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.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
70 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
72 Report and patch from Dmitry Banschikov.
74 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
75 Removed SSL_clear() after SSL_new() which led to protocol negotiation
76 failures. We appear to now support TLS1.1+ with Exim.
78 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
79 lets Exim select keys and certificates based upon TLS SNI from client.
80 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
81 before an outbound SMTP session. New log_selector, +tls_sni.
83 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
84 NULL dereference. Report and patch from Alun Jones.
86 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
87 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
88 Not seeing resolver debug output on NetBSD, but suspect this is a
89 resolver implementation change.
91 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
92 Left warnings. Added "eximon gdb" invocation mode.
94 PP/21 Defaulting "accept_8bitmime" to true, not false.
96 PP/22 Added -bw for inetd wait mode support.
98 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
99 locate the relevant includes and libraries. Made this the default.
101 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
102 Bugzilla 1246, report and most of solution from Tomasz Kusy.
104 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
105 This may cause build issues on older platforms.
107 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
108 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
109 gnutls_require_mac & gnutls_require_protocols (no longer supported).
110 Added SNI support via GnuTLS too.
111 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
113 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
115 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
116 Applied second patch from Janne, implementing suggestion to default
117 multiple-strings-in-record handling to match SPF spec.
119 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
121 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
122 read-only, out of scope).
123 Patch from Wolfgang Breyha, report from Stuart Northfield.
125 PP/29 Fix three issues highlighted by clang analyser static analysis.
126 Only crash-plausible issue would require the Cambridge-specific
127 iplookup router and a misconfiguration.
128 Report from Marcin Mirosław.
134 PP/01 Solaris build fix for Oracle's LDAP libraries.
135 Bugzilla 1109, patch from Stephen Usher.
137 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
139 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
142 TF/02 Fix a couple more cases where we did not log the error message
143 when unlink() failed. See also change 4.74-TF/03.
145 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
146 lock up or crash if it happened to be inside a call to libc when it
147 got a SIGUSR1 from exiwhat.
149 The SIGUSR1 handler appends the current process status to the process
150 log which is later printed by exiwhat. It used to use the general
151 purpose logging code to do this, but several functions it calls are
152 not safe for signals.
154 The new output code in the SIGUSR1 handler is specific to the process
155 log, and simple enough that it's easy to inspect for signal safety.
156 Removing some special cases also simplifies the general logging code.
157 Removing the spurious timestamps from the process log simplifies
160 TF/04 Improved ratelimit ACL condition.
162 The /noupdate option has been deprecated in favour of /readonly which
163 has clearer semantics. The /leaky, /strict, and /readonly update modes
164 are mutually exclusive. The update mode is no longer included in the
165 database key; it just determines when the database is updated. (This
166 means that when you upgrde Exim will forget old rate measurements.)
168 Exim now checks that the per_* options are used with an update mode that
169 makes sense for the current ACL. For example, when Exim is processing a
170 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
171 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
172 must specify per_mail/readonly. If you omit the update mode it defaults to
173 /leaky where that makes sense (as before) or /readonly where required.
175 The /noupdate option is now undocumented but still supported for
176 backwards compatibility. It is equivalent to /readonly except that in
177 ACLs where /readonly is required you may specify /leaky/noupdate or
178 /strict/noupdate which are treated the same as /readonly.
180 A useful new feature is the /count= option. This is a generalization
181 of the per_byte option, so that you can measure the throughput of other
182 aggregate values. For example, the per_byte option is now equivalent
183 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
185 The per_rcpt option has been generalized using the /count= mechanism
186 (though it's more complicated than the per_byte equivalence). When it is
187 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
188 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
189 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
190 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
191 Note that using per_rcpt with a non-readonly update mode in more than
192 one ACL will cause the recipients to be double-counted. (The per_mail
193 and per_byte options don't have this problem.)
195 The handling of very low rates has changed slightly. If the computed rate
196 is less than the event's count (usually one) then this event is the first
197 after a long gap. In this case the rate is set to the same as this event's
198 count, so that the first message of a spam run is counted properly.
200 The major new feature is a mechanism for counting the rate of unique
201 events. The new per_addr option counts the number of different
202 recipients that someone has sent messages to in the last time period. It
203 behaves like per_rcpt if all the recipient addresses are different, but
204 duplicate recipient addresses do not increase the measured rate. Like
205 the /count= option this is a general mechanism, so the per_addr option
206 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
207 example, measure the rate that a client uses different sender addresses
208 with the options per_mail/unique=$sender_address. There are further
209 details in the main documentation.
211 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
213 TF/06 Removed a few PCRE remnants.
215 TF/07 Automatically extract Exim's version number from tags in the git
216 repository when doing development or release builds.
218 PP/02 Raise smtp_cmd_buffer_size to 16kB.
219 Bugzilla 879. Patch from Paul Fisher.
221 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
222 Heavily based on revision 40f9a89a from Simon Arlott's tree.
225 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
227 PP/05 Variable $av_failed, true if the AV scanner deferred.
228 Bugzilla 1078. Patch from John Horne.
230 PP/06 Stop make process more reliably on build failure.
231 Bugzilla 1087. Patch from Heiko Schlittermann.
233 PP/07 Make maildir_use_size_file an _expandable_ boolean.
234 Bugzilla 1089. Patch from Heiko Schlittermann.
236 PP/08 Handle ${run} returning more data than OS pipe buffer size.
237 Bugzilla 1131. Patch from Holger Weiß.
239 PP/09 Handle IPv6 addresses with SPF.
240 Bugzilla 860. Patch from Wolfgang Breyha.
242 PP/10 GnuTLS: support TLS 1.2 & 1.1.
244 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
247 PP/11 match_* no longer expand right-hand-side by default.
248 New compile-time build option, EXPAND_LISTMATCH_RHS.
249 New expansion conditions, "inlist", "inlisti".
251 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
253 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
255 PP/14 fix log_write() format string regression from TF/03.
256 Bugzilla 1152. Patch from Dmitry Isaikin.
262 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
264 PP/02 Harmonised TLS library version reporting; only show if debugging.
265 Layout now matches that introduced for other libraries in 4.74 PP/03.
267 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
269 PP/04 New "dns_use_edns0" global option.
271 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
274 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
275 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
277 TK/01 Updated PolarSSL code to 0.14.2.
278 Bugzilla 1097. Patch from Andreas Metzler.
280 PP/07 Catch divide-by-zero in ${eval:...}.
283 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
286 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
287 format-string attack -- SECURITY: remote arbitrary code execution.
289 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
290 time unintentionally subject to list matching rules, letting the header
291 cause arbitrary Exim lookups (of items which can occur in lists, *not*
292 arbitrary string expansion). This allowed for information disclosure.
294 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
295 INT_MIN/-1 -- value coerced to INT_MAX.
301 NM/01 Workround for PCRE version dependency in version reporting
304 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
305 This fixes portability to compilers other than gcc, notably
306 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
308 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
309 makefiles for portability to HP-UX and POSIX correctness.
311 PP/01 Permit LOOKUP_foo enabling on the make command-line.
312 Also via indented variable definition in the Makefile.
313 (Debugging by Oliver Heesakkers).
315 PP/02 Restore caching of spamd results with expanded spamd_address.
316 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
318 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
319 Improves build reliability. Fix from: Frank Elsner
321 NM/02 Fix wide character breakage in the rfc2047 coding
322 Fixes bug 1064. Patch from Andrey N. Oktyabrski
324 NM/03 Allow underscore in dnslist lookups
325 Fixes bug 1026. Patch from Graeme Fowler
327 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
328 Code patches from Adam Ciarcinski of NetBSD.
330 NM/04 Fixed exiqgrep to cope with mailq missing size issue
333 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
334 is logged, to avoid truncation. Patch from John Horne.
336 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
337 Patch from Jakob Hirsch.
339 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
340 SQL string expansion failure details.
341 Patch from Andrey Oktyabrski.
343 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
344 Patch from Simon Arlott.
346 PP/09 New lookups functionality failed to compile on old gcc which rejects
347 extern declarations in function scope.
