be a
1237 candidate for sending over the same connection that was used for a
1238 successful delivery of some other message. This seems entirely
1239 reasonable: after all the message is NOT "waiting for some host".
1240 This is so "obvious" that I'm not sure why it wasn't done
1241 previously. Hope I haven't missed anything, but it can't do any
1242 harm, as the worst effect is to miss an optimization.
1244 (ii) If, despite (i), such a delivery is accidentally attempted, the
1245 routing retry time is respected, so at least it doesn't keep
1246 hammering the server.
1248 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1249 in ${readsocket because some servers need this prod.
1251 PH/32 Added some extra debug output when updating a wait-xxx database.
1253 PH/33 The hint "could be header name not terminated by colon", which has been
1254 given for certain expansion errors for a long time, was not being given
1255 for the ${if def:h_colon_omitted{... case.
1257 PH/34 The spec says: "With one important exception, whenever a domain list is
1258 being scanned, $domain contains the subject domain." There was at least
1259 one case where this was not true.
1261 PH/35 The error "getsockname() failed: connection reset by peer" was being
1262 written to the panic log as well as the main log, but it isn't really
1263 panic-worthy as it just means the connection died rather early on. I have
1264 removed the panic log writing for the ECONNRESET error when getsockname()
1267 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1268 runs only) independently of the message's sender address. This meant
1269 that, if the 4xx error was in fact related to the sender, a different
1270 message to the same recipient with a different sender could confuse
1271 things. In particualar, this can happen when sending to a greylisting
1272 server, but other circumstances could also provoke similar problems.
1273 I have changed the default so that the retry time for these errors is now
1274 based a combination of the sender and recipient addresses. This change
1275 can be overridden by setting address_retry_include_sender=false in the
1278 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1279 remote server are returned as part of bounce messages. This was not
1280 happening for LMTP over a pipe (the lmtp transport), but now it is the
1281 same for both kinds of LMTP.
1283 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1284 in header lines that were in fact CNAMEs. This is no longer the case.
1286 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1287 and queue runs started by the daemon processed all messages. This has
1288 been fixed so that -R and -S can now usefully be given with -q<time>.
1290 PH/40 Import PCRE release 6.7 (fixes some bugs).
1292 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1294 PH/42 Give an error if -q is specified more than once.
1296 PH/43 Renamed the variables $interface_address and $interface_port as
1297 $received_ip_address and $received_port, to make it clear that these
1298 values apply to message reception, and not to the outgoing interface when
1299 a message is delivered. (The old names remain recognized, of course.)
1301 PH/44 There was no timeout on the connect() call when using a Unix domain
1302 socket in the ${readsocket expansion. There now is.
1304 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1305 be meaningful with "accept".
1307 SC/01 Eximstats V1.43
1308 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1310 SC/02 Eximstats V1.44
1311 Use a glob alias rather than an array ref in the generated
1312 parser. This improves both readability and performance.
1314 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1315 Collect SpamAssassin and rejection statistics.
1316 Don't display local sender or destination tables unless
1317 there is data to show.
1318 Added average volumes into the top table text output.
1320 SC/04 Eximstats V1.46
1321 Collect data on the number of addresses (recipients)
1322 as well as the number of messages.
1324 SC/05 Eximstats V1.47
1325 Added 'Message too big' to the list of mail rejection
1326 reasons (thanks to Marco Gaiarin).
1328 SC/06 Eximstats V1.48
1329 Mainlog lines which have GMT offsets and are too short to
1330 have a flag are now skipped.
1332 SC/07 Eximstats V1.49 (Alain Williams)
1333 Added the -emptyok flag.
1335 SC/08 Eximstats V1.50
1336 Fixes for obtaining the IP address from reject messages.
1338 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1339 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1340 whitesspace changes from 4.64-PH/27
1342 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1345 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1346 are found, allow negative numbers in numeric criteria)
1348 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1350 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1351 to match changes made in 4.64-PH/43
1353 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1355 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1356 "domain" is actually an IP address, and force a failure. This locks out
1357 those revolvers/nameservers that support "A-for-A" lookups, in
1358 contravention of the specifications.
1360 PH/48 When a host name was looked up from an IP address, and the subsequent
1361 forward lookup of the name timed out, the host name was left in
1362 $sender_host_name, contrary to the specification.
1364 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1365 restricted to single-key lookups, Exim was not diagnosing an error if
1366 * or *@ was used with a query-style lookup.
1368 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1370 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1371 long ago, but noone interested enough thought of it. Let's just say that
1372 the "1.1" means that there are some new functions that weren't there at
1373 some point in the past.
1375 PH/51 Error processing for expansion failure of helo_data from an smtp
1376 transport during callout processing was broken.
1378 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1379 tested/used via the -bh/-bhc/-bs options.
1381 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1382 bug, fixed in subsequent PCRE releases).
1384 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1385 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1387 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1393 SC/01 Use a glob alias rather than an array ref in eximstats generated
1394 parser. This improves both readability and performance.
1396 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1397 Don't display local sender or destination tables in eximstats unless
1398 there is data to show.
1399 Added average volumes into the eximstats top table text output.
1401 SC/03 Collect data on the number of addresses (recipients) as well
1402 as the number of messages in eximstats.
1404 TF/01 Correct an error in the documentation for the redirect router. Exim
1405 does (usually) call initgroups() when daemonizing.
1407 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1408 with consistent privilege compared to when running as a daemon.
1410 TF/03 Note in the spec that $authenticated_id is not set for local
1411 submissions from trusted users.
1413 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1414 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1416 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1417 by adding some example configuration directives to the default
1418 configuration file. A little bit of work is required to uncomment the
1419 directives and define how usernames and passwords are checked, but
1420 there is now a framework to start from.
1422 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1423 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1424 without this. I don't know how relevant this is to other LDAP libraries.
1426 PH/02 Add the verb name to the "unknown ACL verb" error.
1428 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1430 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1432 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1433 directory not expanded when it should be if an expanded home directory
1434 was set for the address (which is overridden by the transport).
1436 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1439 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1440 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1441 because it is too late at that time, and has no effect.
1443 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1444 security issue with \' (bugzilla #107). I could not use the
1445 PQescapeStringConn() function, because it needs a PGconn value as one of
1448 PH/08 When testing addresses using -bt, indicate those final addresses that
1449 are duplicates that would not cause an additional delivery. At least one
1450 person was confused, thinking that -bt output corresponded to deliveries.
1451 (Suppressing duplicates isn't a good idea as you lose the information
1452 about possibly different redirections that led to the duplicates.)
1454 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1455 systems where poll() doesn't work, in particular OS X.
1457 PH/10 Added more information to debugging output for retry time not reached.
1459 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1460 operations in malware.c.
1462 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1465 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1466 syslog_duplication set false, log lines that would normally be written
1467 both the the main log and to the reject log were not written to syslog at
1470 PH/14 In the default configuration, change the use of "message" in ACL warn
1471 statements to "add_header".
1473 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1474 not followed by a command (e.g. "seen endif").
1476 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1477 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1480 PH/17 Added extra conditions to the default value of delay_warning_condition
1484 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1485 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1486 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1489 The Auto-Submitted: and various List- headers are standardised, whereas I
1490 don't think Precedence: ever was.
1492 PH/18 Refactored debugging code in route_finduser() to show more information,
1493 in particular, the error code if getpwnam() issues one.
1495 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1496 This is apparently needed in addition to the PH/07 change above to avoid
1497 any possible encoding problems.
1499 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1500 but not after initializing Perl.
1502 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1503 output them only if debugging. By default they are written stderr,
1504 apparently, which is not desirable.
1506 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1509 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1512 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1514 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1515 authenticated or an ident call has been made. Suppress the default
1516 values for $authenticated_id and $authenticated_sender (but permit -oMai
1517 and -oMas) when testing with -bh.
1519 PH/24 Re-jigged the order of the tests in the default configuration so that the
1520 tests for valid domains and recipients precede the DNS black list and CSA
1521 tests, on the grounds that those ones are more expensive.
1523 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1524 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1525 This bug exists in every version of Exim that I still have, right back to
1528 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1529 However, an attempt to turn on case-sensitivity in a regex key by
1530 including (?-i) didn't work because the subject string was already
1531 lowercased, and the effects were non-intuitive. It turns out that a
1532 one-line patch can be used to allow (?-i) to work as expected.
1538 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1539 other effects) broke the use of negated acl sub-conditions.
1541 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1544 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1545 "Deny" causes Exim to reject the incoming connection with a 554 error.
1546 Unfortunately, if there is a major crisis, such as a disk failure,
1547 tcp-wrappers gives "deny", whereas what one would like would be some
1548 kind of temporary error. A kludge has been added to help with this.
1549 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1550 554 error is used if errno is still zero or contains ENOENT (which occurs
1551 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1554 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1556 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1557 errors. Otherwise a message that provokes a temporary error (when other
1558 messages do not) can cause a whole host to time out.
1560 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1561 addresses were routed directly to files or pipes from a redirect router.
1562 File deliveries just didn't batch; pipe deliveries might have suffered
1565 PH/06 A failure to get a lock for a hints database would erroneously always say
1566 "Failed to get write lock", even when it was really a read lock.
1568 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1569 of 0600. This has been changed to use the value of the lockfile_mode
1570 option (which defaults to 0600).
1572 PH/08 Applied small patch from the Sieve maintainer.
1574 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1575 folder from quota calculations, a direct delivery into this folder messed
1576 up the contents of the maildirsize file. This was because the regex was
1577 used only to exclude .Trash (or whatever) when the size of the mailbox
1578 was calculated. There was no check that a delivery was happening into an
1579 excluded directory. This bug has been fixed by ignoring all quota
1580 processing for deliveries into excluded directories.
1582 PH/10 Added the maildirfolder_create_regex option to appendfile.
1588 PH/01 The code for finding all the local interface addresses on a FreeBSD
1589 system running IPv6 was broken. This may well have applied to all BSD
1590 systems, as well as to others that have similar system calls. The broken
1591 code found IPv4 interfaces correctly, but gave incorrect values for the
1592 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1593 that it would not match correctly against @[] and not recognize the IPv6
1596 PH/02 The ipliteral router was not recognizing addresses of the form user@
1597 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1599 PH/03 Added disable_ipv6.
1601 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1602 decoded form, because it is most often used to construct To: headers
1603 lines in autoreplies, and the decoded form may well be syntactically
1604 invalid. However, $reply_address has leading white space removed, and all
1605 newlines turned into spaces so that the autoreply transport does not
1608 PH/05 If group was specified without a user on a router, and no group or user
1609 was specified on a transport, the group from the router was ignored.
1611 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1612 for visible compile-time settings that can be used to change these
1613 numbers, for those that want even more. Backwards compatibility with old
1614 spool files has been maintained. However, going back to a previous Exim
1615 release will lost any variables that are in spool files.
1617 PH/07 Two small changes when running in the test harness: increase delay when
1618 passing a TCP/IP connection to a new process, in case the original
1619 process has to generate a bounce, and remove special handling of
1620 127.0.0.2 (sic), which is no longer necessary.
1622 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1623 be the same on different OS.
1625 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1628 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1629 whether --show-vars was specified or not
1631 JJ/02 exipick: Added support for new ACL variable spool format introduced
1634 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1635 syntactically invalid From: or Reply-to: line, and a filter used this to
1636 generate an autoreply, and therefore failed to obtain an address for the
1637 autoreply, Exim could try to deliver to a non-existent relative file
1638 name, causing unrelated and misleading errors. What now happens is that
1639 it logs this as a hard delivery error, but does not attempt to create a
1642 PH/11 The exinext utility has a -C option for testing purposes, but although
1643 the given file was scanned by exinext itself; it wasn't being passed on
1644 when Exim was called.
1646 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1647 an end-of-file indication when reading a command response.
1649 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1650 compiled. In many other places in Exim, IPv6 addresses are always
1651 recognized, so I have changed this. It also means that IPv4 domain
1652 literals of the form [IPV4:n.n.n.n] are now always recognized.
1654 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1655 used if the router is not running as root, for example, when verifying at
1656 ACL time, or when using -bh. The debugging output from this situation was
1657 non-existent - all you got was a failure to exec. I have made two
1660 (a) Failures to set uid/gid, the current directory, or a process leader
1661 in a subprocess such as that created by queryprogram now generate
1662 suitable debugging ouput when -d is set.
1664 (b) The queryprogram router detects when it is not running as root,
1665 outputs suitable debugging information if -d is set, and then runs
1666 the subprocess without attempting to change uid/gid.
1668 PH/15 Minor change to Makefile for building test_host (undocumented testing
1671 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1672 additional section of a DNS packet that returns MX or SRV records.
1673 Instead, it always explicitly searches for A/AAAA records. This avoids
1674 major problems that occur when a DNS server includes only records of one
1675 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1676 fixed another bug: if SRV records were looked up and the corresponding
1677 address records were *not* found in the additional section, the port
1678 values from the SRV records were lost.
1680 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1681 using the correct key (the original address) when searching the retry
1682 rules in order to find which one to use for generating the retry hint.
1684 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1685 adding the default one. Similarly, if it contains a Reply-To: header, the
1686 errors_reply_to option, if set, is not used.
1688 PH/19 When calculating a retry time, Exim used to measure the "time since
1689 failure" by looking at the "first failed" field in the retry record. Now
1690 it does not use this if it is later than than the arrival time of the
1691 message. Instead it uses the arrival time. This makes for better
1692 behaviour in cases where some deliveries succeed, thus re-setting the
1693 "first failed" field. An example is a quota failure for a huge message
1694 when small messages continue to be delivered. Without this change, the
1695 "time since failure" will always be short, possible causing more frequent
1696 delivery attempts for the huge message than are intended.
1697 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1699 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1700 $1, $2, $3) because the numerical variables can be reset during some
1701 expansion items (e.g. "match"), thereby losing the authentication data.
1703 PH/21 Make -bV show the size of off_t variables so that the test suite can
1704 decide whether to run tests for quotas > 2G.
1706 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1707 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1708 filecount value is greater than 2G or if a quota value is greater than 2G
1709 on a system where the size of off_t is not greater than 4, a panic error
1712 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1713 never match. The debug and -bh output now contains an explicit error
1714 message indicating a malformed IPv4 address or mask.
1716 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1717 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1718 PH/23 above applies.
1720 PH/25 Do not write to syslog when running in the test harness. The only
1721 occasion when this arises is a failure to open the main or panic logs
1722 (for which there is an explicit test).
1724 PH/26 Added the /no_tell option to "control=freeze".
1726 PH/27 If a host name lookup failed very early in a connection, for example, if
1727 the IP address matched host_lookup and the reverse lookup yielded a name
1728 that did not have a forward lookup, an error message of the form "no IP
1729 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1730 could be logged. Now it outputs the IP address instead of "NULL".
1732 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1733 allows the sender and the authenticated sender to be set when
1734 submitting a message from within Exim. Since child_open_exim() is
1735 documented for local_scan(), the new function should be too.
1737 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1738 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1739 results in an empty string is now treated as unset.
1741 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1743 PH/31 Added sender_verify_fail logging option.
1745 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1746 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1747 tidied the source and removed it altogether.
1749 PH/33 When a queue run was abandoned because the load average was too high, a
1750 log line was always written; now it is written only if the queue_run log
1751 selector is set. In addition, the log line for abandonment now contains
1752 information about the queue run such as the pid. This is always present
1753 in "start" and "stop" lines but was omitted from the "abandon" line.
1755 PH/34 Omit spaces between a header name and the colon in the error message that
1756 is given when verify = headers_syntax fails (if there are lots of them,
1757 the message gets confusing).
1759 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1760 names, as there are now some PTR records that contain slashes. This check
1761 is only to protect against broken name servers that fall over on strange
1762 characters, so the fact that it applies to all lookups doesn't matter.
1764 PH/36 Now that the new test suite is complete, we can remove some of the
1765 special code in Exim that was needed for the old test suite. For example,
1766 sorting DNS records because real resolvers return them in an arbitrary
1767 order. The new test suite's fake resolver always returns records in the
1770 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1771 bounces) except when queue_only is set, to avoid logging races between
1772 the different processes.
1774 PH/38 Panic-die if .include specifies a non-absolute path.
1776 PH/39 A tweak to the "H" retry rule from its user.
1778 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1779 a label. They prevented compilation on older perls.
1781 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1782 a warning to be raised on newish perls.
1784 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1785 on queue. Changes to match documented behaviour of showing count of
1786 messages matching specified criteria.
1788 PH/40 Changed the default ident timeout from 30s to 5s.
1790 PH/41 Added support for the use of login_cap features, on those BSD systems
1791 that have them, for controlling the resources used by pipe deliveries.
1793 PH/42 The content-scanning code uses fopen() to create files in which to put
1794 message data. Previously it was not paying any attention to the mode of
1795 the files. Exim runs with umask(0) because the rest of the code creates
1796 files with open(), and sets the required mode explicitly. Thus, these
1797 files were ending up world-writeable. This was not a big issue, because,
1798 being within the spool directory, they were not world-accessible. I have
1799 created a function called modefopen, which takes an additional mode
1800 argument. It sets umask(777), creates the file, chmods it to the required
1801 mode, then resets the umask. All the relevant calls to fopen() in the
1802 content scanning code have been changed to use this function.
1804 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1805 to 24 hours. This avoids potential overflow problems when processing G
1806 and H retry rules. I suspect nobody ever tinkers with this value.
1808 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1810 PH/45 When the plaintext authenticator is running as a client, the server's
1811 challenges are checked to ensure they are valid base64 strings. By
1812 default, the authentication attempt is cancelled if an invalid string is
1813 received. Setting client_ignore_invalid_base64 true ignores these errors.