348 Patch from Oliver Fleischmann
350 PP/10 Use sig_atomic_t for flags set from signal handlers.
351 Check getgroups() return and improve debugging.
352 Fixed developed for diagnosis in bug 927 (which turned out to be
355 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
356 Patch from Mark Zealey.
358 PP/12 Bugzilla 1056: Improved spamd server selection.
359 Patch from Mark Zealey.
361 PP/13 Bugzilla 1086: Deal with maildir quota file races.
362 Based on patch from Heiko Schlittermann.
364 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
365 Patch from Uwe Doering, sign-off by Michael Haardt.
367 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
368 variable declaration deep within a block. Bug and patch from
371 PP/15 lookups-Makefile IRIX compatibilty coercion.
373 PP/16 Make DISABLE_DKIM build knob functional.
375 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
376 Patch by Simon Arlott
378 TF/03 Fix valgrind.h portability to C89 compilers that do not support
379 variable argument macros. Our copy now differs from upstream.
385 TF/01 Failure to get a lock on a hints database can have serious
386 consequences so log it to the panic log.
388 TF/02 Log LMTP confirmation messages in the same way as SMTP,
389 controlled using the smtp_confirmation log selector.
391 TF/03 Include the error message when we fail to unlink a spool file.
393 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
394 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
395 for maintaining out-of-tree patches for some time.
397 PP/01 Bugzilla 139: Documentation and portability issues.
398 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
399 Handle per-OS dynamic-module compilation flags.
401 PP/02 Let /dev/null have normal permissions.
402 The 4.73 fixes were a little too stringent and complained about the
403 permissions on /dev/null. Exempt it from some checks.
404 Reported by Andreas M. Kirchwitz.
406 PP/03 Report version information for many libraries, including
407 Exim version information for dynamically loaded libraries. Created
408 version.h, now support a version extension string for distributors
409 who patch heavily. Dynamic module ABI change.
411 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
412 privilege escalation vulnerability whereby the Exim run-time user
413 can cause root to append content of the attacker's choosing to
416 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
419 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
420 If dropping privileges for untrusted macros, we disabled normal logging
421 on the basis that it would fail; for the Exim run-time user, this is not
422 the case, and it resulted in successful deliveries going unlogged.
423 Fixed. Reported by Andreas Metzler.
429 PP/01 Date: & Message-Id: revert to normally being appended to a message,
430 only prepend for the Resent-* case. Fixes regression introduced in
431 Exim 4.70 by NM/22 for Bugzilla 607.
433 PP/02 Include check_rfc2047_length in configure.default because we're seeing
434 increasing numbers of administrators be bitten by this.
436 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
438 PP/03 Bugzilla 994: added openssl_options main configuration option.
440 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
442 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
444 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
446 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
447 without a peer certificate, leading to a segfault because of an
448 assumption that peers always have certificates. Be a little more
449 paranoid. Problem reported by Martin Tscholak.
451 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
452 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
453 NB: ClamAV planning to remove STREAM in "middle of 2010".
454 CL also introduces -bmalware, various -d+acl logging additions and
455 more caution in buffer sizes.
457 PP/09 Implemented reverse_ip expansion operator.
459 PP/10 Bugzilla 937: provide a "debug" ACL control.
461 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
463 PP/12 Bugzilla 973: Implement --version.
465 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
467 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
469 PP/15 Bugzilla 816: support multiple condition rules on Routers.
471 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
472 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
473 ignore trailing whitespace.
475 JJ/02 prevent non-panic DKIM error from being sent to paniclog
477 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
480 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
481 Notification from Dr Andrew Aitchison.
483 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
484 ExtendedDetectionInfo response format.
485 Notification from John Horne.
487 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
490 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
491 XSL and documented dependency on system catalogs, with examples of how
494 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
497 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
498 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
499 configuration file which is writeable by the Exim user or group.
501 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
502 of configuration files to cover files specified with the -C option if
503 they are going to be used with root privileges, not just the default
506 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
507 option (effectively making it always true).
509 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
510 files to be used while preserving root privileges.
512 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
513 that rogue child processes cannot use them.
515 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
516 run-time user, instead of root.
518 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
519 Exim run-time user without dropping privileges.
521 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
522 result string, instead of calling string_vformat() twice with the same
525 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
526 for other users. Others should always drop root privileges if they use
527 -C on the command line, even for a whitelisted configure file.
529 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
531 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
537 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
538 $data_path, and $header_path variables; fixed documentation bugs and
541 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
542 exipick to access non-standard spools, including the "frozen" queue
545 NM/01 Bugzilla 965: Support mysql stored procedures.
546 Patch from Alain Williams
548 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
550 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
551 Patch from Andreas Metzler
553 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
554 Patch from Kirill Miazine
556 NM/05 Bugzilla 671: Added umask to procmail example.
558 JJ/03 installed exipick 20100323.0, fixing doc bug
560 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
561 directory. Notification and patch from Dan Rosenberg.
563 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
565 TK/02 Improve log output when DKIM signing operation fails.
567 MH/01 Treat the transport option dkim_domain as a colon separated
568 list, not as a single string, and sign the message with each element,
569 omitting multiple occurences of the same signer.
571 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
572 Bugzilla 985, 986. Patch by Simon Arlott
574 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
575 Patch by Simon Arlott
577 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
578 MBX locking. Notification from Dan Rosenberg.
584 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
586 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
588 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
590 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
592 NM/04 Bugzilla 915: Flag broken perl installation during build.
598 TK/01 Added patch by Johannes Berg that expands the main option
599 "spamd_address" if it starts with a dollar sign.
601 TK/02 Write list of recipients to X-Envelope-Sender header when building
602 the mbox-format spool file for content scanning (suggested by Jakob
605 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
606 (http://www.dcc-servers.net/) support via dccifd. Activated by
607 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
609 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
610 by Mark Daniel Reidel <mr@df.eu>.
612 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
613 When building exim an external PCRE library is now needed -
614 PCRE is a system library on the majority of modern systems.
615 See entry on PCRE_LIBS in EDITME file.
617 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
618 conversation. Added nologin parameter to request.
619 Patch contributed by Kirill Miazine.
621 TF/01 Do not log submission mode rewrites if they do not change the address.
623 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
625 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
626 log files in place. Contributed by Roberto Lima.
628 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
630 TF/03 Bugzilla 615: When checking the local_parts router precondition
631 after a local_part_suffix or local_part_prefix option, Exim now
632 does not use the address's named list lookup cache, since this
633 contains cached lookups for the whole local part.
635 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
636 Robert Millan. Documentation is in experimental-spec.txt.
638 TF/04 Bugzilla 668: Fix parallel build (make -j).
640 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
642 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
643 Patch provided by Jan Srzednicki.
645 TF/05 Leading white space used to be stripped from $spam_report which
646 wrecked the formatting. Now it is preserved.
648 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
649 that they are available at delivery time.
651 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
653 TF/08 TLS error reporting now respects the incoming_interface and
654 incoming_port log selectors.
656 TF/09 Produce a more useful error message if an SMTP transport's hosts
657 setting expands to an empty string.
659 NM/06 Bugzilla 744: EXPN did not work under TLS.
660 Patch provided by Phil Pennock.
662 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
663 Patch provided by Richard Godbee.
665 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
666 acl_smtp_notquit, added index entry.
668 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
669 Patch provided by Eugene Bujak.
671 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
672 accept(). Patch provided by Maxim Dounin.
674 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
675 Patch provided by Phil Pennock.
677 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
679 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
680 Patch provided by Brad "anomie" Jorsch.
682 NM/14 Bugzilla 622: Added timeout setting to transport filter.
683 Patch provided by Dean Brooks.
685 TK/05 Add native DKIM support (does not depend on external libraries).
687 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
688 Patch provided by Graeme Fowler.
690 NM/16 Bugzilla 851: Documentation example syntax fix.
692 NM/17 Changed NOTICE file to remove references to embedded PCRE.
694 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
697 NM/19 Bugzilla 745: TLS version reporting.
698 Patch provided by Phil Pennock.
700 NM/20 Bugzilla 167: bool: condition support.
701 Patch provided by Phil Pennock.
703 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
704 clients. Patch provided by Phil Pennock.