1814 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1815 they are received. Thus, the responses can be made to depend on the
1816 challenges. If an invalid string is ignored, an empty string is placed in
1819 PH/46 Messages that are created by the autoreply transport now contains a
1820 References: header, in accordance with RFCs 2822 and 3834.
1822 PH/47 Added authenticated_sender_force to the smtp transport.
1824 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1826 PH/49 Installed latest patch from the Sieve maintainer.
1828 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1829 was also set, the appendfile transport was unnecessarily scanning a
1830 directory of message files (e.g. for maildir delivery) to find the count
1831 of files (along with the size), even though it did not need this
1832 information. It now does the scan only if it needs to find either the
1833 size of the count of files.
1835 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1837 PH/52 Two bugs concerned with error handling when the smtp transport is
1840 (i) Exim was not creating retry information for temporary errors given
1841 for individual recipients after the DATA command when the smtp transport
1842 was used in LMTP mode. This meant that they could be retried too
1843 frequently, and not timed out correctly.
1845 (ii) Exim was setting the flag that allows error details to be returned
1846 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1847 recipients that were returned after the DATA command.
1849 PH/53 This is related to PH/52, but is more general: for any failing address,
1850 when detailed error information was permitted to be returned to the
1851 sender, but the error was temporary, then after the final timeout, only
1852 "retry timeout exceeded" was returned. Now it returns the full error as
1853 well as "retry timeout exceeded".
1855 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1856 do this, and (what is worse) MTAs that accept it.
1858 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1859 will now be deprecated.
1861 PH/56 New os.c-cygwin from the Cygwin maintainer.
1863 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1864 formats (previously only available in exim formats via -bpr, -bpru,
1865 and -bpra. Now also available in native and exiqgrep formats)
1867 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1868 with very large, slow to parse queues
1870 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1872 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1874 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1875 responses to authentication challenges, though it was showing the
1876 challenges; (ii) I've removed the CR characters from the debug output for
1879 PH/58 Allow for the insertion of a newline as well as a space when a string
1880 is turned into more than one encoded-word during RFC 2047 encoding. The
1881 Sieve code now uses this.
1883 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1884 data_4xx, lost_connection, tls_required.
1886 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1887 message was being sent as an SMTP response.
1889 PH/61 Add -l and -k options to exicyclog.
1891 PH/62 When verifying, if an address was redirected to one new address, so that
1892 verification continued, and the new address failed or deferred after
1893 having set something in $address_data, the value of $address_data was not
1894 passed back to the ACL. This was different to the case when no
1895 redirection occurred. The value is now passed back in both cases.
1897 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1898 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1899 use login_cap.h, so on its own it isn't the distinguishing feature. The
1900 new name refers directly to the setclassresources() function.
1902 PH/65 Added configuration files for NetBSD3.
1904 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1906 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1907 is preferred over IPv4.
1909 PH/68 The bounce_return_message and bounce_return_body options were not being
1910 honoured for bounces generated during the reception of non-SMTP messages.
1911 In particular, this applied to messages rejected by the ACL. This bug has
1912 been fixed. However, if bounce_return_message is true and bounce_return_
1913 body is false, the headers that are returned for a non-SMTP message
1914 include only those that have been read before the error was detected.
1915 (In the case of an ACL rejection, they have all been read.)
1917 PH/69 The HTML version of the specification is now built in a directory called
1918 spec_html instead of spec.html, because the latter looks like a path with
1919 a MIME-type, and this confuses some software.
1921 PH/70 Catch two compiler warnings in sieve.c.
1923 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1924 function verify_get_ident() calls ip_connect() to connect a socket, but
1925 if the "connect()" function timed out, ip_connect() used to close the
1926 socket. However, verify_get_ident() also closes the socket later, and in
1927 between Exim writes to the log, which may get opened at this point. When
1928 the socket was closed in ip_connect(), the log could get the same file
1929 descriptor number as the socket. This naturally causes chaos. The fix is
1930 not to close the socket in ip_connect(); the socket should be closed by
1931 the function that creates it. There was only one place in the code where
1932 this was missing, in the iplookup router, which I don't think anybody now
1933 uses, but I've fixed it anyway.
1935 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1936 well as to direct DNS lookups. Otherwise the handling of names in host
1937 lists is inconsistent and therefore confusing.
1943 PH/01 Two changes to the default runtime configuration:
1945 (1) Move the checks for relay_from_hosts and authenticated clients from
1946 after to before the (commented out) DNS black list checks.
1948 (2) Add control=submission to the relay_from_hosts and authenticated
1949 clients checks, on the grounds that messages accepted by these
1950 statements are most likely to be submissions.
1952 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1954 (1) Generate an error if the third argument for the ${prvs expansion is
1957 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1960 (3) Reset the variables that are obtained from the first argument of
1961 ${prvscheck and used in the second argument before leaving the code,
1962 because their memory is reclaimed, so using them afterwards may do
1965 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1966 one (it's much easier than Tom thought :-).
1968 (5) Because of (4), we can now allow for the use of $prvscheck_result
1969 inside the third argument.
1971 PH/03 For some reason, the default setting of PATH when running a command from
1972 a pipe transport was just "/usr/bin". I have changed it to
1975 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1976 anything to be listed in the output from -bV.
1978 PH/05 When a filter generated an autoreply, the entire To: header line was
1979 quoted in the delivery log line, like this:
1981 => >A.N.Other <ano@some.domain> <original@ddress> ...
1983 This has been changed so that it extracts the operative address. There
1984 may be more than one such address. If so, they are comma-separated, like
1987 => >ano@some.domain,ona@other.domain <original@ddress> ...
1989 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1990 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1991 not being looked up in the rDNS to get a host name, Exim was showing the
1992 IP address twice in Received: lines, even though the IP addresses were
1993 identical. For example:
1995 Received: from [1.2.3.4] (helo=[1.2.3.4])
1997 However, if the real host name was known, it was omitting the HELO data
1998 if it matched the actual IP address. This has been tidied up so that it
1999 doesn't show the same IP address twice.
2001 PH/07 When both +timestamp and +memory debugging was on, the value given by
2002 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2003 called by the debug printing, thereby overwriting the timestamp buffer.
2004 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2006 PH/08 When the original message was included in an autoreply transport, it
2007 always said "this is a copy of the message, including all the headers",
2008 even if body_only or headers_only was set. It now gives an appropriate
2011 PH/09 Applied a patch from the Sieve maintainer which:
2013 o fixes some comments
2014 o adds the (disabled) notify extension core
2015 o adds some debug output for the result of if/elsif tests
2016 o points to the current vacation draft in the documentation
2017 and documents the missing references header update
2021 o fixes a bug in processing the envelope test (when testing
2022 multiple envelope elements, the last element determinted the
2025 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2026 Electronic Mail") by including:
2028 Auto-submitted: auto-generated
2030 in the messages that it generates (bounce messages and others, such as
2031 warnings). In the case of bounce messages for non-SMTP mesages, there was
2032 also a typo: it was using "Auto_submitted" (underscore instead of
2033 hyphen). Since every message generated by Exim is necessarily in response
2034 to another message, thes have all been changed to:
2036 Auto-Submitted: auto-replied
2038 in accordance with these statements in the RFC:
2040 The auto-replied keyword:
2042 - SHOULD be used on messages sent in direct response to another
2043 message by an automatic process,
2045 - MUST NOT be used on manually-generated messages,
2047 - MAY be used on Delivery Status Notifications (DSNs) and Message
2048 Disposition Notifications (MDNs),
2050 - MUST NOT be used on messages generated by automatic or periodic
2051 processes, except for messages which are automatic responses to
2054 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2055 to the default Received: header definition.
2057 PH/12 Added log selector acl_warn_skipped (default on).
2059 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2060 variables because (a) they are in the wrong storage pool and (b) even if
2061 they were copied, it wouldn't work properly because of the caching.
2063 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2064 checking when decoding. Apparently there are clients that generate
2065 overlong encoded strings. Why am I not surprised?
2067 PH/15 If the first argument of "${if match_address" was not empty, but did not
2068 contain an "@" character, Exim crashed. Now it writes a panic log message
2069 and treats the condition as false.
2071 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2073 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2074 extension "envelope-auth". The code is finished and in agreement with
2075 other implementations, but there is no documentation so far and in fact,
2076 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2077 not changing the active code.
2079 Print executed "if" and "elsif" statements when debugging is used. This
2080 helps a great deal to understand what a filter does.
2082 Document more things not specified clearly in RFC3028. I had all this
2083 sorted out, when out of a sudden new issues came to my mind. Oops."
2085 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2088 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2089 canonical form (as documented). However, after a host name lookup from
2090 the IP address, check_host() was doing a simple string comparison with
2091 addresses acquired from the DNS when checking that the found name did
2092 have the original IP as one of its addresses. Since any found IPv6
2093 addresses are likely to be in abbreviated form, the comparison could
2094 fail. Luckily, there already exists a function for doing the comparison
2095 by converting both addresses to binary, so now that is used instead of
2096 the text comparison.
2098 PH/20 There was another similar case to PH/19, when a complete host name was
2099 given in a host list; looking up its IP address could give an abbreviated
2100 form, whereas the current host's name might or might not be abbreviated.
2101 The same fix has been applied.
2107 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2108 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2111 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2112 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2114 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2116 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2117 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2118 Also, for "Auto-Submitted", treat anything other than "no" as
2119 non-personal, in accordance with RFC 3834. (Previously it treated
2120 anything starting "auto-" as non-personal.)
2122 TF/01 The control=submission/name=... option had a problem with syntax
2123 errors if the name included a slash character. The /name= option
2124 now slurps the rest of the string, so it can include any characters
2125 but it must come last in the list of options (after /sender_retain
2128 PH/05 Some modifications to the interface to the fake nameserver for the new
2136 TK/01 Added the "success_on_redirect" address verification option. See
2137 NewStuff for rationale and an example.
2139 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2141 PH/02 Patch to exigrep to allow it to work on syslog lines.
2143 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2144 fread() to skip over the body file's header line, because in Cygwin the
2145 header line is locked and is inaccessible.
2147 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2148 co-exist for some time) to make it clear that it is the Exim ID that is
2149 referenced, not the Message-ID: header line.
2151 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2152 string_format() function, because snprintf() does not exist on all
2155 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2156 ${stat: expansion item.
2158 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2159 protocol synchronization error", to keep the pedants happy.
2161 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2162 well as for IRIX systems, when gcc is being used. See the host.c source
2165 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2167 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2170 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2171 transport and to the smtp transport in LMTP mode.
2173 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2175 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2176 run a filter in a subprocess. This could lead to confusion in subsequent
2177 lookups in the parent process. There should also be a search_tidyup() at
2178 the end of the subprocess.
2180 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2181 only if the host matched helo_try_verify_hosts, which caused the
2182 verification to occur when the EHLO/HELO command was issued. The ACL just
2183 tested the remembered result. Now, if a previous verification attempt has
2184 not happened, "verify = helo" does it there and then.
2186 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2188 TK/03 Fix log output including CR from clamd.
2190 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2191 exist provoked a memory error which could cause a segfault.
2193 PH/15 Installed PCRE 6.2
2195 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2197 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2198 of the problem. Specifically, suggested +O2 rather than +O1 for the
2201 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2203 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2204 record, the port was not correctly logged when the outgoing_port log
2205 selector was set (it logged the transort's default port).
2207 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2208 fallback_hosts, and "hosts" in the smtp transport.
2210 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2211 host errors such as "Connection refused".
2213 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2214 authentication with radiusclient 0.4.9:
2216 - Error returned from rc_read_config was caught wrongly
2217 - Username/password not passed on to radius server due to wrong length.
2219 The presumption is that some radiusclient API changes for 4.51/PH/17
2220 were not taken care of correctly. The code is still untested by me (my
2221 Linux distribution still has 0.3.2 of radiusclient), but it was
2222 contributed by a Radius user.
2224 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2225 expanding the "port" option of the smtp transport.
2227 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2228 while reading a MIME header. Thanks to Tom Hughes for a patch.
2230 PH/24 Include config.h inside local_scan.h so that configuration settings are
2233 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2234 that in an ACL for RCPT (for example), you can examine exactly what was
2237 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2238 commands, but it was not correctly comparing the address with the actual
2239 client host address. Thus, it would show the EHLO address in Received:
2240 header lines when this was not necessary.
2242 PH/27 Added the % operator to ${eval:}.
2244 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2245 it should be ignoring failures (because with -C, for example, it has lost
2246 privilege). It wasn't ignoring creation failures other than "already
2249 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2252 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2253 because an input line was too long, either on its own, or by virtue of
2254 too many continuations, the temporary file was not being removed, and the
2255 return code was incorrect.
2257 PH/31 Missing "BOOL" in function definition in filtertest.c.
2259 PH/32 Applied Sieve patches from the maintainer.
2261 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2263 PH/33 Added "verify = not_blind".
2265 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2266 Local/Makefile (with some defaults set). These are used in built scripts
2267 such as exicyclog, but they have never been used in the exim_install
2268 script (though there are many overriding facilities there). I have
2269 arranged that the exim_install script now takes note of these two
2272 PH/35 Installed configuration files for Dragonfly.
2274 PH/36 When a locally submitted message by a trusted user did not contain a
2275 From: header, and the sender address was obtained from -f or from an SMTP
2276 MAIL command, and the trusted user did not use -F to supply a sender
2277 name, $originator_name was incorrectly used when constructing a From:
2278 header. Furthermore, $originator_name was used for submission mode
2279 messages from external hosts without From: headers in a similar way,
2280 which is clearly wrong.
2282 PH/37 Added control=suppress_local_fixups.
2284 PH/38 When log_selector = +received_sender was set, and the addition of the
2285 sender made the log line's construction buffer exactly full, or one byte
2286 less than full, an overflow happened when the terminating "\n" was
2289 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2290 when the result of a list match is failure because a DNS lookup failed.
2292 PH/40 RM_COMMAND is now used in the building process.
2294 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2295 the "build-* directories that it finds.
2297 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2298 address in a domain literal was a prefix of an interface address.
2300 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2301 when verifying a sender address, unless rewrite_headers is false.
2303 PH/44 Wrote a long comment about why errors_to addresses are verified as
2304 recipients, not senders.
2306 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2307 the ratelimit ACL was added.
2309 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2311 PH/46 Added extra information about PostgreSQL errors to the error string.
2313 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2314 avoiding the need to install special zones in a real server. This is
2315 backwards compatible; if it can't find the fake resolver, it drops back.
2316 Thus, both old and new test suites can be run.
2318 TF/02 Added util/ratelimit.pl
2320 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2321 clock is set back in time.
2323 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2324 Candler <B.Candler@pobox.com>.
2326 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2327 OK for addresses that are the result of redirections.
2329 PH/48 A number of further additions for the benefit of the new test suite,
2330 including a fake gethostbyname() that interfaces to the fake DNS resolver
2333 TF/06 The fix for widen_domains has also been applied to qualify_single and
2334 search_parents which are the other dnslookup options that can cause
2337 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2340 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2342 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2343 service" if it encountered an MX record with an empty target hostname.
2344 The message is now "an MX or SRV record indicated no SMTP service".
2346 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2347 if the DNS of the sending site is misconfigured. This is quite a
2348 common situation. This change restores the behaviour of treating a
2349 helo verification defer as a failure.
2351 PH/51 If self=fail was set on a router, the bounce message did not include the
2352 actual error message.
2358 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2360 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2361 command itself ended in error, the underlying message about the transport
2362 filter timeout was being overwritten with the pipe command error. Now the
2363 underlying error message should be appended to the second error message.
2365 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2367 PH/02 Reduce the amount of output that "make" produces by default. Full output
2368 can still be requested.
2370 PH/03 The warning log line about a condition test deferring for a "warn" verb
2371 was being output only once per connection, rather than after each
2372 occurrence (because it was using the same function as for successful
2373 "warn" verbs). This seems wrong, so I have changed it.
2375 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2376 it should not have, which might have caused a crash in the right
2377 circumstances, but probably never did.
2379 PH/04 Installed a modified version of Tony Finch's patch to make submission
2380 mode fix the return path as well as the Sender: header line, and to
2381 add a /name= option so that you can make the user's friendly name appear
2384 TF/03 Added the control = fakedefer ACL modifier.
2386 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2387 Mark Lowes for thorough testing.
2389 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2391 TK/03 Merged latest SRS patch from Miles Wilton.
2393 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2394 with the definition in sysexits.h (which is #included earlier).
2395 Fortunately, Exim does not actually use EX_OK. The code used to try to
2396 preserve the sysexits.h value, by assumimg that macro definitions were
2397 scanned for macro replacements. I have been disabused of this notion,
2398 so now the code just undefines EX_OK before #including unistd.h.
2400 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2401 in the smtp transport. When a block could not be written in a single
2402 write() function, the timeout was being re-applied to each part-write.
2403 This seems wrong - if the receiver was accepting one byte at a time it
2404 would take for ever. The timeout is now adjusted when this happens. It
2405 doesn't have to be particularly precise.
2407 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2408 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2410 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2411 without a domain if the check to <postmaster@domain> fails.
2413 SC/01 Eximstats: added -xls and the ability to specify output files
2414 (patch written by Frank Heydlauf).