706 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
707 Patch provided by Brad "anomie" Jorsch.
709 NM/23 Bugzilla 687: Fix misparses in eximstats.
710 Patch provided by Heiko Schlittermann.
712 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
713 Patch provided by Heiko Schlittermann.
715 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
716 plus update to original patch.
718 NM/26 Bugzilla 799: Documentation correction for ratelimit.
720 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
721 Patch provided by David Brownlee.
723 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
725 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
727 NM/30 Bugzilla 888: TLS documentation bugfixes.
729 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
731 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
732 Unlike the original bugzilla I have changed all shell scripts in src tree.
734 NM/33 Bugzilla 898: Transport filter timeout fix.
735 Patch by Todd Rinaldo.
737 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
738 Patch by Serge Demonchaux.
740 NM/35 Bugzilla 39: Base64 decode bug fixes.
741 Patch by Jakob Hirsch.
743 NM/36 Bugzilla 909: Correct connect() call in dcc code.
745 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
747 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
749 NM/39 Bugzilla 911: Fixed MakeLinks build script.
755 TK/01 Add preliminary DKIM support. Currently requires a forked version of
756 ALT-N's libdkim that I have put here:
757 http://duncanthrax.net/exim-experimental/
759 Note to Michael Haardt: I had to rename some vars in sieve.c. They
760 were called 'true' and it seems that C99 defines that as a reserved
761 keyword to be used with 'bool' variable types. That means you could
762 not include C99-style headers which use bools without triggering
763 build errors in sieve.c.
765 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
766 as mailq or other aliases. Changed the --help handling significantly
767 to do whats expected. exim_usage() emits usage/help information.
769 SC/01 Added the -bylocaldomain option to eximstats.
771 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
773 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
775 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
781 PH/01 Another patch from the Sieve maintainer.
783 PH/02 When an IPv6 address is converted to a string for single-key lookup
784 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
785 dots are used instead of colons so that keys in lsearch files need not
786 contain colons. This was done some time before quoting was made available
787 in lsearch files. However, iplsearch files do require colons in IPv6 keys
788 (notated using the quote facility) so as to distinguish them from IPv4
789 keys. This meant that lookups for IP addresses in host lists did not work
790 for iplsearch lookups.
792 This has been fixed by arranging for IPv6 addresses to be expressed with
793 colons if the lookup type is iplsearch. This is not incompatible, because
794 previously such lookups could never work.
796 The situation is now rather anomolous, since one *can* have colons in
797 ordinary lsearch keys. However, making the change in all cases is
798 incompatible and would probably break a number of configurations.
800 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
803 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
804 conversion specification without a maximum field width, thereby enabling
805 a rogue spamd server to cause a buffer overflow. While nobody in their
806 right mind would setup Exim to query an untrusted spamd server, an
807 attacker that gains access to a server running spamd could potentially
808 exploit this vulnerability to run arbitrary code as the Exim user.
810 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
811 $primary_hostname instead of what libspf2 thinks the hosts name is.
813 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
814 a directory entry by the name of the lookup key. Previously, if a
815 symlink pointed to a non-existing file or a file in a directory that
816 Exim lacked permissions to read, a lookup for a key matching that
817 symlink would fail. Now it is enough that a matching directory entry
818 exists, symlink or not. (Bugzilla 503.)
820 PH/03 The body_linecount and body_zerocount variables are now exported in the
823 PH/04 Added the $dnslist_matched variable.
825 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
826 This means they are set thereafter only if the connection becomes
829 PH/06 Added the client_condition to authenticators so that some can be skipped
830 by clients under certain conditions.
832 PH/07 The error message for a badly-placed control=no_multiline_responses left
833 "_responses" off the end of the name.
835 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
837 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
838 (without spaces) instead of just copying the configuration text.
840 PH/10 Added the /noupdate option to the ratelimit ACL condition.
842 PH/11 Added $max_received_linelength.
844 PH/12 Added +ignore_defer and +include_defer to host lists.
846 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
847 way in which PCRE > 7.0 is built.
849 PH/14 Implemented queue_only_load_latch.
851 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
852 MAIL command. The effect was to mangle the value on 64-bit systems.
854 PH/16 Another patch from the Sieve maintainer.
856 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
858 PH/18 If a system quota error occurred while trying to create the file for
859 a maildir delivery, the message "Mailbox is full" was not appended to the
860 bounce if the delivery eventually timed out. Change 4.67/27 below applied
861 only to a quota excession during the actual writing of the file.
863 PH/19 It seems that peer DN values may contain newlines (and other non-printing
864 characters?) which causes problems in log lines. The DN values are now
865 passed through string_printing() before being added to log lines.
867 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
868 and InterBase are left for another time.)
870 PH/21 Added message_body_newlines option.
872 PH/22 Guard against possible overflow in moan_check_errorcopy().
874 PH/23 POSIX allows open() to be a macro; guard against that.
876 PH/24 If the recipient of an error message contained an @ in the local part
877 (suitably quoted, of course), incorrect values were put in $domain and
878 $local_part during the evaluation of errors_copy.
884 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
885 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
888 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
889 issue a MAIL command.
891 PH/02 In an ACL statement such as
893 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
895 if a client was not listed at all, or was listed with a value other than
896 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
897 the condition was not true (as it should be), so access was not denied.
898 The bug was that the ! inversion was incorrectly passed on to the second
899 item. This has been fixed.
901 PH/03 Added additional dnslists conditions == and =& which are different from
902 = and & when the dns lookup returns more than one IP address.
904 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
905 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
907 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
908 FSYNC, which compiles an option called disable_fsync that allows for
909 bypassing fsync(). The documentation is heavily laced with warnings.
911 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
913 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
914 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
915 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
916 including adding "make clean"; (3) Added -fPIC when compiling the test
917 dynamically loaded module, to get rid of a warning.
919 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
920 message fails, move_frozen_messages = true and ignore_bounce_errors_after
921 = 0s. The bug is otherwise harmless.
923 PH/07 There was a bug in the dovecot authenticator such that the value of
924 $auth1 could be overwritten, and so not correctly preserved, after a
925 successful authentication. This usually meant that the value preserved by
926 the server_setid option was incorrect.
928 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
930 PH/09 Installed PCRE release 7.0.
932 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
933 run for batched SMTP input. It is now run at the start of every message
934 in the batch. While fixing this I discovered that the process information
935 (output by running exiwhat) was not always getting set for -bs and -bS
936 input. This is fixed, and it now also says "batched" for BSMTP.
938 PH/11 Added control=no_pipelining.
940 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
941 patch, slightly modified), and move the expansion of helo_data till after
942 the connection is made in the smtp transport (so it can use these
945 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
947 PH/14 Added log_selector = +pid.
949 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
951 PH/16 Add ${if forany and ${if forall.
953 PH/17 Added dsn_from option to vary the From: line in DSNs.
955 PH/18 Flush SMTP output before performing a callout, unless control =
956 no_callout_flush is set.
958 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
959 was true (the default) a successful delivery failed to delete the retry
960 item, thus causing premature timeout of the address. The bug is now
963 PH/20 Added hosts_avoid_pipelining to the smtp transport.
965 PH/21 Long custom messages for fakedefer and fakereject are now split up
966 into multiline reponses in the same way that messages for "deny" and
967 other ACL rejections are.
969 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
970 with slight modification.
972 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
973 draft, changing the syntax and factoring some duplicate code".
975 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
976 for deliveries of the second and subsequent messages over the same SMTP
979 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
980 ${reduce, with only minor "tidies".
982 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
984 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
985 expansion side effects.
987 PH/27 When a message times out after an over-quota error from an Exim-imposed
988 quota, the bounce message says "mailbox is full". This message was not
989 being given when it was a system quota that was exceeded. It now should
992 MH/03 Made $recipients available in local_scan(). local_scan() already has
993 better access to the recipient list through recipients_list[], but
994 $recipients can be useful in postmaster-provided expansion strings.
996 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
997 in the case of a MAIL command with additional options following the
998 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
999 were accidentally chopped off.
1001 PH/29 SMTP synchronization checks are implemented when a command is read -
1002 there is a check that no more input is waiting when there shouldn't be
1003 any. However, for some commands, a delay in an ACL can mean that it is
1004 some time before the response is written. In this time, more input might
1005 arrive, invalidly. So now there are extra checks after an ACL has run for
1006 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1007 pipelining has not been advertised.
1009 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1011 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1012 returned in the SMTP response when smtp_return_error_details was false.
1013 This has been fixed.
1015 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1016 instead of the C I/O that was originally supplied, because problems were
1017 reported on Solaris.
1019 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1020 Exim which did not show up earlier: it was assuming that a call to
1021 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1022 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1023 was a macro that became an assignment, so it seemed to work. This has
1024 changed to a proper function call with a void return, hence the compile
1025 error. Exim's code has been fixed.
1027 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1030 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1032 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1034 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1035 because it turns out to be too big or there is a timeout) there is no
1036 "Completed" line in the log. When some messages of this type were
1037 selected by exigrep, they were listed as "not completed". Others were
1038 picked up by some special patterns. I have improved the selection
1039 criteria to be more general.
1041 PH/38 The host_find_failed option in the manualroute router can now be set
1042 to "ignore", to completely ignore a host whose IP address cannot be
1043 found. If all hosts are ignored, the behaviour is controlled by the new
1044 host_all_ignored option.
1046 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1047 homing or because of multiple MX records with /mx) generated more than
1048 one IP address, and the following item turned out to be the local host,
1049 all the secondary addresses of the first item were incorrectly removed
1050 from the list, along with the local host and any following hosts (which
1051 is what is supposed to happen).
1053 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1054 whoever called Exim into the -H file. In the case of the daemon it was
1055 behaving confusingly. When first started, it used values for whoever
1056 started the daemon, but after a SIGHUP it used the Exim user (because it
1057 calls itself on a restart). I have changed the code so that it now always
1060 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1061 message are rejected with the same error (e.g. no authentication or bad
1062 sender address), and a DATA command is nevertheless sent (as can happen
1063 with PIPELINING or a stupid MUA), the error message that was given to the
1064 RCPT commands is included in the rejection of the DATA command. This is
1065 intended to be helpful for MUAs that show only the final error to their
1068 PH/42 Another patch from the Sieve maintainer.
1070 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1071 Eximstats - Fixed some broken HTML links and added missing column headers
1073 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1074 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1076 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1078 PH/43 Yet another patch from the Sieve maintainer.
1080 PH/44 I found a way to check for a TCP/IP connection going away before sending
1081 the response to the final '.' that terminates a message, but only in the
1082 case where the client has not sent further data following the '.'
1083 (unfortunately, this is allowed). However, in many cases there won't be
1084 any further data because there won't be any more messages to send. A call
1085 to select() can be used: if it shows that the input is "ready", there is
1086 either input waiting, or the socket has been closed. An attempt to read
1087 the next input character can distinguish the two cases. Previously, Exim
1088 would have sent an OK response which the client would never have see.
1089 This could lead to message repetition. This fix should cure that, at
1090 least in a lot of common cases.
1092 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1093 advertised in response to EHLO.
1099 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1100 fixed by 4.65/MH/01 (is this a record?) are fixed:
1102 (i) An empty string was always treated as zero by the numeric comparison
1103 operators. This behaviour has been restored.
1105 (ii) It is documented that the numeric comparison operators always treat
1106 their arguments as decimal numbers. This was broken in that numbers
1107 starting with 0 were being interpreted as octal.
1109 While fixing these problems I realized that there was another issue that
1110 hadn't been noticed. Values of message_size_limit (both the global option
1111 and the transport option) were treated as octal if they started with 0.
1112 The documentation was vague. These values are now always treated as
1113 decimal, and I will make that clear in the documentation.
1119 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1120 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1123 MH/01 Don't check that the operands of numeric comparison operators are
1124 integers when their expansion is in "skipping" mode (fixes bug
1125 introduced by 4.64-PH/07).
1127 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1128 child addresses, Exim now panics and dies. Previously, because the count
1129 is held in a short int, deliveries were likely to be lost. As such a
1130 large number of recipients for a single message is ridiculous
1131 (performance will be very, very poor), I have chosen to impose a limit
1132 rather than extend the field.
1138 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1139 leftover -K file (the existence of which was triggered by #402).
1140 While we were at it, introduced process PID as part of the -K
1141 filename. This should rule out race conditions when creating
1144 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1145 processing considerably. Previous code took too long for large mails,
1146 triggering a timeout which in turn triggers #401.
1148 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1149 in the DK code in transports.c. sendfile() is not really portable,
1150 hence the _LINUX specificness.
1152 TF/01 In the add_headers option to the mail command in an Exim filter,
1153 there was a bug that Exim would claim a syntax error in any
1154 header after the first one which had an odd number of characters
1157 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1158 callout verification, Exim cached a "reject" for the entire domain. This
1159 is correct for most verifications, but it is not correct for a recipient
1160 verification with use_sender or use_postmaster set, because in that case
1161 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1162 case of MAIL FROM:<> rejection from other early rejections (e.g.
1163 rejection of HELO). When verifying a recipient using a non-null MAIL
1164 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1165 Whatever the result of the callout, the value of the domain cache is
1166 left unchanged (for any other kind of callout, getting as far as trying
1167 RCPT means that the domain itself is ok).
1169 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1172 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1173 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1174 session, the daemon ignores ECONNECT errors and logs others; it now
1175 ignores EPIPE as well.
1177 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1178 (quoted-printable decoding).
1180 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1181 later the small subsequent patch to fix an introduced bug.
1183 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1185 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1187 PH/08 An error is now given if message_size_limit is specified negative.
1189 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1190 to be given (somewhat) arbitrary names.
1192 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1195 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1196 miscellaneous code fixes
1198 PH/10 Added the log_reject_target ACL modifier to specify where to log
1201 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1202 hostname. This is wrong, because it relates to the incoming message (and
1203 probably the interface on which it is arriving) and not to the outgoing
1204 callout (which could be using a different interface). This has been
1205 changed to use the value of the helo_data option from the smtp transport
1206 instead - this is what is used when a message is actually being sent. If
1207 there is no remote transport (possible with a router that sets up host
1208 addresses), $smtp_active_hostname is used.
1210 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1211 tweaks were necessary in order to get it to work (see also 21 below):
1212 (a) The code assumed that strncpy() returns a negative number on buffer
1213 overflow, which isn't the case. Replaced with Exim's string_format()
1215 (b) There were several signed/unsigned issues. I just did the minimum
1216 hacking in of casts. There is scope for a larger refactoring.
1217 (c) The code used strcasecmp() which is not a standard C function.
1218 Replaced with Exim's strcmpic() function.
1219 (d) The code set only $1; it now sets $auth1 as well.
1220 (e) A simple test gave the error "authentication client didn't specify
1221 service in request". It would seem that Dovecot has changed its
1222 interface. Fortunately there's a specification; I followed it and
1223 changed what the client sends and it appears to be working now.
1225 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1228 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1229 address A is aliased to B and C, where B exists and C does not. Without
1230 -v the output is "A verified" because verification stops after a
1231 successful redirection if more than one address is generated. However,
1232 with -v the child addresses are also verified. Exim was outputting "A
1233 failed to verify" and then showing the successful verification for C,
1234 with its parentage. It now outputs "B failed to verify", showing B's
1235 parentage before showing the successful verification of C.
1237 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1238 look up a TXT record in a specific list after matching in a combined
1241 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1242 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1243 they consult the DNS. I had assumed they would set it the way they
1244 wanted; and indeed my experiments on Linux seem to show that in some
1245 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1246 To be on the safe side, however, I have now made the interface to
1247 host_find_byname() similar to host_find_bydns(), with an argument
1248 containing the DNS resolver options. The host_find_byname() function now
1249 sets these options at its start, just as host_find_bydns() does. The smtp
1250 transport options dns_qualify_single and dns_search_parents are passed to
1251 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1252 of host_find_byname() use the default settings of RES_DEFNAMES
1253 (qualify_single) but not RES_DNSRCH (search_parents).