2416 SC/02 Eximstats: use FileHandles for outputing results.
2418 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2420 SC/04 Eximstats: fixed display of large numbers with -nvr option
2422 SC/05 Eximstats: fixed merging of reports with empty tables.
2424 SC/06 Eximstats: added the -include_original_destination flag
2426 SC/07 Eximstats: removed tabs and trailing whitespace.
2428 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2430 TK/06 MBOX spool code: Add real "From " MBOX separator line
2431 so the .eml file is really in mbox format (even though
2432 most programs do not really care). Patch from Alex Miller.
2434 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2435 The latter is generated from $received_to and is only set if the
2436 message has one envelope recipient. SA can use these headers,
2437 obviously out-of-the-box. Patch from Alex Miller.
2439 PH/08 The ${def test on a variable was returning false if the variable's
2440 value was "0", contrary to what the specification has always said!
2441 The result should be true unless the variable is empty.
2443 PH/09 The syntax error of a character other than { following "${if
2444 def:variable_name" (after optional whitespace) was not being diagnosed.
2445 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2446 accidental colon was present, for example, could give incorrect results.
2448 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2449 result is used (not including appendfile, where other changes are about
2452 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2453 This involved changing a lot of size variables from int to off_t. It
2454 should work with maildirs and everything.
2456 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2457 spamd dying while we are connected to it.
2459 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2462 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2463 function for BDB 4.3.
2465 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2467 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2468 the sources for PCRE have been re-arranged and more files are now
2471 PH/15 The code I had for printing potentially long long variables in PH/11
2472 above was not the best (it lost precision). The length of off_t variables
2473 is now inspected at build time, and an appropriate printing format (%ld
2474 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2475 to be "long long int" or "long int". This is needed for the internal
2476 formatting function string_vformat().
2478 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2479 the configuration file to be ":syslog", then the script "guesses" where
2480 the logs files are, rather than using the compiled in default. In our
2481 case the guess is not the same as the compiled default, so the script
2482 suddenly stopped working when I started to use syslog. The patch checks
2483 to see if log_file_path is "". If so, it attempts to read it from exim
2484 with no configuration file to get the compiled in version, before it
2485 falls back to the previous guessing code."
2487 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2488 implementing BATV in an Exim configuration. See NewStuff for the gory
2491 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2492 Makefile that are specific to HP-UX.
2494 PH/18 If the "use_postmaster" option was set for a recipient callout together
2495 with the "random" option, the postmaster address was used as the MAIL
2496 FROM address for the random test, but not for the subsequent recipient
2497 test. It is now used for both.
2499 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2500 patch removes a few documentation additions to RFC 3028, because the
2501 latest draft now contains them. It adds the new en;ascii-case comparator
2502 and a new error check for 8bit text in MIME parts. Comparator and
2503 require names are now matched exactly. I enabled the subaddress
2504 extension, but it is not well tested yet (read: it works for me)."
2506 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2507 rework some of the code of TK/09 above to avoid the hardwired use of
2508 "%lld" and "long long". Replaced the call to snprintf() with a call to
2511 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2512 records point to non-existent hosts", "retry timeout exceeded", and
2513 "retry time not reached for any host after a long failure period".
2515 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2516 experimental DomainKeys support:
2518 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2519 (2) On an error such as an illegally used "control", the wrong name for
2520 the control was given.
2522 These problems did NOT occur unless DomainKeys support was compiled.
2524 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2526 PH/24 Added ${if match_ip condition.
2528 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2529 databases so that it will be absolutely obvious if a crash occurs in the
2530 DB library. This is a regular occurrence (often caused by mis-matched
2533 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2534 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2535 picked up on a user's system that detects such things. There doesn't seem
2536 to be a gcc warning option for this - only an attribute that has to be
2537 put on the function's prototype. It seems that in Fedora Core 4 they have
2538 set this on a number of new functions. No doubt there will be more in due
2541 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2542 specify a transport. However, if an address that was verified by such a
2543 router was the subject of a callout, Exim crashed because it tried to
2544 read the rcpt_include_affixes from the non-existent transport. Now it
2545 just assumes that the setting of that option is false. This bug was
2546 introduced by 4.51/PH/31.
2548 PH/28 Changed -d+all to exclude +memory, because that information is very
2549 rarely of interest, but it makes the output a lot bigger. People tend to
2550 do -d+all out of habit.
2552 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2553 code in os-type was giving problems when libc.so lives in lib64, like on
2556 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2557 aren't the modern standard, and it seems that some systems' include files
2558 don't always have them. Exim was already checking for some of the newer
2559 ones like T_AAAA, and defining it itself. I've added checks for all the
2560 record types that Exim uses.
2562 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2563 not automatically generating a new one, as it is supposed to. This
2564 prevented TLS from working. If the file did exist, but contained invalid
2565 data, a new version was generated, as expected. It was only the case of a
2566 non-existent file that was broken.
2568 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2569 with a change in libdomainkeys > 0.64.
2571 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2572 from DNS. If the selector record carries the flag, it now has
2573 precedence over the domain-wide flag.
2575 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2577 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2578 the use of a port name that isn't defined in /etc/services) occurred, the
2579 message was deferred as in a normal delivery, and thus remained on the
2580 spool, instead of being failed because of the mua_wrapper setting. This
2581 is now fixed, and I tidied up some of the mua_wrapper messages at the
2584 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2585 the messages in a hash of arrays rather than using individual hashes.
2586 This is a bit cleaner and results in dramatic memory savings, albeit
2587 at a slight CPU cost.
2589 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2590 as requested by Marc Sherman.
2592 SC/10 Eximstats: added histograms for user specified patterns as requested
2595 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2597 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2598 fopen() in the content-scanning modules that did not already have it.
2604 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2605 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2607 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2609 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2611 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2612 to test Sieve filters that use "vacation".
2614 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2615 that changes the way the GnuTLS parameters are stored in the cache file.
2616 The new format can be generated externally. For backward compatibility,
2617 if the data in the cache doesn't make sense, Exim assumes it has read an
2618 old-format file, and it generates new data and writes a new file. This
2619 means that you can't go back to an older release without removing the
2622 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2623 work if there are any delivery delays because "one_time" forces the
2624 parent to be marked "delivered", so its unseen clone is never tried
2625 again. For this reason, Exim now forbids the simultaneous setting of
2628 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2629 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2630 entry if you want to know the details. The fix, however, neglected to
2631 consider the case when local delivery batching is involved. The test for
2632 "previously delivered" was not happening when checking to see if an
2633 address could be batched with a previous (undelivered) one; under
2634 certain circumstances this could lead to multiple deliveries to the same
2637 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2638 in its include files, and this causes problems building Exim.
2640 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2641 header_syntax) but Exim was just ignoring anything given after a slash.
2642 In particular, this caused confusion with an attempt to use "verify =
2643 reverse_host_lookup/defer_ok". An error is now given when options are
2644 supplied for verify items that do not have them. (Maybe reverse_host_
2645 lookup should have a defer_ok option, but that's a different point.)
2647 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2648 defined by RFC 821) to 2048, because there were problems with some AUTH
2649 commands, and RFC 1869 says the size should be increased for extended
2650 SMTP commands that take arguments.
2652 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2655 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2656 "unknown" error; now it says that the functionality isn't in the binary.
2658 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2659 an address' error message when a string expansion fails (syntax or
2660 whatever). Otherwise the password may appear in the log. Following change
2661 PH/42 below, there is no longer a chance of it appearing in a bounce
2664 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2666 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2667 of its addresses was ever tried. (Bugzilla bug #2.)
2669 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2670 the result incorrectly in the debug output. (It correctly added a newline
2671 to what was transported.)
2673 TF/01 Added $received_time.
2675 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2676 commented out examples of how to interface to a virus scanner and to
2677 SpamAssassin. Also added commented examples of av_scanner and
2678 spamd_address settings.
2680 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2681 and controls are allowed in which ACLs. There were a couple of minor
2682 errors. Some of the entries in the conditions table (which is a table of
2683 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2684 negation of where the condition IS allowed.
2686 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2688 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2689 header file does not have a version number, so I've had to invent a new
2690 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2691 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2692 radiusclient), but it was contributed by a Radius user.
2694 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2695 files or directories, for OpenSSL.
2697 PH/19 When an Exim process that is running as root has to create an Exim log
2698 file, it does so in a subprocess that runs as exim:exim so as to get the
2699 ownership right at creation (otherwise, other Exim processes might see
2700 the file with the wrong ownership). There was no test for failure of this
2701 fork() call, which would lead to the process getting stuck as it waited
2702 for a non-existent subprocess. Forks do occasionally fail when resources
2703 run out. I reviewed all the other calls to fork(); they all seem to check
2706 PH/20 When checking for unexpected SMTP input at connect time (before writing
2707 the banner), Exim was not dealing correctly with a non-positive return
2708 from the read() function. If the client had disconnected by this time,
2709 the result was a log entry for a synchronization error with an empty
2710 string after "input=" when read() returned zero. If read() returned -1
2711 (an event I could not check), uninitialized data bytes were printed.
2712 There were reports of junk text (parts of files, etc) appearing after
2715 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2717 PH/22 Added support for macro redefinition, and (re)definition in between
2718 driver and ACL definitions.
2720 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2721 forgetting to use the resulting value; it was using the unexpanded value.
2723 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2724 hadn't been configured. The fix is from Juergen Kreileder, who
2725 understands it better than I do:
2727 "Here's what I see happening with three configured cyrus_sasl
2728 authenticators configured (plain, login, cram-md5):
2730 On startup auth_cyrus_sasl_init() gets called for each of these.
2731 This means three calls to sasl_listmech() without a specified mech_list.
2732 => SASL tests which mechs of all available mechs actually work
2733 => three warnings about OTP not working
2734 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2736 With the patch, sasl_listmech() also gets called three times. But now
2737 SASL's mech_list option is set to the server_mech specified in the the
2738 authenticator. Or in other words, the answer from sasl_listmech()
2739 gets limited to just the mech you're testing for (which is different
2741 => the return list contains just 'plain' or 'login', 'cram-md5' or
2742 nothing depending on the value of ob->server_mech.
2744 I've just tested the patch: Authentication still works fine,
2745 unavailable mechs specified in the exim configuration are still
2746 caught, and the auth.log warnings about OTP are gone."
2748 PH/25 When debugging is enabled, the contents of the command line are added
2749 to the debugging output, even when log_selector=+arguments is not
2752 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2753 answer is "GNU", and only if the return is "GNU/something" is the answer
2756 PH/27 $acl_verify_message is now set immediately after the failure of a
2757 verification in an ACL, and so is available in subsequent modifiers. In
2758 particular, the message can be preserved by coding like this:
2760 warn !verify = sender
2761 set acl_m0 = $acl_verify_message
2763 Previously, $acl_verify_message was set only while expanding "message"
2764 and "log_message" when a very denied access.
2766 PH/28 Modified OS/os.c-Linux with
2768 -#ifndef OS_LOAD_AVERAGE
2769 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2771 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2772 nomenclature these days.)
2774 PH/29 Installed patch from the Sieve maintainer that adds the options
2775 sieve_useraddress and sieve_subaddress to the redirect router.
2777 PH/30 In these circumstances:
2778 . Two addresses routed to the same list of hosts;
2779 . First host does not offer TLS;
2780 . First host accepts first address;
2781 . First host gives temporary error to second address;
2782 . Second host offers TLS and a TLS session is established;
2783 . Second host accepts second address.
2784 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2785 and peerdn, if requested) that were in fact used only for the second
2788 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2789 attention to any local part prefix or suffix that was matched by the
2790 router that accepted the address. It now behaves in the same way as it
2791 does for delivery: the affixes are removed from the local part unless
2792 rcpt_include_affixes is set on the transport.
2794 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2795 timeout during the DATA phase of an incoming message.
2797 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2798 applied a patch sanctioned by the Sieve maintainer.
2800 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2801 the uid or gid is negative. A case of a negative gid caused this to be
2802 noticed. The fix allows for either to be negative.
2804 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2805 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2808 PH/36 The change PH/12 above was broken. Fixed it.
2810 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2811 the grounds that routing the same address twice would always produce the
2812 same answer. This might have been true once, but it is certainly no
2813 longer true now. Routing a child address may depend on the previous
2814 routing that produced that child. Some complicated redirection strategies
2815 went wrong when messages had multiple recipients, and made Exim's
2816 behaviour dependent on the order in which the addresses were given.
2818 I have moved the duplicate checking until after the routing is complete.
2819 Exim scans the addresses that are assigned to local and remote
2820 transports, and removes any duplicates. This means that more work will be
2821 done, as duplicates will always all be routed, but duplicates are
2822 presumably rare, so I don't expect this is of any significance.
2824 For deliveries to pipes, files, and autoreplies, the duplicate checking
2825 still happens during the routing process, since they are not going to be
2828 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2829 It corrects a timeout issue with spamd. This is Ian's comment: "The
2830 background is that sometimes spamd either never reads data from a
2831 connection it has accepted, or it never writes response data. The exiscan
2832 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2833 blindly assumes that writes won't block so it may never time out."
2835 PH/39 Allow G after quota size as well as K and M.
2837 PH/40 The value set for $authenticated_id in an authenticator may not contain
2838 binary zeroes or newlines because the value is written to log lines and
2839 to spool files. There was no check on this. Now the value is run through
2840 the string_printing() function so that such characters are converted to
2841 printable escape sequences.
2843 PH/41 $message_linecount is a new variable that contains the total number of
2844 lines in the message. Compare $body_linecount, which is the count for the
2847 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2848 bounce and delay warning messages, except in certain special cases, which
2851 (a) An SMTP error message from a remote host;
2852 (b) A message specified in a :fail: redirection;
2853 (c) A message specified in a "fail" command in a system filter;
2854 (d) A message specified in a FAIL return from the queryprogram router;
2855 (e) A message specified by the cannot_route_message router option.
2857 In these cases only, Exim does include the error details in bounce and
2858 warning messages. There are also a few cases where bland messages such
2859 as "unrouteable address" or "local delivery error" are given.
2861 PH/43 $value is now also set for the "else" part of a ${run expansion.
2863 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2864 being worked on, but at least Exim now implements the latest version to
2867 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2868 process to complete was treated as a delivery failure, a timeout while
2869 writing the message to the pipe was logged, but erroneously treated as a
2870 successful delivery. Such timeouts include transport filter timeouts. For
2871 consistency with the overall process timeout, these timeouts are now
2872 treated as errors, giving rise to delivery failures by default. However,
2873 there is now a new Boolean option for the pipe transport called
2874 timeout_defer, which, if set TRUE, converts the failures into defers for
2875 both kinds of timeout. A transport filter timeout is now identified in
2878 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2879 systems where "make" and "gmake" are different, calling "gmake" at top
2880 level broke things. I've arranged for the value of $(MAKE) to be passed
2881 from the Makefile to this script so that it can call the same version of
2885 A note about Exim versions 4.44 and 4.50
2886 ----------------------------------------
2888 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2889 changes of various kinds. As a consequence, a big documentation update was
2890 needed. This delayed the release for rather longer than seemed good, especially
2891 in the light of a couple of (minor) security issues. Therefore, the changes
2892 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2893 release. So 4.44 and 4.50 are in effect two different branches that both start
2896 I have left the 4.50 change log unchanged; it contains all the changes since
2897 4.43. The change log for 4.44 is below; many of its items are identical to
2898 those for 4.50. This seems to be the most sensible way to preserve the
2899 historical information.
2905 1. Minor wording change to the doc/README.SIEVE file.
2907 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2908 computation of the current number of files was incorrect.
2910 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2911 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2914 4. Give more explanation in the error message when the command for a transport
2915 filter fails to execute.
2917 5. There are several places where Exim runs a non-Exim command in a
2918 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2919 was being done only for the command run by the queryprogram router. It is
2920 now done for all such subprocesses. The other cases are: ${run, transport
2921 filters, and the commands run by the lmtp and pipe transports.
2923 6. Added CONFIGURE_GROUP build-time option.
2925 7. Some older OS have a limit of 256 on the maximum number of file
2926 descriptors. Exim was using setrlimit() to set 1000 as a large value
2927 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2928 systems. I've change it so that if it can't get 1000, it tries for 256.
2930 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2931 was an oversight, and furthermore, ever since the addition of extra
2932 controls (e.g. 4.43/32), the checks on when to allow different forms of
2933 "control" were broken. There should now be diagnostics for all cases when a
2934 control that does not make sense is encountered.
2936 9. Added the /retain_sender option to "control=submission".
2938 10. $recipients is now available in the predata ACL (oversight).
2940 11. Tidy the search cache before the fork to do a delivery from a message
2941 received from the command line. Otherwise the child will trigger a lookup
2942 failure and thereby defer the delivery if it tries to use (for example) a
2943 cached ldap connection that the parent has called unbind on.
2945 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2946 of $address_data from the recipient verification was clobbered by the
2947 sender verification.
2949 13. The value of address_data from a sender verification is now available in
2950 $sender_address_data in subsequent conditions in the ACL statement.
2952 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2954 15. Added a new option "connect=<time>" to callout options, to set a different
2957 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2958 was its contents. (It was OK if the option was not defined at all.)
2960 17. A "Completed" log line is now written for messages that are removed from
2961 the spool by the -Mrm option.
2963 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2964 information about exactly what failed.
2966 19. Added -dd to debug only the daemon process.
2968 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2969 handles timeouts, both on the server side and network timeouts. Renamed the
2970 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2972 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2973 It is now set to "smtps".
2975 22. $host_address is now set to the target address during the checking of
2976 ignore_target_hosts.
2978 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2979 being passed; this would have caused $sender_host_name to have been used if
2980 matching the list had actually called for a host name (not very likely,
2981 since this list is usually IP addresses). A host name is now passed as
2984 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2985 code that specifies a non-restarting handler (typically sigaction() in
2986 modern systems) in an attempt to fix a rare and obscure crash bug.