1255 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1256 spool_read_header() do less string testing, by means of a preliminary
1257 switch on the second character of optional "-foo" lines. (This is
1258 overdue, caused by the large number of possibilities that now exist.
1259 Originally there were few.) While I was there, I also converted the
1260 str(n)cmp tests so they don't re-test the leading "-" and the first
1261 character, in the hope this might squeeze out yet more improvement.
1263 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1264 flag allowing group syntax was set by the header_syntax check but not
1265 turned off, possible causing trouble later; (2) The flag was not being
1266 set at all for the header_verify test, causing "group"-style headers to
1267 be rejected. I have now set it in this case, and also caused header_
1268 verify to ignore an empty address taken from a group. While doing this, I
1269 came across some other cases where the code for allowing group syntax
1270 while scanning a header line wasn't quite right (mostly, not resetting
1271 the flag correctly in the right place). These bugs could have caused
1272 trouble for malformed header lines. I hope it is now all correct.
1274 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1275 with the "reply" argument non-NULL. The code, however (which originally
1276 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1277 but it didn't always do it. This confused somebody who was copying the
1278 code for some other use. I have removed all the tests.
1280 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1281 feature that was used to support insecure browsers during the U.S. crypto
1282 embargo. It requires special client support, and Exim is probably the
1283 only MTA that supported it -- and would never use it because real RSA is
1284 always available. This code has been removed, because it had the bad
1285 effect of slowing Exim down by computing (never used) parameters for the
1286 RSA_EXPORT functionality.
1288 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1289 authenticator to fail if there's a tab character in the incoming data
1290 (there should never be unless someone is messing about, as it's supposed
1291 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1292 if the connection is using TLS or if the remote IP is the same as the
1293 local IP, and the "valid-client-cert option" if a client certificate has
1296 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1297 authenticators. This can be used for authorization after authentication
1298 succeeds. (In the case of plaintext, it servers for both authentication
1301 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1302 if any retry times were supplied.
1304 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1305 connection, where there is no client IP address to check. In this
1306 situation, the verify now always succeeds.
1308 PH/25 Applied John Jetmore's -Mset patch.
1310 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1312 PH/27 In a string expansion for a processed (not raw) header when multiple
1313 headers of the same name were present, leading whitespace was being
1314 removed from all of them, but trailing whitespace was being removed only
1315 from the last one. Now trailing whitespace is removed from each header
1316 before concatenation. Completely empty headers in a concatenation (as
1317 before) are ignored.
1319 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1320 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1322 PH/29 [Removed. This was a change that I later backed out, and forgot to
1323 correct the ChangeLog entry (that I had efficiently created) before
1324 committing the later change.]
1326 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1327 address errors (4xx response to RCPT) over the same connection as other
1328 messages routed to the same hosts. Such deliveries are always "forced",
1329 so retry times are not inspected. This resulted in far too many retries
1330 for the affected addresses. The effect occurred only when there were more
1331 hosts than the hosts_max_try setting in the smtp transport when it had
1332 the 4xx errors. Those hosts that it had tried were not added to the list
1333 of hosts for which the message was waiting, so if all were tried, there
1334 was no problem. Two fixes have been applied:
1336 (i) If there are any address or message errors in an SMTP delivery, none
1337 of the hosts (tried or untried) are now added to the list of hosts
1338 for which the message is waiting, so the message should not be a
1339 candidate for sending over the same connection that was used for a
1340 successful delivery of some other message. This seems entirely
1341 reasonable: after all the message is NOT "waiting for some host".
1342 This is so "obvious" that I'm not sure why it wasn't done
1343 previously. Hope I haven't missed anything, but it can't do any
1344 harm, as the worst effect is to miss an optimization.
1346 (ii) If, despite (i), such a delivery is accidentally attempted, the
1347 routing retry time is respected, so at least it doesn't keep
1348 hammering the server.
1350 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1351 in ${readsocket because some servers need this prod.
1353 PH/32 Added some extra debug output when updating a wait-xxx database.
1355 PH/33 The hint "could be header name not terminated by colon", which has been
1356 given for certain expansion errors for a long time, was not being given
1357 for the ${if def:h_colon_omitted{... case.
1359 PH/34 The spec says: "With one important exception, whenever a domain list is
1360 being scanned, $domain contains the subject domain." There was at least
1361 one case where this was not true.
1363 PH/35 The error "getsockname() failed: connection reset by peer" was being
1364 written to the panic log as well as the main log, but it isn't really
1365 panic-worthy as it just means the connection died rather early on. I have
1366 removed the panic log writing for the ECONNRESET error when getsockname()
1369 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1370 runs only) independently of the message's sender address. This meant
1371 that, if the 4xx error was in fact related to the sender, a different
1372 message to the same recipient with a different sender could confuse
1373 things. In particualar, this can happen when sending to a greylisting
1374 server, but other circumstances could also provoke similar problems.
1375 I have changed the default so that the retry time for these errors is now
1376 based a combination of the sender and recipient addresses. This change
1377 can be overridden by setting address_retry_include_sender=false in the
1380 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1381 remote server are returned as part of bounce messages. This was not
1382 happening for LMTP over a pipe (the lmtp transport), but now it is the
1383 same for both kinds of LMTP.
1385 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1386 in header lines that were in fact CNAMEs. This is no longer the case.
1388 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1389 and queue runs started by the daemon processed all messages. This has
1390 been fixed so that -R and -S can now usefully be given with -q<time>.
1392 PH/40 Import PCRE release 6.7 (fixes some bugs).
1394 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1396 PH/42 Give an error if -q is specified more than once.
1398 PH/43 Renamed the variables $interface_address and $interface_port as
1399 $received_ip_address and $received_port, to make it clear that these
1400 values apply to message reception, and not to the outgoing interface when
1401 a message is delivered. (The old names remain recognized, of course.)
1403 PH/44 There was no timeout on the connect() call when using a Unix domain
1404 socket in the ${readsocket expansion. There now is.
1406 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1407 be meaningful with "accept".
1409 SC/01 Eximstats V1.43
1410 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1412 SC/02 Eximstats V1.44
1413 Use a glob alias rather than an array ref in the generated
1414 parser. This improves both readability and performance.
1416 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1417 Collect SpamAssassin and rejection statistics.
1418 Don't display local sender or destination tables unless
1419 there is data to show.
1420 Added average volumes into the top table text output.
1422 SC/04 Eximstats V1.46
1423 Collect data on the number of addresses (recipients)
1424 as well as the number of messages.
1426 SC/05 Eximstats V1.47
1427 Added 'Message too big' to the list of mail rejection
1428 reasons (thanks to Marco Gaiarin).
1430 SC/06 Eximstats V1.48
1431 Mainlog lines which have GMT offsets and are too short to
1432 have a flag are now skipped.
1434 SC/07 Eximstats V1.49 (Alain Williams)
1435 Added the -emptyok flag.
1437 SC/08 Eximstats V1.50
1438 Fixes for obtaining the IP address from reject messages.
1440 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1441 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1442 whitesspace changes from 4.64-PH/27
1444 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1447 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1448 are found, allow negative numbers in numeric criteria)
1450 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1452 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1453 to match changes made in 4.64-PH/43
1455 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1457 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1458 "domain" is actually an IP address, and force a failure. This locks out
1459 those revolvers/nameservers that support "A-for-A" lookups, in
1460 contravention of the specifications.
1462 PH/48 When a host name was looked up from an IP address, and the subsequent
1463 forward lookup of the name timed out, the host name was left in
1464 $sender_host_name, contrary to the specification.
1466 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1467 restricted to single-key lookups, Exim was not diagnosing an error if
1468 * or *@ was used with a query-style lookup.
1470 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1472 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1473 long ago, but noone interested enough thought of it. Let's just say that
1474 the "1.1" means that there are some new functions that weren't there at
1475 some point in the past.
1477 PH/51 Error processing for expansion failure of helo_data from an smtp
1478 transport during callout processing was broken.
1480 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1481 tested/used via the -bh/-bhc/-bs options.
1483 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1484 bug, fixed in subsequent PCRE releases).
1486 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1487 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1489 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1495 SC/01 Use a glob alias rather than an array ref in eximstats generated
1496 parser. This improves both readability and performance.