2988 25. Narrowed the window for a race in the daemon that could cause it to ignore
2989 SIGCHLD signals. This is not a major problem, because they are used only to
2990 wake it up if nothing else does.
2992 26. A malformed maildirsize file could cause Exim to calculate negative values
2993 for the mailbox size or file count. Odd effects could occur as a result.
2994 The maildirsize information is now recalculated if the size or filecount
2997 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2998 support for a long time. Removed HAVE_SYS_VFS_H.
3000 28. Installed the latest version of exipick from John Jetmore.
3002 29. In an address list, if the pattern was not a regular expression, an empty
3003 subject address (from a bounce message) matched only if the pattern was an
3004 empty string. Non-empty patterns were not even tested. This was the wrong
3005 because it is perfectly reasonable to use an empty address as part of a
3006 database query. An empty address is now tested by patterns that are
3007 lookups. However, all the other forms of pattern expect the subject to
3008 contain a local part and a domain, and therefore, for them, an empty
3009 address still always fails if the pattern is not itself empty.
3011 30. Exim went into a mad DNS loop when attempting to do a callout where the
3012 host was specified on an smtp transport, and looking it up yielded more
3013 than one IP address.
3015 31. Re-factored the code for checking spool and log partition space into a
3016 function that finds that data and another that does the check. The former
3017 is then used to implement four new variables: $spool_space, $log_space,
3018 $spool_inodes, and $log_inodes.
3020 32. The RFC2047 encoding function was originally intended for short strings
3021 such as real names; it was not keeping to the 75-character limit for
3022 encoded words that the RFC imposes. It now respects the limit, and
3023 generates multiple encoded words if necessary. To be on the safe side, I
3024 have increased the buffer size for the ${rfc2047: expansion operator from
3027 33. It is now permitted to omit both strings after an "if" condition; if the
3028 condition is true, the result is "true". As before, when the second string
3029 is omitted, a false condition yields an empty string. This makes it less
3030 cumbersome to write custom ACL and router conditions.
3032 34. Failure to deliver a bounce message always caused it to be frozen, even if
3033 there was an errors_to setting on the router. The errors_to setting is now
3036 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3037 canonical form (fully expanded) before being placed in
3038 $sender_host_address.
3040 36. The table in the code that translates DNS record types into text (T_A to
3041 "A" for instance) was missing entries for NS and CNAME. It is just possible
3042 that this could have caused confusion if both these types were looked up
3043 for the same domain, because the text type is used as part of Exim's
3044 per-process caching. But the chance of anyone hitting this buglet seems
3047 37. The dnsdb lookup has been extended in a number of ways.
3049 (1) There is a new type, "zns", which walks up the domain tree until it
3050 finds some nameserver records. It should be used with care.
3052 (2) There is a new type, "mxh", which is like "mx" except that it returns
3053 just the host names, not the priorities.
3055 (3) It is now possible to give a list of domains (or IP addresses) to be
3056 looked up. The behaviour when one of the lookups defers can be
3057 controlled by a keyword.
3059 (4) It is now possible to specify the separator character for use when
3060 multiple records are returned.
3062 38. The dnslists ACL condition has been extended: it is now possible to supply
3063 a list of IP addresses and/or domains to be looked up in a particular DNS
3066 39. Added log_selector=+queue_time_overall.
3068 40. When running the queue in the test harness, wait just a tad after forking a
3069 delivery process, to get repeatability of debugging output.
3071 41. Include certificate and key file names in error message when GnuTLS fails
3072 to set them up, because the GnuTLS error message doesn't include the name
3073 of the failing file when there is a problem reading it.
3075 42. Allow both -bf and -bF in the same test run.
3077 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3079 44. The "Exiscan patch" is now merged into the mainline Exim source.
3081 45. Sometimes the final signoff response after QUIT could fail to get
3082 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3083 < 0 before doing a fflush(). This bug looks as though it goes back to the
3084 introduction of TLS in release 3.20, but "sometimes" must have been rare
3085 because the tests only now provoked it.
3087 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3088 (this can affect the format of dates).
3090 47. exim_tidydb, when checking for the continued existence of a message for
3091 which it has found a message-specific retry record, was not finding
3092 messages that were in split spool directories. Consequently, it was
3093 deleting retry records that should have stayed in existence.
3095 48. Steve fixed some bugs in eximstats.
3097 49. The SPA authentication driver was not abandoning authentication and moving
3098 on to the next authenticator when an expansion was forced to fail,
3099 contradicting the general specification for all authenticators. Instead it
3100 was generating a temporary error. It now behaves as specified.
3102 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3103 (the order specifies the preference for clients). The order is now AES256,
3104 AES128, 3DES, ARCFOUR128.
3106 51. Small patch to Sieve code - explicitly set From: when generating an
3109 52. Exim crashed if a remote delivery caused a very long error message to be
3110 recorded - for instance if somebody sent an entire SpamAssassin report back
3111 as a large number of 550 error lines. This bug was coincidentally fixed by
3112 increasing the size of one of Exim's internal buffers (big_buffer) that
3113 happened as part of the Exiscan merge. However, to be on the safe side, I
3114 have made the code more robust (and fixed the comments that describe what
3117 53. Now that there can be additional text after "Completed" in log lines (if
3118 the queue_time_overall log selector is set), a one-byte patch to exigrep
3119 was needed to allow it to recognize "Completed" as not the last thing in
3122 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3123 patch that reportedly fixes this has been added. I am not expert enough to
3124 create a test for it. This is what the patch creator wrote:
3126 "I found a little strange behaviour of ldap code when working with
3127 Windows 2003 AD Domain, where users was placed in more than one
3128 Organization Units. When I tried to give exim partial DN, the exit code
3129 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3130 But simultaneously result of request was absolutely normal ldap result,
3131 so I produce this patch..."
3133 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3134 REFERENCE, so I have modified the code to exclude the patch when that macro
3137 55. Some experimental protocols are using DNS PTR records for new purposes. The
3138 keys for these records are domain names, not reversed IP addresses. The
3139 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3140 leaves it alone. Component reversal etc. now happens only for IP addresses.
3143 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3145 57. Double the size of the debug message buffer (to 2048) so that more of very
3146 long debug lines gets shown.
3148 58. The exicyclog utility now does better if the number of log files to keep
3149 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3151 59. Two changes related to the smtp_active_hostname option:
3153 (1) $smtp_active_hostname is now available as a variable.
3154 (2) The default for smtp_banner uses $smtp_active_hostname instead
3155 of $primary_hostname.
3157 60. The host_aton() function is supposed to be passed a string that is known
3158 to be a valid IP address. However, in the case of IPv6 addresses, it was
3159 not checking this. This is a hostage to fortune. Exim now panics and dies
3160 if the condition is not met. A case was found where this could be provoked
3161 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3162 components; fortuitously, this particular loophole had already been fixed
3163 by change 4.50/55 above.
3165 If there are any other similar loopholes, the new check in host_aton()
3166 itself should stop them being exploited. The report I received stated that
3167 data on the command line could provoke the exploit when Exim was running as
3168 exim, but did not say which command line option was involved. All I could
3169 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3170 running as the user.
3173 61. There was a buffer overflow vulnerability in the SPA authentication code
3174 (which came originally from the Samba project). I have added a test to the
3175 spa_base64_to_bits() function which I hope fixes it.
3178 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3179 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3181 63. The daemon start-up calls getloadavg() while still root for those OS that
3182 need the first call to be done as root, but it missed one case: when
3183 deliver_queue_load_max is set with deliver_drop_privilege. This is
3184 necessary for the benefit of the queue runner, because there is no re-exec
3185 when deliver_drop_privilege is set.
3187 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3188 This has been fixed.
3190 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3191 was in use, was not putting the data itself into the right store pool;
3192 consequently, it could be overwritten for a subsequent message in the same
3193 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3196 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3198 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3199 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3200 the function were treating the return as a boolean value, which happened to
3201 work because 0=false and not-0=true, but is not correct code.
3203 68. The host_aton() function was not handling scoped IPv6 addresses (those
3204 with, for example, "%eth0" on the end) correctly.
3206 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3207 negated items (that is, ~something) in unsigned ints. Some compilers
3208 apparently mutter when there is no cast.
3210 70. If an address verification called from an ACL failed, and did not produce a
3211 user-specific message (i.e. there was only a "system" message), nothing was
3212 put in $acl_verify_message. In this situation, it now puts the system
3215 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3216 session; change 4.31/43 added the unwanted input to the log line - except
3217 that it did not do this in the start of session case. It now does.
3219 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3220 This is wrong and can cause the other end to generate a synchronization
3221 error if it is another Exim or anything else that does the synchronization
3222 check. A QUIT command is no longer sent after a timeout.
3224 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3225 during host lookups.
3227 74. The defer_ok option of callout verification was not working if it was used
3228 when verifying addresses in header lines, that is, for this case:
3230 verify = header_sender/callout=defer_ok
3232 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3233 those file descriptors could be used for SMTP connections. If anything
3234 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3235 could be sent to the SMTP client, causing chaos. The daemon now opens
3236 stdin, stdout, and stderr to /dev/null when it puts itself into the
3239 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3240 log by default. The user can override this with suitable Perl magic.
3242 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3243 the log message when discard triggers, was not working for the DATA ACL or
3244 for the non-SMTP ACL.
3246 78. Error message wording change in sieve.c.
3248 79. If smtp_accept_max_per_host was set, the number of connections could be
3249 restricted to fewer than expected, because the daemon was trying to set up
3250 a new connection before checking whether the processes handling previous
3251 connections had finished. The check for completed processes is now done
3252 earlier. On busy systems, this bug wouldn't be noticed because something
3253 else would have woken the daemon, and it would have reaped the completed
3256 80. If a message was submitted locally by a user whose login name contained one
3257 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3258 It caused a spool format error. I have fixed the spool reading code. A
3259 related problem was that the "from" clause in the Received: line became
3260 illegal because of the space(s). It is now covered by ${quote_local_part.
3262 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3265 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3267 83. Patch from Sieve maintainer to fix unterminated string problem in
3268 "vacation" handling.
3270 84. Some minor changes to the Linux configuration files to help with other
3271 OS variants using glibc.
3273 85. One more patch for Sieve to update vacation handling to latest spec.
3276 ----------------------------------------------------
3277 See the note above about the 4.44 and 4.50 releases.
3278 ----------------------------------------------------
3284 1. Change 4.43/35 introduced a bug that caused file counts to be
3285 incorrectly computed when quota_filecount was set in an appendfile
3288 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3289 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3292 3. Give more explanation in the error message when the command for a transport
3293 filter fails to execute.
3295 4. There are several places where Exim runs a non-Exim command in a
3296 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3297 was being done only for the command run by the queryprogram router. It is
3298 now done for all such subprocesses. The other cases are: ${run, transport
3299 filters, and the commands run by the lmtp and pipe transports.
3301 5. Some older OS have a limit of 256 on the maximum number of file
3302 descriptors. Exim was using setrlimit() to set 1000 as a large value
3303 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3304 systems. I've change it so that if it can't get 1000, it tries for 256.
3306 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3307 was an oversight, and furthermore, ever since the addition of extra
3308 controls (e.g. 4.43/32), the checks on when to allow different forms of
3309 "control" were broken. There should now be diagnostics for all cases when a
3310 control that does not make sense is encountered.
3312 7. $recipients is now available in the predata ACL (oversight).
3314 8. Tidy the search cache before the fork to do a delivery from a message
3315 received from the command line. Otherwise the child will trigger a lookup
3316 failure and thereby defer the delivery if it tries to use (for example) a
3317 cached ldap connection that the parent has called unbind on.
3319 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3320 of $address_data from the recipient verification was clobbered by the
3321 sender verification.
3323 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3324 was its contents. (It was OK if the option was not defined at all.)
3326 11. A "Completed" log line is now written for messages that are removed from
3327 the spool by the -Mrm option.
3329 12. $host_address is now set to the target address during the checking of
3330 ignore_target_hosts.
3332 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3333 being passed; this would have caused $sender_host_name to have been used if
3334 matching the list had actually called for a host name (not very likely,
3335 since this list is usually IP addresses). A host name is now passed as
3338 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3339 code that specifies a non-restarting handler (typically sigaction() in
3340 modern systems) in an attempt to fix a rare and obscure crash bug.
3342 15. Narrowed the window for a race in the daemon that could cause it to ignore
3343 SIGCHLD signals. This is not a major problem, because they are used only to
3344 wake it up if nothing else does.
3346 16. A malformed maildirsize file could cause Exim to calculate negative values
3347 for the mailbox size or file count. Odd effects could occur as a result.
3348 The maildirsize information is now recalculated if the size or filecount
3351 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3352 support for a long time. Removed HAVE_SYS_VFS_H.
3354 18. Updated exipick to current release from John Jetmore.
3356 19. Allow an empty sender to be matched against a lookup in an address list.
3357 Previously the only cases considered were a regular expression, or an
3360 20. Exim went into a mad DNS lookup loop when doing a callout where the
3361 host was specified on the transport, if the DNS lookup yielded more than
3364 21. The RFC2047 encoding function was originally intended for short strings
3365 such as real names; it was not keeping to the 75-character limit for
3366 encoded words that the RFC imposes. It now respects the limit, and
3367 generates multiple encoded words if necessary. To be on the safe side, I
3368 have increased the buffer size for the ${rfc2047: expansion operator from
3371 22. Failure to deliver a bounce message always caused it to be frozen, even if
3372 there was an errors_to setting on the router. The errors_to setting is now
3375 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3376 canonical form (fully expanded) before being placed in
3377 $sender_host_address.
3379 24. Updated eximstats to version 1.33
3381 25. Include certificate and key file names in error message when GnuTLS fails
3382 to set them up, because the GnuTLS error message doesn't include the name
3383 of the failing file when there is a problem reading it.
3385 26. Expand error message when OpenSSL has problems setting up cert/key files.
3388 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3389 (this can affect the format of dates).
3391 28. exim_tidydb, when checking for the continued existence of a message for
3392 which it has found a message-specific retry record, was not finding
3393 messages that were in split spool directories. Consequently, it was
3394 deleting retry records that should have stayed in existence.
3396 29. eximstats updated to version 1.35
3397 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3398 1.35 - bugfix such that pie charts by volume are generated correctly
3400 30. The SPA authentication driver was not abandoning authentication and moving
3401 on to the next authenticator when an expansion was forced to fail,
3402 contradicting the general specification for all authenticators. Instead it
3403 was generating a temporary error. It now behaves as specified.
3405 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3406 (the order specifies the preference for clients). The order is now AES256,
3407 AES128, 3DES, ARCFOUR128.
3409 31. Small patch to Sieve code - explicitly set From: when generating an
3412 32. Exim crashed if a remote delivery caused a very long error message to be
3413 recorded - for instance if somebody sent an entire SpamAssassin report back
3414 as a large number of 550 error lines. This bug was coincidentally fixed by
3415 increasing the size of one of Exim's internal buffers (big_buffer) that
3416 happened as part of the Exiscan merge. However, to be on the safe side, I
3417 have made the code more robust (and fixed the comments that describe what
3420 33. Some experimental protocols are using DNS PTR records for new purposes. The
3421 keys for these records are domain names, not reversed IP addresses. The
3422 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3423 leaves it alone. Component reversal etc. now happens only for IP addresses.
3426 34. The host_aton() function is supposed to be passed a string that is known
3427 to be a valid IP address. However, in the case of IPv6 addresses, it was
3428 not checking this. This is a hostage to fortune. Exim now panics and dies
3429 if the condition is not met. A case was found where this could be provoked
3430 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3431 components; fortuitously, this particular loophole had already been fixed
3432 by change 4.50/55 or 4.44/33 above.
3434 If there are any other similar loopholes, the new check in host_aton()
3435 itself should stop them being exploited. The report I received stated that
3436 data on the command line could provoke the exploit when Exim was running as
3437 exim, but did not say which command line option was involved. All I could
3438 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3439 running as the user.
3442 35. There was a buffer overflow vulnerability in the SPA authentication code
3443 (which came originally from the Samba project). I have added a test to the
3444 spa_base64_to_bits() function which I hope fixes it.
3447 36. The daemon start-up calls getloadavg() while still root for those OS that
3448 need the first call to be done as root, but it missed one case: when
3449 deliver_queue_load_max is set with deliver_drop_privilege. This is
3450 necessary for the benefit of the queue runner, because there is no re-exec
3451 when deliver_drop_privilege is set.
3453 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3454 was in use, was not putting the data itself into the right store pool;
3455 consequently, it could be overwritten for a subsequent message in the same
3456 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3459 38. Sometimes the final signoff response after QUIT could fail to get
3460 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3461 < 0 before doing a fflush(). This bug looks as though it goes back to the
3462 introduction of TLS in release 3.20, but "sometimes" must have been rare
3463 because the tests only now provoked it.
3469 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3470 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3471 Now it returns an integer. A number of calls were still expecting a T/F
3472 return. Fortuitously, in all cases, the tests worked in OK situations,
3473 which is the norm. However, things would have gone wrong on any write
3474 failures on the smtp file descriptor. This function is used when sending
3475 messages over SMTP and also when doing verify callouts.
3477 2. When Exim is called to do synchronous delivery of a locally submitted
3478 message (the -odf or -odi options), it no longer closes stderr before doing
3481 3. Implemented the mua_wrapper option.
3483 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3485 5. Implemented the functions header_remove(), header_testname(),
3486 header_add_at_position(), and receive_remove_recipient(), and exported them
3489 6. If an ACL "warn" statement specified the addition of headers, Exim already
3490 inserted X-ACL-Warn: at the start if there was no header name. However, it
3491 was not making this test for the second and subsequent header lines if
3492 there were newlines in the string. This meant that an invalid header could
3493 be inserted if Exim was badly configured.
3495 7. Allow an ACL "warn" statement to add header lines at the start or after all
3496 the Received: headers, as well as at the end.
3498 8. Added the rcpt_4xx retry error code.
3500 9. Added postmaster_mailfrom=xxx to callout verification option.
3502 10. Added mailfrom=xxxx to the callout verification option, for verify=
3505 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3506 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3508 12. Inserted some casts to stop certain compilers warning when using pointer
3509 differences as field lengths or precisions in printf-type calls (mostly
3510 affecting debugging statements).
3512 13. Added optional readline() support for -be (dynamically loaded).
3514 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3515 same clock tick as a message's arrival, so that its received time was the
3516 same as the "first fail" time on the retry record, and that message
3517 remained on the queue past the ultimate address timeout, every queue runner
3518 would try a delivery (because it was past the ultimate address timeout) but
3519 after another failure, the ultimate address timeout, which should have then
3520 bounced the address, did not kick in. This was a "< instead of <=" error;
3521 in most cases the first failure would have been in the next clock tick
3522 after the received time, and all would be well.
3524 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3525 being recognized when the domain list was tested by the match_domain
3526 condition in an expansion string.
3528 16. Added the ${str2b64: operator.
3530 17. Exim was always calling setrlimit() to set a large limit for the number of
3531 processes, without checking whether the existing limit was already
3532 adequate. (It did check for the limit on file descriptors.) Furthermore,
3533 errors from getrlimit() and setrlimit() were being ignored. Now they are
3534 logged to the main and panic logs, but Exim does carry on, to try to do its
3535 job under whatever limits there are.
3537 18. Imported PCRE 5.0.
3539 19. Trivial typo in log message " temporarily refused connection" (the leading
3542 20. If the log selector return_path_on_delivery was set and an address was
3543 redirected to /dev/null, the delivery process crashed because it assumed
3544 that a return path would always be set for a "successful" delivery. In this
3545 case, the whole delivery is bypassed as an optimization, and therefore no
3548 21. Internal re-arrangement: the function for sending a challenge and reading
3549 a response while authentication was assuming a zero-terminated challenge
3550 string. It's now changed to take a pointer and a length, to allow for
3551 binary data in such strings.
3553 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3555 23. Exim was not respecting finduser_retries when seeking the login of the
3556 uid under which it was called; it was always trying 10 times. (The default
3557 setting of finduser_retries is zero.) Also, it was sleeping after the final
3558 failure, which is pointless.
3560 24. Implemented tls_on_connect_ports.
3562 25. Implemented acl_smtp_predata.
3564 26. If the domain in control=submission is set empty, Exim assumes that the
3565 authenticated id is a complete email address when it generates From: or
3566 Sender: header lines.
3568 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3569 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3570 chown and chgrp in /bin and hostname in /usr/bin.
3572 28. Exim was keeping the "process log" file open after each use, just as it
3573 does for the main log. This opens the possibility of it remaining open for
3574 long periods when the USR1 signal hits a daemon. Occasional processlog
3575 errors were reported, that could have been caused by this. Anyway, it seems
3576 much more sensible not to leave this file open at all, so that is what now
3579 29. The long-running daemon process does not normally write to the log once it
3580 has entered its main loop, and it closes the log before doing so. This is
3581 so that log files can straightforwardly be renamed and moved. However,
3582 there are a couple of unusual error situations where the daemon does write
3583 log entries, and I had neglected to close the log afterwards.
3585 30. The text of an SMTP error response that was received during a remote
3586 delivery was being truncated at 512 bytes. This is too short for some of
3587 the long messages that one sometimes sees. I've increased the limit to
3590 31. It is now possible to make retry rules that apply only when a message has a
3591 specific sender, in particular, an empty sender.
3593 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3594 it possible to be selective about when SMTP synchronization is enforced.
3596 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3598 32. Implemented hosts_connection_nolog.
3600 33. Added an ACL for QUIT.
3602 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3605 35. Added mailbox_size and mailbox_filecount to appendfile.
3607 36. Added control = no_multiline_responses to ACLs.
3609 37. There was a bug in the logic of the code that waits for the clock to tick
3610 in the case where the clock went backwards by a substantial amount such
3611 that the microsecond fraction of "now" was more than the microsecond
3612 fraction of "then" (but the whole seconds number was less).
3614 38. Added support for the libradius Radius client library this is found on
3615 FreeBSD (previously only the radiusclient library was supported).
3621 1. When certain lookups returned multiple values in the form name=value, the
3622 quoting of the values was not always being done properly. Specifically:
3623 (a) If the value started with a double quote, but contained no whitespace,
3625 (b) If the value contained whitespace other than a space character (i.e.
3626 tabs or newlines or carriage returns) it was not quoted.
3627 This fix has been applied to the mysql and pgsql lookups by writing a
3628 separate quoting function and calling it from the lookup code. The fix
3629 should probably also be applied to nisplus, ibase and oracle lookups, but
3630 since I cannot test any of those, I have not disturbed their existing code.
3632 2. A hit in the callout cache for a specific address caused a log line with no
3633 reason for rejecting RCPT. Now it says "Previous (cached) callout
3634 verification failure".
3636 3. There was an off-by-one bug in the queryprogram router. An over-long
3637 return line was truncated at 256 instead of 255 characters, thereby
3638 overflowing its buffer with the terminating zero. As well as fixing this, I
3639 have increased the buffer size to 1024 (and made a note to document this).
3641 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3642 when Exim is waiting for an SMTP response from a remote server, Exim
3643 restarts its select() call on the socket, thereby resetting its timeout.
3644 This is not a problem when such interrupts are rare. Somebody set up a cron
3645 job to run exiwhat every 2 minutes, which is less than the normal select()
3646 timeout (5 or 10 minutes). This meant that the select() timeout never
3647 kicked in because it was always reset. I have fixed this by comparing the
3648 time when an interrupt arrives with the time at the start of the first call
3649 to select(). If more time than the timeout has elapsed, the interrupt is
3650 treated as a timeout.
3652 5. Some internal re-factoring in preparation for the addition of Sieve
3653 extensions (by MH). In particular, the "personal" test is moved to a
3654 separate function, and given an option for scanning Cc: and Bcc: (which is
3655 not set for Exim filters).
3657 6. When Exim created an email address using the login of the caller as the
3658 local part (e.g. when creating a From: or Sender: header line), it was not
3659 quoting the local part when it contained special characters such as @.
3661 7. Installed new OpenBSD configuration files.
3663 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3664 try to make them clearer.
3666 9. Callout options, other than the timeout value, were being ignored when
3667 verifying sender addresses in header lines. For example, when using
3669 verify = header_sender/callout=no_cache
3671 the cache was (incorrectly) being used.
3673 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3674 only the Cygwin environment.
3676 11. When return_path_on_delivery was set as a log selector, if different remote
3677 addresses in the same message used different return paths and parallel
3678 remote delivery occurred, the wrong values would sometimes be logged.
3679 (Whenever a remote delivery process finished, the return path value from
3680 the most recently started remote delivery process was logged.)
3682 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3683 lines when AUTH and/or TLS are in use. This is the "received protocol"
3684 field. Exim used to use "asmtp" for authenticated SMTP, without any
3685 indication (in the protocol name) for TLS use. Now it follows the RFC and
3686 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3687 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3688 names appear in log lines as well as in Received: header lines.
3690 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3691 extensions, and comparison tests, and to fix some bugs.
3693 14. Changes to the "personal" filter test:
3695 (1) The test was buggy in that it was just doing the equivalent of
3696 "contains" tests on header lines. For example, if a user's address was
3697 anne@some.where, the "personal" test would incorrectly be true for
3699 To: susanne@some.where
3701 This test is now done by extracting each address from the header in turn,
3702 and checking the entire address. Other tests that are part of "personal"
3703 are now done using regular expressions (for example, to check local parts
3704 of addresses in From: header lines).
3706 (2) The list of non-personal local parts in From: addresses has been
3707 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3708 taken from the Sieve specification recommendations.
3710 (3) If the message contains any header line starting with "List-" it is
3711 treated as non-personal.
3713 (4) The test for "circular" in the Subject: header line has been removed
3714 because it now seems ill-conceived.
3716 15. Minor typos in src/EDITME comments corrected.
3718 16. Installed latest exipick from John Jetmore.
3720 17. If headers_add on a router specified a text string that was too long for
3721 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3722 of string_sprintf() is now avoided.
3724 18. $message_body_size was not set (it was always zero) when running the DATA
3725 ACL and the local_scan() function.
3727 19. For the "mail" command in an Exim filter, no default was being set for
3728 the once_repeat time, causing a random time value to be used if "once" was
3729 specified. (If the value happened to be <= 0, no repeat happened.) The
3730 default is now 0s, meaning "never repeat". The "vacation" command was OK
3731 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3732 (I found it when inspecting the code).
3734 20. There is now an overall timeout for performing a callout verification. It
3735 defaults to 4 times the callout timeout, which applies to individual SMTP
3736 commands during the callout. The overall timeout applies when there is more
3737 than one host that can be tried. The timeout is checked before trying the
3738 next host. This prevents very long delays if there are a large number of
3739 hosts and all are timing out (e.g. when the network connections are timing
3740 out). The value of the overall timeout can be changed by specifying an
3741 additional sub-option for "callout", called "maxwait". For example:
3743 verify = sender/callout=5s,maxwait=20s
3745 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3746 to the end before writing, but this should make it even safer).
3748 22. Exim was forgetting that it had advertised PIPELINING for the second and
3749 subsequent messages on an SMTP connection. It was also not resetting its
3750 memory on STARTTLS and an internal HELO.
3752 23. When Exim logs an SMTP synchronization error within a session, it now
3753 records whether PIPELINING has been advertised or not.
3755 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3756 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3757 rather than long int.
3759 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3761 26. Added the never_mail option to autoreply.
3767 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3768 crash if the getsockname() call failed; this can happen if a connection is
3769 closed very soon after it is established. The problem was simply in the
3770 order in which certain operations were done, causing Exim to try to write
3771 to the SMTP stream before it had set up the file descriptor. The bug has
3772 been fixed by making things happen in the correct order.
3778 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3779 before the connection was closed, thus losing the rejection response.
3781 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3782 some early Solaris releases, but causes trouble in current releases where
3783 socklen_t is defined.
3785 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3788 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3791 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3792 that releasing the top of it at the end releases what was used for sub-
3793 expansions (unless the block got too big). However, discard this block if
3794 the first thing is a variable or header, so that we can use its block when
3795 it is dynamic (useful for very large $message_headers, for example).
3797 6. Lookups now cache *every* query, not just the most recent. A new, separate
3798 store pool is used for this. It can be recovered when all lookup caches are
3799 flushed. Lookups now release memory at the end of their result strings.
3800 This has involved some general refactoring of the lookup sources.
3802 7. Some code has been added to the store_xxx() functions to reduce the amount
3803 of flapping under certain conditions.
3805 8. log_incoming_interface used to affect only the <= reception log lines. Now
3806 it causes the local interface and port to be added to several more SMTP log
3807 lines, for example "SMTP connection from", and rejection lines.
3809 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3811 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3813 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3814 could be overwritten at the end of the current message (or the start of a
3815 new message if it was set in a HELO ACL). The value is now preserved for
3816 the duration of the SMTP connection.
3818 12. If a transport had a headers_rewrite setting, and a matching header line
3819 contained an unqualified address, that address was qualified, even if it
3820 did not match any rewriting rules. The underlying bug was that the values
3821 of the flags that permit the existence of unqualified sender and recipient
3822 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3823 non-local messages, and by -bnq for local messages) were not being
3824 preserved with the message after it was received.
3826 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3827 "next input=" as part of the text comprising the host identity instead of
3828 the correct text. The code was using the same buffer for two different
3829 strings. However, depending on which order the printing function evaluated
3830 its arguments, the bug did not always show up. Under Linux, for example, my
3831 test suite worked just fine.
3833 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3834 doesn't work with some older versions of Perl. It has been changed to "my",
3835 which in any case is probably the better facility to use.
3837 15. A really picky compiler found some instances of statements for creating
3838 error messages that either had too many or two few arguments for the format
3841 16. The size of the buffer for calls to the DNS resolver has been increased
3842 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3843 for addresses that have a lot of PTR records. This alleviates a problem; it
3844 does not fully solve it.
3846 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3847 buffer now truncates the list and logs the incident, which is the same
3848 action as happens when Exim is looking up a host name and its aliases.
3849 Previously in this situation something unpredictable would happen;
3850 sometimes it was "internal error: store_reset failed".
3852 18. If a server dropped the connection unexpectedly when an Exim client was
3853 using GnuTLS and trying to read a response, the client delivery process
3854 crashed while trying to generate an error log message.
3856 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3857 string, for example:
3859 warn message = H1: something\nH2: something
3861 the text was added as a single header line from Exim's point of view
3862 though it ended up OK in the delivered message. However, searching for the
3863 second and subsequent header lines using $h_h2: did not work. This has been
3864 fixed. Similarly, if a system filter added multiple headers in this way,
3865 the routers could not see them.
3867 20. Expanded the error message when iplsearch is called with an invalid key to
3868 suggest using net-iplsearch in a host list.
3870 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3871 ACLs are no longer actually imposed (and a message to that effect is
3874 22. If a "gecos" field in a passwd entry contained escaped characters, in
3875 particular, if it contained a \" sequence, Exim got it wrong when building
3876 a From: or a Sender: header from that name. A second bug also caused
3877 incorrect handling when an unquoted " was present following a character
3878 that needed quoting.
3880 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3881 was not being matched caselessly.
3883 24. Arranged for all hyphens in the exim.8 source to be escaped with
3886 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3887 properly. Recipient callout cache records were still being keyed to include
3888 the sender, even when use_sender was set false. This led to far more
3889 callouts that were necessary. The sender is no longer included in the key
3890 when use_sender is false.
3892 26. Added "control = submission" modifier to ACLs.
3894 27. Added the ${base62d: operator to decode base 62 numbers.
3896 28. dnsdb lookups can now access SRV records.
3898 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3899 the configuration file.
3901 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3902 (-v) mode. This makes the output for a verbose queue run more intelligible.
3904 31. Added a use_postmaster feature to recipient callouts.
3906 32. Added the $body_zerocount variable, containing the number of binary zero
3907 bytes in the message body.
3909 33. The time of last modification of the "new" subdirectory is now used as the
3910 "mailbox time last read" when there is a quota error for a maildir
3913 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3915 35. Added +ignore_unknown as a special item in host lists.
3917 36. Code for decoding IPv6 addresses in host lists is now included, even if
3918 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3919 address was recognized as an IP address, but was then not correctly decoded
3920 into binary, causing unexpected and incorrect effects when compared with
3927 1. Very minor rewording of debugging text in manualroute to say "list of
3928 hosts" instead of "hostlist".
3930 2. If verify=header_syntax was set, and a header line with an unqualified
3931 address (no domain) and a large number of spaces between the end of the
3932 name and the colon was received, the reception process suffered a buffer
3933 overflow, and (when I tested it) crashed. This was caused by some obsolete
3934 code that should have been removed. The fix is to remove it!
3936 3. When running in the test harness, delay a bit after writing a bounce
3937 message to get a bit more predictability in the log output.
3939 4. Added a call to search_tidyup() just before forking a reception process. In
3940 theory, someone could use a lookup in the expansion of smtp_accept_max_
3941 per_host which, without the tidyup, could leave open a database connection.
3943 5. Added the variables $recipient_data and $sender_data which get set from a
3944 lookup success in an ACL "recipients" or "senders" condition, or a router
3945 "senders" option, similar to $domain_data and $local_part_data.
3947 6. Moved the writing of debug_print from before to after the "senders" test
3950 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3951 problems for message scanning, either using a data ACL, or using
3952 local_scan() because the Received: header was not generated till after they
3953 were called (in order to set the time as the time of reception completion).
3954 I have revised the way this works. The header is now generated after the
3955 body is received, but before the ACL or local_scan() are called. After they
3956 are run, the timestamp in the header is updated.
3962 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3963 before starting a queue runner without re-exec. This happened only when
3964 deliver_drop_privilege was set or when the Exim user was set to root. The
3965 effect of the bug was that timeouts during subsequent deliveries caused
3966 crashes instead of being properly handled. The handler is now left at its
3967 default (and expected) setting.
3969 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3970 message, again when deliver_drop_privilege is set or Exim is run as root.
3971 The bug described in (1) was not present in this case, but the tidying up
3972 of the other signals was missing. I have made the two cases consistent.
3974 3. The ignore_target_hosts setting on a manualroute router was being ignored
3975 for hosts that were looked up using the /MX notation.
3977 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3980 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3981 operated on the sender address. After changing the $sender_address to <>
3982 for the sender address verify, Exim was re-instated it as the original
3983 (before rewriting) address, but remembering that it had rewritten it, so it
3984 wasn't rewriting it again. This bug also had the effect of breaking the
3985 sender address verification caching when the sender address was rewritten.
3987 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3988 This has been changed so that if the ip literal address matches
3989 ignore_target_hosts, the router declines.
3991 7. Added expansion conditions match_domain, match_address, and match_local_
3992 part (NOT match_host).