1498 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1499 Don't display local sender or destination tables in eximstats unless
1500 there is data to show.
1501 Added average volumes into the eximstats top table text output.
1503 SC/03 Collect data on the number of addresses (recipients) as well
1504 as the number of messages in eximstats.
1506 TF/01 Correct an error in the documentation for the redirect router. Exim
1507 does (usually) call initgroups() when daemonizing.
1509 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1510 with consistent privilege compared to when running as a daemon.
1512 TF/03 Note in the spec that $authenticated_id is not set for local
1513 submissions from trusted users.
1515 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1516 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1518 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1519 by adding some example configuration directives to the default
1520 configuration file. A little bit of work is required to uncomment the
1521 directives and define how usernames and passwords are checked, but
1522 there is now a framework to start from.
1524 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1525 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1526 without this. I don't know how relevant this is to other LDAP libraries.
1528 PH/02 Add the verb name to the "unknown ACL verb" error.
1530 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1532 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1534 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1535 directory not expanded when it should be if an expanded home directory
1536 was set for the address (which is overridden by the transport).
1538 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1541 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1542 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1543 because it is too late at that time, and has no effect.
1545 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1546 security issue with \' (bugzilla #107). I could not use the
1547 PQescapeStringConn() function, because it needs a PGconn value as one of
1550 PH/08 When testing addresses using -bt, indicate those final addresses that
1551 are duplicates that would not cause an additional delivery. At least one
1552 person was confused, thinking that -bt output corresponded to deliveries.
1553 (Suppressing duplicates isn't a good idea as you lose the information
1554 about possibly different redirections that led to the duplicates.)
1556 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1557 systems where poll() doesn't work, in particular OS X.
1559 PH/10 Added more information to debugging output for retry time not reached.
1561 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1562 operations in malware.c.
1564 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1567 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1568 syslog_duplication set false, log lines that would normally be written
1569 both the the main log and to the reject log were not written to syslog at
1572 PH/14 In the default configuration, change the use of "message" in ACL warn
1573 statements to "add_header".
1575 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1576 not followed by a command (e.g. "seen endif").
1578 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1579 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1582 PH/17 Added extra conditions to the default value of delay_warning_condition
1586 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1587 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1588 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1591 The Auto-Submitted: and various List- headers are standardised, whereas I
1592 don't think Precedence: ever was.
1594 PH/18 Refactored debugging code in route_finduser() to show more information,
1595 in particular, the error code if getpwnam() issues one.
1597 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1598 This is apparently needed in addition to the PH/07 change above to avoid
1599 any possible encoding problems.
1601 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1602 but not after initializing Perl.
1604 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1605 output them only if debugging. By default they are written stderr,
1606 apparently, which is not desirable.
1608 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1611 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1614 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1616 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1617 authenticated or an ident call has been made. Suppress the default
1618 values for $authenticated_id and $authenticated_sender (but permit -oMai
1619 and -oMas) when testing with -bh.
1621 PH/24 Re-jigged the order of the tests in the default configuration so that the
1622 tests for valid domains and recipients precede the DNS black list and CSA
1623 tests, on the grounds that those ones are more expensive.
1625 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1626 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1627 This bug exists in every version of Exim that I still have, right back to
1630 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1631 However, an attempt to turn on case-sensitivity in a regex key by
1632 including (?-i) didn't work because the subject string was already
1633 lowercased, and the effects were non-intuitive. It turns out that a
1634 one-line patch can be used to allow (?-i) to work as expected.
1640 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1641 other effects) broke the use of negated acl sub-conditions.
1643 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1646 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1647 "Deny" causes Exim to reject the incoming connection with a 554 error.
1648 Unfortunately, if there is a major crisis, such as a disk failure,
1649 tcp-wrappers gives "deny", whereas what one would like would be some
1650 kind of temporary error. A kludge has been added to help with this.
1651 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1652 554 error is used if errno is still zero or contains ENOENT (which occurs
1653 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1656 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1658 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1659 errors. Otherwise a message that provokes a temporary error (when other
1660 messages do not) can cause a whole host to time out.
1662 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1663 addresses were routed directly to files or pipes from a redirect router.
1664 File deliveries just didn't batch; pipe deliveries might have suffered
1667 PH/06 A failure to get a lock for a hints database would erroneously always say
1668 "Failed to get write lock", even when it was really a read lock.
1670 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1671 of 0600. This has been changed to use the value of the lockfile_mode
1672 option (which defaults to 0600).
1674 PH/08 Applied small patch from the Sieve maintainer.
1676 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1677 folder from quota calculations, a direct delivery into this folder messed
1678 up the contents of the maildirsize file. This was because the regex was
1679 used only to exclude .Trash (or whatever) when the size of the mailbox
1680 was calculated. There was no check that a delivery was happening into an
1681 excluded directory. This bug has been fixed by ignoring all quota
1682 processing for deliveries into excluded directories.
1684 PH/10 Added the maildirfolder_create_regex option to appendfile.
1690 PH/01 The code for finding all the local interface addresses on a FreeBSD
1691 system running IPv6 was broken. This may well have applied to all BSD
1692 systems, as well as to others that have similar system calls. The broken
1693 code found IPv4 interfaces correctly, but gave incorrect values for the
1694 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1695 that it would not match correctly against @[] and not recognize the IPv6
1698 PH/02 The ipliteral router was not recognizing addresses of the form user@
1699 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1701 PH/03 Added disable_ipv6.
1703 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1704 decoded form, because it is most often used to construct To: headers
1705 lines in autoreplies, and the decoded form may well be syntactically
1706 invalid. However, $reply_address has leading white space removed, and all
1707 newlines turned into spaces so that the autoreply transport does not
1710 PH/05 If group was specified without a user on a router, and no group or user
1711 was specified on a transport, the group from the router was ignored.
1713 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1714 for visible compile-time settings that can be used to change these
1715 numbers, for those that want even more. Backwards compatibility with old
1716 spool files has been maintained. However, going back to a previous Exim
1717 release will lost any variables that are in spool files.
1719 PH/07 Two small changes when running in the test harness: increase delay when
1720 passing a TCP/IP connection to a new process, in case the original
1721 process has to generate a bounce, and remove special handling of
1722 127.0.0.2 (sic), which is no longer necessary.
1724 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1725 be the same on different OS.
1727 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1730 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1731 whether --show-vars was specified or not
1733 JJ/02 exipick: Added support for new ACL variable spool format introduced
1736 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1737 syntactically invalid From: or Reply-to: line, and a filter used this to
1738 generate an autoreply, and therefore failed to obtain an address for the
1739 autoreply, Exim could try to deliver to a non-existent relative file
1740 name, causing unrelated and misleading errors. What now happens is that
1741 it logs this as a hard delivery error, but does not attempt to create a
1744 PH/11 The exinext utility has a -C option for testing purposes, but although
1745 the given file was scanned by exinext itself; it wasn't being passed on
1746 when Exim was called.
1748 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1749 an end-of-file indication when reading a command response.
1751 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1752 compiled. In many other places in Exim, IPv6 addresses are always
1753 recognized, so I have changed this. It also means that IPv4 domain
1754 literals of the form [IPV4:n.n.n.n] are now always recognized.
1756 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1757 used if the router is not running as root, for example, when verifying at
1758 ACL time, or when using -bh. The debugging output from this situation was
1759 non-existent - all you got was a failure to exec. I have made two
1762 (a) Failures to set uid/gid, the current directory, or a process leader
1763 in a subprocess such as that created by queryprogram now generate
1764 suitable debugging ouput when -d is set.
1766 (b) The queryprogram router detects when it is not running as root,
1767 outputs suitable debugging information if -d is set, and then runs
1768 the subprocess without attempting to change uid/gid.
1770 PH/15 Minor change to Makefile for building test_host (undocumented testing
1773 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1774 additional section of a DNS packet that returns MX or SRV records.
1775 Instead, it always explicitly searches for A/AAAA records. This avoids
1776 major problems that occur when a DNS server includes only records of one
1777 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1778 fixed another bug: if SRV records were looked up and the corresponding
1779 address records were *not* found in the additional section, the port
1780 values from the SRV records were lost.
1782 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1783 using the correct key (the original address) when searching the retry
1784 rules in order to find which one to use for generating the retry hint.
1786 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1787 adding the default one. Similarly, if it contains a Reply-To: header, the
1788 errors_reply_to option, if set, is not used.
1790 PH/19 When calculating a retry time, Exim used to measure the "time since
1791 failure" by looking at the "first failed" field in the retry record. Now
1792 it does not use this if it is later than than the arrival time of the
1793 message. Instead it uses the arrival time. This makes for better
1794 behaviour in cases where some deliveries succeed, thus re-setting the
1795 "first failed" field. An example is a quota failure for a huge message
1796 when small messages continue to be delivered. Without this change, the
1797 "time since failure" will always be short, possible causing more frequent
1798 delivery attempts for the huge message than are intended.
1799 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1801 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1802 $1, $2, $3) because the numerical variables can be reset during some
1803 expansion items (e.g. "match"), thereby losing the authentication data.
1805 PH/21 Make -bV show the size of off_t variables so that the test suite can
1806 decide whether to run tests for quotas > 2G.
1808 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1809 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1810 filecount value is greater than 2G or if a quota value is greater than 2G
1811 on a system where the size of off_t is not greater than 4, a panic error
1814 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1815 never match. The debug and -bh output now contains an explicit error
1816 message indicating a malformed IPv4 address or mask.
1818 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1819 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1820 PH/23 above applies.
1822 PH/25 Do not write to syslog when running in the test harness. The only
1823 occasion when this arises is a failure to open the main or panic logs
1824 (for which there is an explicit test).
1826 PH/26 Added the /no_tell option to "control=freeze".
1828 PH/27 If a host name lookup failed very early in a connection, for example, if
1829 the IP address matched host_lookup and the reverse lookup yielded a name
1830 that did not have a forward lookup, an error message of the form "no IP
1831 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1832 could be logged. Now it outputs the IP address instead of "NULL".
1834 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1835 allows the sender and the authenticated sender to be set when
1836 submitting a message from within Exim. Since child_open_exim() is
1837 documented for local_scan(), the new function should be too.
1839 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1840 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1841 results in an empty string is now treated as unset.
1843 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1845 PH/31 Added sender_verify_fail logging option.
1847 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1848 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1849 tidied the source and removed it altogether.
1851 PH/33 When a queue run was abandoned because the load average was too high, a
1852 log line was always written; now it is written only if the queue_run log
1853 selector is set. In addition, the log line for abandonment now contains
1854 information about the queue run such as the pid. This is always present
1855 in "start" and "stop" lines but was omitted from the "abandon" line.
1857 PH/34 Omit spaces between a header name and the colon in the error message that
1858 is given when verify = headers_syntax fails (if there are lots of them,
1859 the message gets confusing).
1861 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1862 names, as there are now some PTR records that contain slashes. This check
1863 is only to protect against broken name servers that fall over on strange
1864 characters, so the fact that it applies to all lookups doesn't matter.
1866 PH/36 Now that the new test suite is complete, we can remove some of the
1867 special code in Exim that was needed for the old test suite. For example,
1868 sorting DNS records because real resolvers return them in an arbitrary
1869 order. The new test suite's fake resolver always returns records in the
1872 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1873 bounces) except when queue_only is set, to avoid logging races between
1874 the different processes.
1876 PH/38 Panic-die if .include specifies a non-absolute path.
1878 PH/39 A tweak to the "H" retry rule from its user.
1880 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1881 a label. They prevented compilation on older perls.
1883 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1884 a warning to be raised on newish perls.
1886 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1887 on queue. Changes to match documented behaviour of showing count of
1888 messages matching specified criteria.
1890 PH/40 Changed the default ident timeout from 30s to 5s.
1892 PH/41 Added support for the use of login_cap features, on those BSD systems
1893 that have them, for controlling the resources used by pipe deliveries.
1895 PH/42 The content-scanning code uses fopen() to create files in which to put
1896 message data. Previously it was not paying any attention to the mode of
1897 the files. Exim runs with umask(0) because the rest of the code creates
1898 files with open(), and sets the required mode explicitly. Thus, these
1899 files were ending up world-writeable. This was not a big issue, because,
1900 being within the spool directory, they were not world-accessible. I have
1901 created a function called modefopen, which takes an additional mode
1902 argument. It sets umask(777), creates the file, chmods it to the required
1903 mode, then resets the umask. All the relevant calls to fopen() in the
1904 content scanning code have been changed to use this function.
1906 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1907 to 24 hours. This avoids potential overflow problems when processing G
1908 and H retry rules. I suspect nobody ever tinkers with this value.
1910 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1912 PH/45 When the plaintext authenticator is running as a client, the server's
1913 challenges are checked to ensure they are valid base64 strings. By
1914 default, the authentication attempt is cancelled if an invalid string is
1915 received. Setting client_ignore_invalid_base64 true ignores these errors.
1916 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1917 they are received. Thus, the responses can be made to depend on the
1918 challenges. If an invalid string is ignored, an empty string is placed in
1921 PH/46 Messages that are created by the autoreply transport now contains a
1922 References: header, in accordance with RFCs 2822 and 3834.
1924 PH/47 Added authenticated_sender_force to the smtp transport.
1926 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1928 PH/49 Installed latest patch from the Sieve maintainer.
1930 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1931 was also set, the appendfile transport was unnecessarily scanning a
1932 directory of message files (e.g. for maildir delivery) to find the count
1933 of files (along with the size), even though it did not need this
1934 information. It now does the scan only if it needs to find either the
1935 size of the count of files.
1937 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1939 PH/52 Two bugs concerned with error handling when the smtp transport is
1942 (i) Exim was not creating retry information for temporary errors given
1943 for individual recipients after the DATA command when the smtp transport
1944 was used in LMTP mode. This meant that they could be retried too
1945 frequently, and not timed out correctly.
1947 (ii) Exim was setting the flag that allows error details to be returned
1948 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1949 recipients that were returned after the DATA command.
1951 PH/53 This is related to PH/52, but is more general: for any failing address,
1952 when detailed error information was permitted to be returned to the
1953 sender, but the error was temporary, then after the final timeout, only
1954 "retry timeout exceeded" was returned. Now it returns the full error as
1955 well as "retry timeout exceeded".
1957 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1958 do this, and (what is worse) MTAs that accept it.
1960 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1961 will now be deprecated.
1963 PH/56 New os.c-cygwin from the Cygwin maintainer.
1965 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1966 formats (previously only available in exim formats via -bpr, -bpru,
1967 and -bpra. Now also available in native and exiqgrep formats)
1969 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1970 with very large, slow to parse queues
1972 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1974 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1976 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1977 responses to authentication challenges, though it was showing the
1978 challenges; (ii) I've removed the CR characters from the debug output for
1981 PH/58 Allow for the insertion of a newline as well as a space when a string
1982 is turned into more than one encoded-word during RFC 2047 encoding. The
1983 Sieve code now uses this.
1985 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1986 data_4xx, lost_connection, tls_required.
1988 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1989 message was being sent as an SMTP response.
1991 PH/61 Add -l and -k options to exicyclog.
1993 PH/62 When verifying, if an address was redirected to one new address, so that
1994 verification continued, and the new address failed or deferred after
1995 having set something in $address_data, the value of $address_data was not
1996 passed back to the ACL. This was different to the case when no
1997 redirection occurred. The value is now passed back in both cases.
1999 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2000 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2001 use login_cap.h, so on its own it isn't the distinguishing feature. The
2002 new name refers directly to the setclassresources() function.
2004 PH/65 Added configuration files for NetBSD3.
2006 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2008 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2009 is preferred over IPv4.
2011 PH/68 The bounce_return_message and bounce_return_body options were not being
2012 honoured for bounces generated during the reception of non-SMTP messages.
2013 In particular, this applied to messages rejected by the ACL. This bug has
2014 been fixed. However, if bounce_return_message is true and bounce_return_
2015 body is false, the headers that are returned for a non-SMTP message
2016 include only those that have been read before the error was detected.
2017 (In the case of an ACL rejection, they have all been read.)
2019 PH/69 The HTML version of the specification is now built in a directory called
2020 spec_html instead of spec.html, because the latter looks like a path with
2021 a MIME-type, and this confuses some software.
2023 PH/70 Catch two compiler warnings in sieve.c.
2025 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2026 function verify_get_ident() calls ip_connect() to connect a socket, but
2027 if the "connect()" function timed out, ip_connect() used to close the
2028 socket. However, verify_get_ident() also closes the socket later, and in
2029 between Exim writes to the log, which may get opened at this point. When
2030 the socket was closed in ip_connect(), the log could get the same file
2031 descriptor number as the socket. This naturally causes chaos. The fix is
2032 not to close the socket in ip_connect(); the socket should be closed by
2033 the function that creates it. There was only one place in the code where
2034 this was missing, in the iplookup router, which I don't think anybody now
2035 uses, but I've fixed it anyway.
2037 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2038 well as to direct DNS lookups. Otherwise the handling of names in host
2039 lists is inconsistent and therefore confusing.
2045 PH/01 Two changes to the default runtime configuration:
2047 (1) Move the checks for relay_from_hosts and authenticated clients from
2048 after to before the (commented out) DNS black list checks.
2050 (2) Add control=submission to the relay_from_hosts and authenticated
2051 clients checks, on the grounds that messages accepted by these
2052 statements are most likely to be submissions.
2054 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2056 (1) Generate an error if the third argument for the ${prvs expansion is
2059 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2062 (3) Reset the variables that are obtained from the first argument of
2063 ${prvscheck and used in the second argument before leaving the code,
2064 because their memory is reclaimed, so using them afterwards may do
2067 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2068 one (it's much easier than Tom thought :-).
2070 (5) Because of (4), we can now allow for the use of $prvscheck_result
2071 inside the third argument.
2073 PH/03 For some reason, the default setting of PATH when running a command from
2074 a pipe transport was just "/usr/bin". I have changed it to
2077 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2078 anything to be listed in the output from -bV.
2080 PH/05 When a filter generated an autoreply, the entire To: header line was
2081 quoted in the delivery log line, like this:
2083 => >A.N.Other <ano@some.domain> <original@ddress> ...
2085 This has been changed so that it extracts the operative address. There
2086 may be more than one such address. If so, they are comma-separated, like
2089 => >ano@some.domain,ona@other.domain <original@ddress> ...
2091 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2092 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2093 not being looked up in the rDNS to get a host name, Exim was showing the
2094 IP address twice in Received: lines, even though the IP addresses were
2095 identical. For example:
2097 Received: from [1.2.3.4] (helo=[1.2.3.4])
2099 However, if the real host name was known, it was omitting the HELO data
2100 if it matched the actual IP address. This has been tidied up so that it
2101 doesn't show the same IP address twice.
2103 PH/07 When both +timestamp and +memory debugging was on, the value given by
2104 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2105 called by the debug printing, thereby overwriting the timestamp buffer.
2106 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2108 PH/08 When the original message was included in an autoreply transport, it
2109 always said "this is a copy of the message, including all the headers",
2110 even if body_only or headers_only was set. It now gives an appropriate
2113 PH/09 Applied a patch from the Sieve maintainer which:
2115 o fixes some comments
2116 o adds the (disabled) notify extension core
2117 o adds some debug output for the result of if/elsif tests
2118 o points to the current vacation draft in the documentation
2119 and documents the missing references header update
2123 o fixes a bug in processing the envelope test (when testing
2124 multiple envelope elements, the last element determinted the
2127 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2128 Electronic Mail") by including:
2130 Auto-submitted: auto-generated
2132 in the messages that it generates (bounce messages and others, such as
2133 warnings). In the case of bounce messages for non-SMTP mesages, there was
2134 also a typo: it was using "Auto_submitted" (underscore instead of
2135 hyphen). Since every message generated by Exim is necessarily in response
2136 to another message, thes have all been changed to:
2138 Auto-Submitted: auto-replied
2140 in accordance with these statements in the RFC:
2142 The auto-replied keyword:
2144 - SHOULD be used on messages sent in direct response to another
2145 message by an automatic process,
2147 - MUST NOT be used on manually-generated messages,
2149 - MAY be used on Delivery Status Notifications (DSNs) and Message
2150 Disposition Notifications (MDNs),
2152 - MUST NOT be used on messages generated by automatic or periodic
2153 processes, except for messages which are automatic responses to
2156 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2157 to the default Received: header definition.
2159 PH/12 Added log selector acl_warn_skipped (default on).
2161 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2162 variables because (a) they are in the wrong storage pool and (b) even if
2163 they were copied, it wouldn't work properly because of the caching.
2165 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2166 checking when decoding. Apparently there are clients that generate
2167 overlong encoded strings. Why am I not surprised?
2169 PH/15 If the first argument of "${if match_address" was not empty, but did not
2170 contain an "@" character, Exim crashed. Now it writes a panic log message
2171 and treats the condition as false.
2173 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2175 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2176 extension "envelope-auth". The code is finished and in agreement with
2177 other implementations, but there is no documentation so far and in fact,
2178 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2179 not changing the active code.
2181 Print executed "if" and "elsif" statements when debugging is used. This
2182 helps a great deal to understand what a filter does.
2184 Document more things not specified clearly in RFC3028. I had all this
2185 sorted out, when out of a sudden new issues came to my mind. Oops."
2187 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2190 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2191 canonical form (as documented). However, after a host name lookup from
2192 the IP address, check_host() was doing a simple string comparison with
2193 addresses acquired from the DNS when checking that the found name did
2194 have the original IP as one of its addresses. Since any found IPv6
2195 addresses are likely to be in abbreviated form, the comparison could
2196 fail. Luckily, there already exists a function for doing the comparison
2197 by converting both addresses to binary, so now that is used instead of
2198 the text comparison.
2200 PH/20 There was another similar case to PH/19, when a complete host name was
2201 given in a host list; looking up its IP address could give an abbreviated
2202 form, whereas the current host's name might or might not be abbreviated.
2203 The same fix has been applied.
2209 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2210 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2213 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2214 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2216 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2218 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2219 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2220 Also, for "Auto-Submitted", treat anything other than "no" as
2221 non-personal, in accordance with RFC 3834. (Previously it treated
2222 anything starting "auto-" as non-personal.)
2224 TF/01 The control=submission/name=... option had a problem with syntax
2225 errors if the name included a slash character. The /name= option
2226 now slurps the rest of the string, so it can include any characters
2227 but it must come last in the list of options (after /sender_retain
2230 PH/05 Some modifications to the interface to the fake nameserver for the new
2238 TK/01 Added the "success_on_redirect" address verification option. See
2239 NewStuff for rationale and an example.
2241 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2243 PH/02 Patch to exigrep to allow it to work on syslog lines.
2245 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2246 fread() to skip over the body file's header line, because in Cygwin the
2247 header line is locked and is inaccessible.
2249 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2250 co-exist for some time) to make it clear that it is the Exim ID that is
2251 referenced, not the Message-ID: header line.
2253 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2254 string_format() function, because snprintf() does not exist on all
2257 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2258 ${stat: expansion item.
2260 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2261 protocol synchronization error", to keep the pedants happy.
2263 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2264 well as for IRIX systems, when gcc is being used. See the host.c source
2267 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2269 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2272 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2273 transport and to the smtp transport in LMTP mode.
2275 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2277 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2278 run a filter in a subprocess. This could lead to confusion in subsequent
2279 lookups in the parent process. There should also be a search_tidyup() at
2280 the end of the subprocess.
2282 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2283 only if the host matched helo_try_verify_hosts, which caused the
2284 verification to occur when the EHLO/HELO command was issued. The ACL just
2285 tested the remembered result. Now, if a previous verification attempt has
2286 not happened, "verify = helo" does it there and then.
2288 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2290 TK/03 Fix log output including CR from clamd.
2292 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2293 exist provoked a memory error which could cause a segfault.
2295 PH/15 Installed PCRE 6.2
2297 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2299 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2300 of the problem