3994 8. The placeholder for the Received: header didn't have a length field set.
3996 9. Added code to Exim itself and to exim_lock to test for a specific race
3997 condition that could lead to file corruption when using MBX delivery. The
3998 issue is with the lockfile that is created in /tmp. If this file is removed
3999 after a process has opened it but before that process has acquired a lock,
4000 there is the potential for a second process to recreate the file and also
4001 acquire a lock. This could lead to two Exim processes writing to the file
4002 at the same time. The added code performs the same test as UW imapd; it
4003 checks after acquiring the lock that its file descriptor still refers to
4004 the same named file.
4006 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4007 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4010 11. Added the smtp_active_hostname option. If used, this will typically be made
4011 to depend on the incoming interface address. Because $interface_address is
4012 not set up until the daemon has forked a reception process, error responses
4013 that can happen earlier (such as "too many connections") no longer contain
4016 12. If an expansion in a condition on a "warn" statement fails because a lookup
4017 defers, the "warn" statement is abandoned, and the next ACL statement is
4018 processed. Previously this caused the whole ACL to be aborted.
4020 13. Added the iplsearch lookup type.
4022 14. Added ident_timeout as a log selector.
4024 15. Added tls_certificate_verified as a log selector.
4026 16. Added a global option tls_require_ciphers (compare the smtp transport
4027 option of the same name). This controls incoming TLS connections.
4029 17. I finally figured out how to make tls_require_ciphers do a similar thing
4030 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4031 before starting the TLS session.
4033 18. Tabs are now shown as \t in -bP output.
4035 19. If the log selector return_path_on_delivery was set, Exim crashed when
4036 bouncing a message because it had too many Received: header lines.
4038 20. If two routers both had headers_remove settings, and the first one included
4039 a superfluous trailing colon, the final name in the first list and the
4040 first name in the second list were incorrectly joined into one item (with a
4041 colon in the middle).
4047 1. Added -C and -D options to the exinext utility, mainly to make it easier
4048 to include in the automated testing, but these could be helpful when
4049 multiple configurations are in use.
4051 2. The exinext utility was not formatting the output nicely when there was
4052 an alternate port involved in the retry record key, nor when there was a
4053 message id as well (for retries that were specific to a specific message
4054 and a specific host). It was also confused by IPv6 addresses, because of
4055 the additional colons they contain. I have fixed the IPv4 problem, and
4056 patched it up to do a reasonable job for IPv6.
4058 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4059 delivery, the log line now contains "pipelined" if PIPELINING was used.
4061 4. An SMTP transport process used to panic and die if the bind() call to set
4062 an explicit outgoing interface failed. This has been changed; it is now
4063 treated in the same way as a connect() failure.
4065 5. A reference to $sender_host_name in the part of a conditional expansion
4066 that was being skipped was still causing a DNS lookup. This no longer
4069 6. The def: expansion condition was not recognizing references to header lines
4070 that used bh_ and bheader_.
4072 7. Added the _cache feature to named lists.
4074 8. The code for checking quota_filecount in the appendfile transport was
4075 allowing one more file than it should have been.
4077 9. For compatibility with Sendmail, the command line option
4085 and sets the incoming protocol and host name (for trusted callers). The
4086 host name and its colon can be omitted when only the protocol is to be set.
4087 Note the Exim already has two private options, -pd and -ps, that refer to
4088 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4089 "s", but I don't think that's a major issue.
4091 10. A number of refactoring changes to the code, none of which should affect
4094 (a) The number of logging options was getting close to filling up the
4095 32-bit word that was used as a bit map. I have split them into two classes:
4096 those that are passed in the argument to log_write(), and those that are
4097 only ever tested independently outside of that function. These are now in
4098 separate 32-bit words, so there is plenty of room for expansion again.
4099 There is no change in the user interface or the logging behaviour.
4101 (b) When building, for example, log lines, the code previously used a
4102 macro that called string_cat() twice, in order to add two strings. This is
4103 not really sufficiently general. Furthermore, there was one instance where
4104 it was actually wrong because one of the argument was used twice, and in
4105 one call a function was used. (As it happened, calling the function twice
4106 did not affect the overall behaviour.) The macro has been replaced by a
4107 function that can join an arbitrary number of extra strings onto a growing
4110 (c) The code for expansion conditions now uses a table and a binary chop
4111 instead of a serial search (which was left over from when there were very
4112 few conditions). Also, it now recognizes conditions like "pam" even when
4113 the relevant support is not compiled in: a suitably worded error message is
4114 given if an attempt is made to use such a condition.
4116 11. Added ${time_interval:xxxxx}.
4118 12. A bug was causing one of the ddress fields not to be passed back correctly
4119 from remote delivery subprocesses. The field in question was not being
4120 subsequently used, so this caused to problems in practice.
4122 13. Added new log selectors queue_time and deliver_time.
4124 14. Might have fixed a bug in maildirsizefile handling that threw up
4125 "unexpected character" debug warnings, and recalculated the data
4126 unnecessarily. In any case, I expanded the warning message to give more
4129 15. Added the message "Restricted characters in address" to the statements in
4130 the default ACL that block characters like @ and % in local parts.
4132 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4133 Three changes have been made:
4135 (a) There was a serious bug; a negative response to MAIL caused the whole
4136 recipient domain to be cached as invalid, thereby blocking all messages
4137 to all local parts at the same domain, from all senders. This bug has
4138 been fixed. The domain is no longer cached after a negative response to
4139 MAIL if the sender used is not empty.
4141 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4144 (c) A new callout option, "use_sender" has been added for people who want
4145 the modified behaviour.
4151 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4154 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4155 indeed breaks things for older releases.
4157 3. Added additional logging to the case where there is a problem reading data
4158 from a filter that is running in a subprocess using a pipe, in order to
4159 try to track down a specific problem.
4161 4. Testing facility fudge: when running in the test harness and attempting
4162 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4163 getting "No route to host". Convert this to a timeout.
4165 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4168 6. Some OS don't have socklen_t but use size_t instead. This affects the
4169 fifth argument of getsockopt() amongst other things. This is now
4170 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4171 can be set for individual OS. I have set it for SunOS5, OSF1, and
4172 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4173 some earlier ones do not.
4175 7. Change 4.30/15 was not doing the test caselessly.
4177 8. The standard form for an IPv6 address literal was being rejected by address
4178 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4179 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4180 this, as well as the form without the "IPv6" on the front (but only when
4181 address literals are enabled, of course).
4183 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4185 10. Exim crashed if a message with an empty sender address specified by -f
4186 encountered a router with an errors_to setting. This could be provoked only
4187 by a command such as
4191 where an empty string was supplied; "<>" did not hit this bug.
4193 11. Installed PCRE release 4.5.
4195 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4196 remained set. It is now erased.
4198 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4199 times from message ids (which are base 36 rather than the normal 62).
4201 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4202 were being counted as actual protocol errors, and logged if the log
4203 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4204 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4205 and DATA following a set of rejected RCPTs do not count as protocol errors.
4206 In other words, Exim assumes they were pipelined, though this may not
4207 actually be the case. Of course, in all cases the client gets an
4208 appropriate error code.
4210 15. If a lookup fails in an ACL condition, a message about the failure may
4211 be available; it is used if testing the ACL cannot continue, because most
4212 such messages specify what the cause of the deferral is. However, some
4213 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4214 that caused an old message to be retained and used if a later statement
4215 caused a defer, replacing the real cause of the deferral.
4217 16. If an IP address had so many PTR records that the DNS lookup buffer
4218 was not large enough to hold them, Exim could crash while trying to process
4219 the truncated data. It now detects and logs this case.
4221 17. Further to 4.21/58, another change has been made: if (and only if) the
4222 first line of a message (the first header line) ends with CRLF, a bare LF
4223 in a subsequent header line has a space inserted after it, so as not to
4224 terminate the header.
4226 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4227 unnecessarily, used atoi() instead of strtol(), and didn't check the
4228 termination when getting file sizes from file names by regex.
4230 19. Completely re-implemented the support for maildirsize files, in the light
4231 of a number of problems with the previous contributed implementation
4232 (4.30/29). In particular:
4234 . If the quota is zero, the maildirsize file is maintained, but no quota is
4237 . If the maildir directory does not exist, it is created before any attempt
4238 to write a maildirsize file.
4240 . The quota value in the file is just a cache; if the quota is changed in
4241 the transport, the new value overrides.
4243 . A regular expression is available for excluding directories from the
4246 20. The autoreply transport checks the characters in options that define the
4247 message's headers; it allows continued headers, but it was checking with
4248 isspace() after an embedded newline instead of explicitly looking for a
4251 21. If all the "regular" hosts to which an address was routed had passed their
4252 expiry times, and had not reached their retry times, the address was
4253 bounced, even if fallback hosts were defined. Now Exim should go on to try
4256 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4257 equivalent code in the SMTP transport. Some hosts send humungous responses
4258 to HELO/EHLO, more than 1024 it seems.
4260 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4261 gives compiler warnings in some environments. I've now done it "properly",
4264 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4265 (because of problems with the built-in one) was declared to return uschar *
4266 instead of char *, causing compiler failure.
4268 25. Fixed a file descriptor leak when processing alias/forward files.
4270 26. Fixed a minor format string issue in dbfn.c.
4272 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4274 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4275 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4278 29. When a sender address is verified, it is cached, to save repeating the test
4279 when there is more than one recipient in a message. However, when the
4280 verification involves a callout, it is possible for different callout
4281 options to be set for different recipients. It is too complicated to keep
4282 track of this in the cache, so now Exim always runs a verification when a
4283 callout is required, relying on the callout cache for the optimization.
4284 The overhead is duplication of the address routing, but this should not be
4287 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4288 to be verified with callout=postmaster, and the main callout worked but the
4289 postmaster check failed, the verification correctly failed. However, if a
4290 subsequent RCPT command asked for sender verification *without* the
4291 postmaster check, incorrect caching caused this verification also to fail,
4294 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4295 it was not caching the DNS options (qualify_single, search_parents) that
4296 were used when the lookup failed. A subsequent lookup with different
4297 options therefore always gave the same answer, though there were cases
4298 where it should not have. (Example: a "domains = !$mx_any" option on a
4299 dnslookup router: the "domains" option is always processed without any
4300 widening, but the router might have qualify_single set.) Now Exim uses the
4301 cached value only when the same options are set.
4303 32. Added John Jetmore's "exipick" utility to the distribution.
4305 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4306 than a timeout (e.g. a certificate is required, and is not provided), an
4307 Exim server now closes the connection immediately. Previously it waited for
4308 the client to close - but if the client is SSL, it seems that they each
4309 wait for each other, leading to a delay before one of them times out.
4311 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4312 maintained 0.8.x compatibility because I don't think many are using it, and
4313 it is clearly obsolete.
4315 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4318 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4319 client certificate was expired. A simple patch fixes this, though I don't
4320 understand the full logic of why the verify callback is called multiple
4323 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4324 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4325 which causes problems with some clients (such as the Certicom SSL Plus
4326 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4327 disables the coutermeasure allowing Eudora to connect."
4329 38. Exim was not checking that a write() to a log file succeeded. This could
4330 lead to Bad Things if a log got too big, in particular if it hit a file
4331 size limit. Exim now panics and dies if it cannot write to a log file, just
4332 as it does if it cannot open a log file.
4334 39. Modified OS/Makefile-Linux so that it now contains
4336 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4338 The two -D definitions ensure that Exim is compiled with large file
4339 support, which makes it possible to handle log files that are bigger than
4342 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4343 instance) a domain was checked against a named list that involved a lookup,
4344 causing $domain_data to be set, then another domain was checked against the
4345 same list, then the first domain was re-checked, the value of $domain_data
4346 after the final check could be wrong. In particular, if the second check
4347 failed, it could be set empty. This bug probably also applied to
4350 41. The strip_trailing_dot option was not being applied to the address given
4351 with the -f command-line option.
4353 42. The code for reading a message's header from the spool was incrementing
4354 $received_count, but never initializing it. This meant that the value was
4355 incorrect (doubled) while delivering a message in the same process in which
4356 it was received. In the most common configuration of Exim, this never
4357 happens - a fresh exec is done - but it can happen when
4358 deliver_drop_privilege is set.
4360 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4361 it now includes up to 150 characters of the unexpected data in the log
4364 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4365 and building data strings. The size of both of these buffers was 10 000
4366 bytes - far larger than anybody would *ever* want, thought I. Needless to
4367 say, somebody hit the limit. I have increased the maximum line length to
4368 20 000 and the maximum data length of concatenated lines to 100 000. I have
4369 also fixed two bugs, because there was no checking on these buffers. Tsk,
4370 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4371 buffer is too small.
4373 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4374 lsearch lookups. Now it does.
4376 46. When parsing a route_list item in a manualroute router, a fixed-length
4377 buffer was used for the list of hosts. I made this 1024 bytes long,
4378 thinking that nobody would ever have a list of hosts that long. Wrong.
4379 Somebody had a whole pile of complicated expansion conditions, and the
4380 string was silently truncated, leading to an expansion error. It turns out
4381 that it is easier to change to an unlimited length (owing to other changes
4382 that have happened since this code was originally written) than to build
4383 structure for giving a limitation error. The length of the item that
4384 expands into the list of hosts is now unlimited.
4386 47. The lsearch lookup could not handle data where the length of text line was
4387 more than 4095 characters. Such lines were truncated, leading to shortened
4388 data being returned. It should now handle lines of any length.
4390 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4391 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4394 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4395 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4396 no longer happens in comment lines. A list of replacements is now placed
4397 at the head of all of the source files, except those whose only change is
4398 to replace PERL_COMMAND in the very first #! line.
4400 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4401 messages on the queue, with a bottom-up merge sort, using code contributed
4402 by Michael Haardt. This should make operations like -bp somewhat faster on
4403 large queues. It won't affect queue runners, except when queue_run_in_order
4406 51. Installed eximstats 1.31 in the distribution.
4408 52. Added support for SRV lookups to the dnslookup router.
4410 53. If an ACL referred to $message_body or $message_body_end, the value was not
4411 reset for any messages that followed in the same SMTP session.
4413 54. The store-handling optimization for building very long strings was not
4414 differentiating between the different store pools. I don't think this
4415 actually made any difference in practice, but I've tidied it.
4417 55. While running the routers to verify a sender address, $sender_address
4418 was still set to the sender address. This is wrong, because when routing to
4419 send a bounce to the sender, it would be empty. Therefore, I have changed
4420 it so that, while verifying a sender address, $sender_address is set to <>.
4421 (There is no change to what happens when verifying a recipient address.)
4423 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4424 target A or AAAA records (if not already returned) without resetting the
4425 qualify_single or search_parents options of the DNS resolver. These are
4426 inappropriate in this case because the targets of MX and SRV records must
4427 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4428 target that, when qualified, matched something in the local domain. These
4429 two options are now turned off when doing these lookups.
4431 57. It seems that at least some releases of Reiserfs (which does not have the
4432 concept of a fixed number of inodes) returns zero and not -1 for the
4433 number of available inodes. This interacted badly with check_spool_inodes,
4434 which assumed that -1 was the "no such thing" setting. What I have done is
4435 to check that the total number of inodes is greater than zero before doing
4436 the test of how many are available.
4438 58. When a "warn" ACL statement has a log_message modifier, the message is
4439 remembered, and not repeated. This is to avoid a lot of repetition when a
4440 message has many recipients that cause the same warning to be written.
4441 Howewer, Exim was preserving the list of already written lines for an
4442 entire SMTP session, which doesn't seem right. The memory is now reset if a
4443 new message is started.
4445 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4446 debugging output unless log_rewrite was also set.
4448 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4449 of (int)(handle) when we know that handle contains (void *)(-1).
4451 61. The Exim daemon panic-logs an error return when it closes the incoming
4452 connection. However "connection reset by peer" seems to be common, and
4453 isn't really an error worthy of noting specially, so that particular error
4456 62. When Exim is trying to find all the local interfaces, it used to panic and
4457 die if the ioctl to get the interface flags failed. However, it seems that
4458 on at least one OS (Solaris 9) it is possible to have an interface that is
4459 included in the list of interfaces, but for which you get a failure error
4460 for this call. This happens when the interface is not "plumbed" into a
4461 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4462 failure of the "get flags" call assumes that the interface is down.
4464 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4465 makes life easier for people who are doing arithmetic on fields extracted
4466 from dates, where you often get leading zeros that should not be
4467 interpreted as octal.
4469 64. Added qualify_domain to the redirect router, to override the global
4472 65. If a pathologically long header line contained very many addresses (the
4473 report of this problem mentioned 10 000) and each of them was rewritten,
4474 Exim could use up a very large amount of memory. (It kept on making new
4475 copies of the header line as it rewrote, and never released the old ones.)
4476 At the expense of a bit more processing, the header rewriting function has
4477 been changed so that it no longer eats memory in this way.
4479 66. The generation of the Received: header has been moved from the time that a
4480 message starts to be received, to the time that it finishes. The timestamp
4481 in the Received: header should now be very close to that of the <= log
4482 line. There are two side-effects of this change:
4484 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4485 logged header lines no longer include the local Received: line, because
4486 it has not yet been created. The same applies to a copy of the message
4487 that is returned to a non-SMTP sender when a message is rejected.
4489 (b) When a filter file is tested using -bf, no additional Received: header
4490 is added to the test message. After some thought, I decided that this
4493 This change does not affect the value of $received_for. It is still set
4494 after address rewriting, but before local_scan() is called.
4496 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4498 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4499 gave an unhelpful panic error message, and a defer error. I have managed to
4500 change this behaviour so that it now rejects any supplied certificate,
4501 which seems right, as the list of acceptable certificates is empty.
4503 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4504 gave an unhelpful defer error. I have not managed to make this reject any
4505 supplied certificates, but the error message it gives is "no certificate
4506 supplied", which is not helpful.
4508 70. exigrep's output now also includes lines that are not associated with any
4509 message, but which match the given pattern. Implemented by a patch from
4510 Martin Sluka, which also tidied up the Perl a bit.
4512 71. Recipient callout verification, like sender verification, was using <> in
4513 the MAIL FROM command. This isn't really the right thing, since the actual
4514 sender may affect whether the remote host accepts the recipient or not. I
4515 have changed it to use the actual sender in the callout; this means that
4516 the cache record is now keyed on a recipient/sender pair, not just the
4517 recipient address. There doesn't seem to be a real danger of callout loops,
4518 since a callout by the remote host to check the sender would use <>.
4519 [SEE ABOVE: changed after hitting problems.]
4521 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4522 temporary errors. However, in the case of such a code being given after
4523 the end of a data transmission (i.e. after ".") Exim was failing to write
4524 a retry record for the message. (Yes, there was some broken host that was
4525 actually sending 8xx at this point.)
4527 73. An unknown lookup type in a host list could cause Exim to panic-die when
4528 the list was checked. (An example that provoked this was putting <; in the
4529 middle of a list instead of at the start.) If this happened during a DATA
4530 ACL check, a -D file could be left lying around. This kind of configuration
4531 error no longer causes Exim to die; instead it causes a defer errror. The
4532 incident is still logged to the main and panic logs.
4534 74. Buglet left over from Exim 3 conversion. The message "too many messages
4535 in one connection" was written to the rejectlog but not the mainlog, except
4536 when address rewriting (yes!) was being logged.
4538 75. Added write_rejectlog option.
4540 76. When a system filter was run not as root (that is, when system_filter_user
4541 was set), the values of the $n variables were not being returned to the
4542 main process; thus, they were not subsequently available in the $sn
4545 77. Added +return_path_on_delivery log selector.
4547 78. A connection timeout was being treated differently from recipients deferred
4548 when testing hosts_max_try with a message that was older than the host's
4549 retry timeout. (The host should not be counted, thus allowing all hosts to
4550 be tried at least once before bouncing.) This may have been the cause of an
4551 occasionally reported bug whereby a message would remain on the queue
4552 longer than the retry timeout, but would be bounced if a delivery was
4553 forced. I say "may" because I never totally pinned down the problem;
4554 setting up timeout/retry tests is difficult. See also the next item.
4556 79. The ultimate address timeout was not being applied to errors that involved
4557 a combination of host plus message (for example, a timeout on a MAIL
4558 command). When an address resolved to a number of possible hosts, and they
4559 were not all tried for each delivery (e.g. because of hosts_max_try), a
4560 message could remain on the queue longer than the retry timeout.
4562 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4565 81. Fixed an obscure SMTP outgoing bug which required at least the following
4566 conditions: (a) there was another message waiting for the same server;
4567 (b) the server returned 5xx to all RCPT commands in the first message so
4568 that the message was not completed; (c) the server dropped the connection
4569 or gave a negative response to the RSET that Exim sends to abort the
4570 transaction. The observed case was a dropped connection after DATA that had
4571 been sent in pipelining mode. That is, the server had advertised PIPELINING
4572 but was not implementing it correctly. The effect of the bug was incorrect
4573 behaviour, such as trying another host, and this could lead to a crash.
4579 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4580 and daemon.c were passed as pointers to ints; they should have been
4581 pointers to socklen_t variables (which are typically unsigned ints).
4583 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4586 3. Fixed a really odd bug that affected only the testing scheme; patching a
4587 certain fixed string in the binary changed the value of another string that
4588 happened to be identical to the end of the original first string.
4590 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4591 name", it returns that address as the IP address. On some operating
4592 systems (e.g. Solaris), it also passes back the IP address string as the
4593 "host name". However, on others (e.g. Linux), it passes back an empty
4594 string. Exim wasn't checking for this, and was changing the host name to an
4595 empty string, assuming it had been canonicized.
4597 5. Although rare, it is permitted to have more than one PTR record for a given
4598 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4599 all the names associated with an address, because they do in Solaris.
4600 However, it seems that they do not in Linux for data that comes from the
4601 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4602 I found this out when I moved to a new Linux workstation and tried to run
4603 the Exim test suite.
4605 To get round this problem I have changed the code so that it now does its
4606 own call to the DNS to look up PTR records when searching for a host name.
4607 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4608 addresses that are only in /etc/hosts are still found.
4610 This behaviour is, however, controlled by an option called host_lookup_
4611 order, which defaults to "bydns:byaddr". If people want to use the other
4612 order, or indeed, just use one or the other means of lookup, they can
4613 specify it in this variable.
4615 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4616 some operating systems, this comes back from gethostbyaddr() as an empty
4617 string, and this is what Exim used to test for. However, it seems that in
4618 other systems, "." is yielded. Exim now tests for this case too.
4620 7. The values of check_spool_space and check_log_space are now held internally
4621 as a number of kilobytes instead of an absolute number of bytes. If a
4622 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4623 kilobyte. This means that much larger values can be stored.
4625 8. Exim monitor: an attempt to get the action menu when not actually pointing
4626 at a message produces an empty menu entitled "No message selected". This
4627 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4628 no entries in it ("Shell widget menu has zero width and/or height"). So I
4629 have added a single, blank menu entry in this case.
4631 9. Added ${quote_local_part.
4633 10. MIME decoding is now applied to the contents of Subject: header lines when
4636 11. Now that a reference to $sender_host_address automatically causes a reverse
4637 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4638 host lookup before query-style lookups in lists that might use this
4639 variable. This has therefore been abolished, and the "net-" prefix is no
4640 longer necessary for query-style lookups.
4642 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4643 have been a typo for LFLAGS, so it has been changed.
4645 13. The install script calls Exim with "-C /dev/null" in order to find the
4646 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4647 to be output. Howeve, since Exim outputs its version number before the
4648 error, it didn't break the script. It just looked ugly. I fixed this by
4649 always allowing "-C /dev/null" if the caller is root.
4651 14. Ignore overlarge ACL variable number when reading spool file - insurance
4652 against a later release with more variables having written the file.
4654 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4655 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4656 this, as well as the form without the "IPv6" on the front.
4658 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4659 OS/Makefile-Darwin file.
4661 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4662 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4664 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4665 variable. However, some OS use ints for inodes. I've added cast to long int
4666 to get rid of the compiler warning.
4668 19. I had forgotten to lock out "/../" in configuration file names when
4669 ALT_CONFIG_PREFIX was set.
4671 20. Routers used for verification do not need to specify transports. However,
4672 if such a router generated a host list, and callout was configured, Exim
4673 crashed, because it could not find a port number from the (non-existent)
4674 transport. It now assumes port 25 in this circumstance.
4676 21. Added the -t option to exigrep.
4678 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4679 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4680 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4681 them is set, LOOKUP_LSEARCH is forced.
4683 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4685 24. Added sender and host information to the "rejected by local_scan()" log
4686 line; previously there was no indication of these.
4688 25. Added .include_if_exists.
4690 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4691 Linux. It turns out that not all file systems support this. Apparently some
4692 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4693 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4694 not supported on the file descriptor, is now ignored when Exim is trying to
4695 sync a directory. This applies only to Linux.
4697 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4699 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4700 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4703 29. Added support for maildirsize files from supplied patch (modified a bit).
4705 30. The use of :fail: followed by an empty string could lead Exim to respond to
4706 sender verification failures with (e.g.):
4708 550 Verification failed for <xxx>
4709 550 Sender verify failed
4711 where the first response line was missing the '-' that indicates it is not
4712 the final line of the response.
4714 31. The loop for finding the name of the user that called Exim had a hardwired
4715 limit of 10; it now uses the value of finduser_retries, which is used for
4716 all other user lookups.
4718 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4721 33. Exim was neglecting to zero errno before one call of strtol() when
4722 expanding a string and expecting an integer value. On some systems this
4723 resulted in spurious "integer overflow" errors. Also, it was casting the
4724 result into an int without checking.
4726 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4727 did not work. The code looks as if it has *never* worked, though it appears
4728 to have been documented since at least releast 1.62. I have made it work.
4730 35. The "timeout_DNS" error in retry rules, also documented since at least
4731 1.62, also never worked. As it isn't clear exactly what this means, and
4732 clearly it isn't a major issue, I have abolished the feature by treating it
4733 as "timeout", and writing a warning to the main and panic logs.
4735 36. The display of retry rules for -brt wasn't always showing the error code
4738 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4739 timeout_connect_A, timeout_connect_MX.
4741 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4742 to the empty sender.
4744 39. The daemon was not analysing the content of -oX till after it had closed
4745 stderr and disconnected from the controlling terminal. This meant that any
4746 syntax errors were only noted on the panic log, and the return code from
4747 the command was 0. By re-arranging the code a little, I've made the
4748 decoding happen first, so such errors now appear on stderr, and the return
4749 code is 1. However, the actual setting up of the sockets still happens in
4750 the disconnected process, so errors there are still only recorded on the
4753 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4754 connections (as happens on some IP stacks) was logged at start up time as
4755 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4756 it from "IPv6 and IPv4", which means that two separate sockets are being
4759 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4760 says whether AF_INET or AF_INET6 was passed as an argument.
4762 42. Exiwhat output was messed up when time zones were included in log
4765 43. Exiwhat now gives more information about the daemon's listening ports,
4766 and whether -tls-on-connect was used.
4768 44. The "port" option of the smtp transport is now expanded.
4770 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4771 silently ignored. Now an error message is written to the main and panic
4774 46. There's a new ACL modifier called "logwrite" which writes to a log file
4775 as soon as it is encountered.
4777 47. Added $local_user_uid and $local_user_gid at routing time.
4779 48. Exim crashed when trying to verify a sender address that was being
4782 49. Exim was recognizing only a space character after ".include". It now also
4783 recognizes a tab character.
4785 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4786 extracting the relevant information from the specification. The man page no
4787 longer contains scrambled data for the -d option, and I've added a section
4788 at the front about calling Exim under different names.
4790 51. Added "extra_headers" argument to the "mail" command in filter files.
4792 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4795 53. Installed eximstats 1.29.
4797 54. Added transport_filter_timeout as a generic transport option.
4799 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4800 Cc: header lines. This was required by RFC 822, but it not required by RFC
4803 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4804 incoming messages that did not have them. Now it does so only if the
4805 message originates locally, that is, if there is no associated remote host
4806 address. When Resent- header lines are present, this applies to the Resent-
4807 lines rather than the non-Resent- lines.
4809 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4810 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4812 58. Messages for configuration errors now include the name of the main
4813 configuration file - useful now that there may be more than one file in a
4814 list (.included file names were always shown).
4816 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4817 for those rare installations that do not start the daemon as root or run it
4818 setuid root. I've cut out the call to initgroups() if the daemon is not
4821 60. The Exim user and group can now be bound into the binary as text strings
4822 that are looked up at the start of Exim's processing.
4824 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4826 62. Added $mailstore_basename variable.
4828 63. Installed patch to sieve.c from Michael Haardt.
4830 64. When Exim failed to open the panic log after failing to open the main log,
4831 the original message it was trying to log was written to stderr and debug
4832 output, but if they were not available (the usual case in production), it
4833 was lost. Now it is written to syslog before the two lines that record the
4834 failures to open the logs.
4836 65. Users' Exim filters run in subprocesses under the user's uid. It is
4837 possible for a "deliver" command or an alias in a "personal" command to
4838 provoke an address rewrite. If logging of address rewriting is configured,
4839 this fails because the process is not running as root or exim. There may be
4840 a better way of dealing with this, but for the moment (because 4.30 needs
4841 to be released), I have disabled address rewrite logging when running a
4842 filter in a non-root, non-exim process.
4848 1. The buildconfig auxiliary program wasn't quoting the value set for
4849 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4850 not defined. This bug was masked in 4.22 by the effect that was fixed in
4853 2. Some messages that were rejected after a message id was allocated were
4854 shown as "incomplete" by exigrep. It no longer does this for messages that
4855 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4857 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4858 have allow_domain_literals set, the ID did not get logged in the <= line.
4859 Domain literals are now always recognized in Message-ID: header lines.
4861 4. The first argument for a ${extract expansion item is the key name or field
4862 number. Leading and trailing spaces in this item were not being ignored,
4863 causing some misleading effects.
4865 5. When deliver_drop_privilege was set, single queue runner processes started
4866 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4867 same command in the process it spins off) were not dropping privilege.
4869 6. When the daemon running as "exim" started a queue runner, it always
4870 re-executed Exim in the spun-off process. This is a waste of effort when
4871 deliver_drop_privilege is set. The new process now just calls the
4872 queue-runner function directly.
4878 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4881 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4882 ignored. Though the use of -oP was forcing the writing of a pid file, it
4883 was always written to the default place.
4885 3. If the message "no IP address found for host xxxx" is generated during
4886 incoming verification, it is now followed by identification of the incoming
4887 connection (so you can more easily find what provoked it).
4889 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4891 5. Added some features to "harden" Exim a bit more against certain attacks:
4893 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4894 be put in Local/Makefile. This is like the never_users runtime option,
4895 but it cannot be overridden. The default setting is "root".
4897 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4898 prefix string with which any file named in a -C command line option
4901 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4902 is retained for -C and -D only if the caller of Exim is root. Without
4903 it, the exim user may also use -C and -D and retain privilege.
4905 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4906 command line option is disabled.
4908 6. Macro names set by the -D option must start with an upper case letter, just
4909 like macro names defined in the configuration file.
4911 7. Added "dereference=" facility to LDAP.
4913 8. Two instances of the typo "uknown" in the source files are fixed.
4915 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4916 the Configure-Makefile script screwed up while processing it.
4918 10. Incorporated PCRE 4.4.
4920 11. The SMTP synchronization check was not operating right at the start of an
4921 SMTP session. For example, it could not catch a HELO sent before the client
4922 waited for the greeting. There is now a check for outstanding input at the
4923 point when the greeting is written. Because of the duplex, asynchronous
4924 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4925 way, but not yet received, when the check is performed.
4927 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4928 on TCP/IP sockets, because this apparently causes some broken clients to
4931 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4932 unchanged) from the Cygwin maintainer.
4934 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4935 was supported instead of testing for mbx. Effectively a typo.
4937 15. The spa authenticator server code was not checking that the input it
4938 received was valid base64.
4940 16. The debug output line for the "set" modifier in ACLs was not showing the
4941 name of the variable that was being set.
4943 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4945 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4946 lookup on its own. Something else was needed to trigger the lookup. For
4947 example, a match in host_lookup or the need for a host name in a host list.
4948 Now, if $sender_host_name is referenced and the host name has not yet been
4949 looked up, a lookup is performed. If the lookup fails, the variable remains
4950 empty, and $host_lookup_failed is set to "1".
4952 19. Added "eqi" as a case-independent comparison operator.
4954 20. The saslauthd authentication condition could segfault if neither service
4955 nor realm was specified.
4957 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4958 error message that was logged was misleading, and incoming SMTP
4959 connections were dropped. The message is now more accurate, and temporary
4960 errors are given to SMTP connections.
4962 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4963 (except RSET) with a 421 error, until QUIT is received. However, it was
4964 failing to send a response to QUIT.
4966 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4967 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4968 ACL is now run after the helo_try_verify_hosts code.
4970 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4971 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4972 case-independent, so other case variants are also recognized). Apparently
4973 some systems use these upper case variants.
4975 25. If more than two messages were waiting for the same host, and a transport
4976 filter was specified for the transport, Exim sent two messages over the
4977 same TCP/IP connection, and then failed with "socket operation on non-
4978 socket" when it tried to send the third.
4980 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4982 27. The extern definition of crypt16() in expand.c was not being excluded when
4983 the OS had its own crypt16() function.
4985 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4986 as a preferred synonym for return_size_limit, both as an option and as an
4989 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4991 30. Changed the default configuration ACL to relax the local part checking rule
4992 for addresses that are not in any local domains. For these addresses,
4993 slashes and pipe symbols are allowed within local parts, but the sequence
4994 /../ is explicitly forbidden.
4996 31. SPA server authentication was not clearing the challenge buffer before
4999 32. log_message in a "warn" ACL statement was writing to the reject log as
5000 well as to the main log, which contradicts the documentation and doesn't
5001 seem right (because no rejection is happening). So I have stopped it.
5003 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5004 However, I am unable to do any testing of this.
5006 34. Fixed an infelicity in the appendfile transport. When checking directories
5007 for a mailbox, to see if any needed to be created, it was accidentally
5008 using path names with one or more superfluous leading slashes; tracing
5009 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5011 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5012 contents are added to the log line that is written for every discarded
5013 recipient. (Previously a log_message setting was ignored.)
5015 36. The ${quote: operator now quotes the string if it is empty.
5017 37. The install script runs exim in order to find its version number. If for
5018 some reason other than non-existence or emptiness, which it checks, it
5019 could not run './exim', it was installing it with an empty version number,
5020 i.e. as "exim-". This error state is now caught, and the installation is
5023 38. An argument was missing from the function that creates an error message
5024 when Exim fails to connect to the socket for saslauthd authentication.
5025 This could cause Exim to crash, or give a corrupted message.
5027 39. Added isip, isip4, and isip6 to ${if conditions.
5029 40. The ACL variables $acl_xx are now saved with the message, and can be
5030 accessed later in routers, transports, and filters.
5032 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5033 strings in the file are not string-expanded.
5035 42. If a MAIL command specified a SIZE value that was too large to fit into an
5036 int variable, the check against message_size_limit failed. Such values are
5037 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5038 day this will have to be increased, but I don't think I want to be around
5039 when emails are that large.
5046 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5047 iconv() is not standard in FreeBSD.
5049 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5050 IPv6 enabled. The observed symptom was a segmentation fault on return from
5051 the function os_common_find_running_interfaces() in src/os.c.
5053 3. In the check_special_case() function in daemon.c I had used "errno" as an
5054 argument name, which causes warnings on some systems. This was basically a
5055 typo, since it was named "eno" in the comments!
5057 4. The code that waits for the clock to tick (at a resolution of some fraction
5058 of a second) so as to ensure message-id uniqueness was always waiting for
5059 at least one whole tick, when it could have waited for less. [This is
5060 almost certainly not relevant at current processor speeds, where it is
5061 unlikely to ever wait at all. But we try to future-proof.]
5063 5. The function that sleeps for a time interval that includes fractions of a
5064 second contained a race. It did not block SIGALRM between setting the
5065 timer, and suspending (a couple of lines later). If the interval was short
5066 and the sigsuspend() was delayed until after it had expired, the suspension
5067 never ended. On busy systems this could lead to processes getting stuck for
5070 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5071 process, before it forks any delivery processes. The open lookup caching
5072 mechanism meant that the open file or database connection was passed into
5073 the delivery process. The problem was that delivery processes always tidy
5074 up cached lookup data. This could cause a problem for the next delivery
5075 process started by the queue runner, because the external queue runner
5076 process does not know about the closure. So the next delivery process
5077 still has data in the lookup cache. In the case of a file lookup, there was
5078 no problem because closing a file descriptor in a subprocess doesn't affect
5079 the parent. However, if the lookup was caching a connection to a database,
5080 the connection was closed, and the second delivery process was likely to
5081 see errors such as "PGSQL: query failed: server closed the connection
5082 unexpectedly". The problem has been fixed by closing all cached lookups
5083 in a queue runner before running a delivery process.
5085 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5086 seem to have the "const" qualifier which it has on other OS. I've
5089 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5090 *of the same type* (client or server) with the same public name that an
5091 error should be diagnosed.
5093 9. When Exim looked up a host name for an IP address, but failed to find the
5094 original IP address when looking up the host name (a safety check), it
5095 output the message "<ip address> does not match any IP for NULL", which was
5096 confusing, to say the least. The bug was that the host name should have
5097 appeared instead of "NULL".
5099 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5100 user that is built into the binary, and the -C or -D options is used, root
5101 privilege is dropped before the configuration file is read. In addition,
5102 logging is switched to stderr instead of the normal log files. If the
5103 configuration then re-defines the Exim user, the unprivileged environment
5104 is probably not what is expected, so Exim logs a panic warning message (but
5107 However, if deliver_drop_privilege is set, the unprivileged state may well
5108 be exactly what is intended, so the warning has been cut out in that case,
5109 and Exim is allowed to try to write to its normal log files.
5115 1. smtp_return_error_details was not giving details for temporary sender
5116 or receiver verification errors.
5118 2. Diagnose a configuration error if two authenticators have the same public
5121 3. Exim used not to create the message log file for a message until the first
5122 delivery attempt. This could be confusing when incoming messages were held
5123 for policy or load reasons. The message log file is now created at the time
5124 the message is received, and an initial "Received" line is written to it.
5126 4. The automatically generated man page for command line options had a minor
5127 bug that caused no ill effects; however, a more serious problem was that
5128 the procedure for building the man page automatically didn't always
5129 operate. Consequently, release 4.20 contains an out-of-date version. This
5130 shouldn't happen again.
5132 5. When building Exim with embedded Perl support, the script that builds the
5133 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5134 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5136 6. The freeze_tell option was not being used for messages that were frozen on
5137 arrival, either by an ACL or by local_scan().
5139 7. Added the smtp_incomplete_transaction log selector.
5141 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5142 was accepting AUTH without a new EHLO.
5144 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5145 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5148 10. Logging of TCP/IP connections (when configured) now happens in the main
5149 daemon process instead of the child process, so that the TCP/IP connection
5150 count is more accurate (but it can never be perfect).
5152 11. The use of "drop" in a nested ACL was not being handled correctly in the
5153 outer ACL. Now, if condition failure induced by the nested "drop" causes
5154 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5155 or "require"), the connection is dropped.
5157 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5158 that yield "discard" can now be used with an "accept" or a "discard" verb,
5159 but an error is generated for any others (because I can't see a useful way
5160 to define what should happen).
5162 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5163 are now processed in the same way as lines in the Exim configuration file.
5164 In particular, continuation lines are supported.
5166 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5168 15. Added -ti meaning -t -i.
5170 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5171 domains, and warn by logging if others are found.
5173 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5174 returned from GIFCONF when Exim is trying to find the list of interfaces on
5175 a host. The code in os.c has been modified to copy each ifreq to an aligned
5176 structure in all cases.
5178 Also, in some cases, the returned ifreq's were being copied to a 'struct
5179 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5180 means the last couple of bytes of an IPv6 address could be chopped if the
5181 ifreq contained only a normal sockaddr (14 bytes storage).
5183 18. Named domain lists were not supported in the hosts_treat_as_local option.
5184 An entry such as +xxxx was not recognized, and was treated as a literal
5187 19. Ensure that header lines added by a DATA ACL are included in the reject log
5188 if the ACL subsequently rejects the message.
5190 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5191 MD5 (which is deprecated).
5193 21. When testing a filter file using -bf, Exim was writing a message when it
5194 took the sender from a "From " line in the message, but it was not doing so
5195 when it took $return_path from a Return-Path: header line. It now does.
5197 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5198 with a valid header line field name (a series of printing characters
5199 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5201 23. Changed "disc" in the source to "disk" to conform to the documentation and
5202 the book and for uniformity.
5204 24. Ignore Sendmail's -Ooption=value command line item.
5206 25. When execve() failed while trying to run a command in a pipe transport,
5207 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5208 could be confused with a return value of 69 from the command itself. This
5209 has been changed to 127, the value the shell returns if it is asked to run
5210 a non-existent command. The wording for the related log line suggests a
5211 non-existent command as the problem.
5213 26. If received_header_text expands to an empty string, do not add a Received:
5214 header line to the message. (Well, it adds a token one on the spool, but
5215 marks it "old" so that it doesn't get used or transmitted.)
5217 27. Installed eximstats 1.28 (addition of -nt option).
5219 28. There was no check for failure on the call to getsockname() in the daemon
5220 code. This can fail if there is a shortage of resources on the system, with
5221 ENOMEM, for example. A temporary error is now given on failure.
5223 29. Contrary to the C standard, it seems that in some environments, the
5224 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5225 program. Exim now does this explicitly; it affects the formatting of
5226 timestamps using strftime().
5228 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5229 complaints. I've now initialized all the variables, to avoid this.
5231 32. Header lines added by a system filter were not being "seen" during
5232 transport-time rewrites.
5234 33. The info_callback() function passed to OpenSSL is set up with type void
5235 (*)(SSL *, int, int), as described somewhere. However, when calling the
5236 function (actually a macro) that sets it up, the type void(*)() is
5237 expected. I've put in a cast to prevent warnings from picky compilers.
5239 34. If a DNS black list lookup found a CNAME record, but there were no A
5240 records associated with the domain it pointed at, Exim crashed.
5242 35. If a DNS black list lookup returned more than one A record, Exim ignored
5243 all but the first. It now scans all returned addresses if a particular IP
5244 value is being sought. In this situation, the contents of the
5245 $dnslist_value variable are a list of all the addresses, separated by a
5248 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5249 to accept a host name and all its aliases if the forward lookup for any of
5250 them yielded the IP address of the incoming connection. Now it accepts only
5251 those names whose forward lookup yields the correct IP address. Any other
5252 names are discarded. This closes a loophole whereby a rogue DNS
5253 administrator could create reverse DNS records to break through a
5254 wildcarded host restriction in an ACL.
5256 37. If a user filter or a system filter that ran in a subprocess used any of
5257 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5258 the wrong values were passed to the pipe command ($thisaddress had the
5259 value of $0, $0 had the value of $1, etc). This bug was introduced by
5260 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5262 38. Improved the line breaking for long SMTP error messages from ACLs.
5263 Previously, if there was no break point between 40 and 75 characters, Exim
5264 left the rest of the message alone. Two changes have been made: (a) I've
5265 reduced the minimum length to 35 characters; (b) if it can't find a break
5266 point between 35 and 75 characters, it looks ahead and uses the first one
5267 that it finds. This may give the occasional overlong line, but at least the
5268 remaining text gets split now.
5270 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5271 file descriptors might be low, and that setting 1000 would always raise it.
5272 It turns out that in some environments, the limit is already over 1000 and
5273 that lowering it causes trouble. So now Exim takes care not to decrease it.
5275 40. When delivering a message, the value of $return_path is set to $sender_
5276 address at the start of routing (routers may change the value). By an
5277 oversight, this default was not being set up when an address was tested by
5278 -bt or -bv, which affected the outcome if any router or filter referred to
5281 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5282 log only when "message" or "log_message" are set. However, if one of the
5283 conditions was an address verification, or a call to a nested ACL, the
5284 messages generated by the underlying test were being passed through. This
5285 no longer happens. The underlying message is available in $acl_verify_
5286 message for both "message" and "log_message" expansions, so it can be
5287 passed through if needed.
5289 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5290 new expansion $bh_ to give the encoded byte string without charset
5291 translation. Translation happens only if iconv() is available; HAVE_ICONV
5292 indicates this at build time. HEADERS_CHARSET gives the charset to
5293 translate to; headers_charset can change it in the configuration, and
5294 "headers charset" can change it in an individual filter file.
5296 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5297 that creates RFC 2047 encoded "words" labels them as that charset instead
5298 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5299 expansion operator; (ii) when Exim creates a From: line for a local
5300 message; (iii) when a header line is rewritten to include a "phrase" part.
5302 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5303 buggy, causing it to skip the first lines of messages whose message ID
5304 ended in 'D'. This would not have bitten before Exim release 4.14, because
5305 message IDs were unlikely to end in 'D' before then. The effect was to have
5306 incorrect size information for certain domains.
5308 45. #include "config.h" was missing at the start of the crypt16.c module. This
5309 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5312 46. If there was a timeout during a "random" callout check, Exim treated it as
5313 a failure of the random address, and carried on sending RSET and the real
5314 address. If the delay was just some slowness somewhere, the response to the
5315 original RCPT would be taken as a response to RSET and so on, causing
5316 mayhem of various kinds.
5318 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5319 when I implemented it. It didn't allow for the fact that some option values
5320 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5321 the right test for positive values.
5323 48. Domain names in DNS records are case-independent. Exim always looks them up
5324 in lower case. Some resolvers return domain names in exactly the case they
5325 appear in the zone file, that is, they may contain uppercase letters. Not
5326 all resolvers do this - some return always lower case. Exim was treating a
5327 change of case by a resolver as a change of domain, similar to a widening
5328 of a domain abbreviation. This triggered its re-routing code and so it was
5329 trying to route what was effectively the same domain again. This normally
5330 caused routing to fail (because the router wouldn't handle the domain
5331 twice). Now Exim checks for this case specially, and just changes the
5332 casing of the domain that it ultimately uses when it transmits the message
5335 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5338 50. If a filter generated a file delivery with a non-absolute name (possible if
5339 no home directory exists for the router), the forbid_file option was not
5342 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5343 the existing equality matching.
5345 52. Exim was using ints instead of ino_t variables in some places where it was
5346 dealing with inode numbers.
5348 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5349 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5350 called TMPDIR, and if it finds it is different, it changes its value.
5352 54. The smtp_printf() function is now made available to local_scan() so
5353 additional output lines can be written before returning. There is also an
5354 smtp_fflush() function to enable the detection of a dropped connection.
5355 The variables smtp_input and smtp_batched_input are exported to
5358 55. Changed the default runtime configuration: the message "Unknown user"
5359 has been removed from the ACL, and instead placed on the localuser router,
5360 using the cannot_route_message feature. This means that any verification
5361 failures that generate their own messages won't get overridden. Similarly,
5362 the "Unrouteable address" message that was in the ACL for unverifiable
5363 relay addresses has also been removed.
5365 56. Added hosts_avoid_esmtp to the smtp transport.
5367 57. The exicyclog script was not checking for the esoteric option
5368 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5369 will work only if exicyclog is run under the appropriate euid.
5371 58. Following a discussion on the list, the rules by which Exim recognises line
5372 endings on incoming messages have been changed. The -dropcr and drop_cr
5373 options are now no-ops, retained only for backwards compatibility. The
5374 following line terminators are recognized: LF CRLF CR. However, special
5375 processing applies to CR:
5377 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5378 nor a local message in the state where . is a terminator.
5380 (ii) If a bare CR is encountered in a header line, an extra space is added
5381 after the line terminator so as not to end the header. The reasoning
5382 behind this is that bare CRs in header lines are most likely either
5383 to be mistakes, or people trying to play silly games.
5385 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5386 was being incorrectly given as 18 bytes larger than it should have been.
5387 This is a VOB (very old bug).
5389 60. This may never have affected anything current, but just in case it has:
5390 When the local host is found other than at the start of a list of hosts,
5391 the local host, those with the same MX, and any that follow, are discarded.
5392 When the list in question was part of a longer list of hosts, the following
5393 hosts (not currently being processed) were also being discarded. This no
5394 longer happens. I'm not sure if this situation could ever has previously
5397 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5400 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5401 header. This is something that is recommended in a new Internet Draft, and
5402 is something that is documented as being done by Sendmail. There are two
5403 possible values. For messages generated by the autoreply transport, Exim
5406 Auto-Submitted: auto-replied
5408 whereas for all other generated messages (e.g. bounces) it adds
5410 Auto-Submitted: auto-generated
5412 63. The "personal" condition in filters now includes a test for the
5413 Auto-Submitted: header. If it contains the string "auto-" the message it
5414 not considered personal.
5416 64. Added rcpt_include_affixes as a generic transport option.
5418 65. Added queue_only_override (default true).
5420 66. Added the syslog_duplication option.
5422 67. If what should have been the first header line of a message consisted of
5423 a space followed by a colon, Exim was mis-interpreting it as a header line.
5424 It isn't of course - it is syntactically invalid and should therefore be
5425 treated as the start of the message body. The misbehaviour could have
5426 caused a number of strange effects, including loss of data in subsequent
5427 header lines, and spool format errors.
5429 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5430 client host had authenticated. This control can now be exercised by an ACL
5431 for more flexibility.
5433 69. By default, callouts do not happen when testing with -bh. There is now a
5434 variant, -bhc, which does actually run the callout code, including
5435 consulting and updating the callout cache.
5437 70. Added support for saslauthd authentication, courtesy of Alexander
5440 71. If statvfs() failed on the spool or log directories while checking their
5441 size for availability, Exim confusingly gave the error "space shortage".
5442 Furthermore, in debugging mode it crashed with a floating point exception.
5443 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5444 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5445 serious problem, Exim now writes to the main and panic logs when this
5446 happens, with details of the failure. It then refuses to accept the
5447 incoming message, giving the message "spool directory problem" or "log
5448 directory problem" with a 421 code for SMTP messages.
5450 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5451 0, 1, and 2 exist, because some OS complain for execs without them (see
5452 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5453 descriptors. However, the code omitted to check that the open succeeded,
5454 causing mysterious errors if for some reason the permissions on /dev/null
5455 got screwed. Now Exim writes a message to the main and panic logs, and
5456 bombs out if it can't open /dev/null.
5458 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5459 interact so that it is all more flexible. It is supposed to remain
5460 backwards compatible. Also added extra_local_interfaces.
5462 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5463 to bomb out with an assertion failure - to the client this appears as a
5464 connection drop. This problem occurs in the part of the code that was taken
5465 from the Samba project. Fortunately, the assertion is in a very simple
5466 function, so I have fixed this by reproducing the function inline in the
5467 one place where it is called, and arranging for authentication to fail
5468 instead of killing the process with assert().
5470 75. The SPA client code was not working when the server requested OEM rather
5471 than Unicode encoding.
5473 76. Added code to make require_files with a specific uid setting more usable in
5474 the case where statting the file as root fails - usually a non-root-mounted
5475 NFS file system. When this happens and the failure is EACCES, Exim now
5476 forks a subprocess and does the per-uid checking as the relevant uid.
5478 77. Added process_log_path.
5480 78. If log_file_path was not explicitly set, a setting of check_log_space or
5481 check_log_inodes was ignored.
5483 79. If a space check for the spool or log partitions fails, the incident is now
5484 logged. Of course, in the latter case the data may get lost...
5486 80. Added the %p formatting code to string_format() so that it can be used to
5487 print addresses in debug_print(). Adjusted all the address printing in the
5488 debugging in store.c to use %p rather than %d.
5490 81. There was a concern that a line of code in smtp_in.c could overflow a
5491 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5492 initially expressed, the concern was not well-founded, because trailing
5493 spaces are removed early. However, if the trailing spaces were followed by
5494 a NULL, they did not get removed, so the overflow was possible. Two fixes
5497 (a) I re-wrote the offending code in a cleaner fashion.
5498 (b) If an incoming SMTP command contains a NULL character, it is rejected
5501 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5502 runs initgroups(), so that if the Exim user is in any additional groups,
5503 they will be used during message reception.
5509 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob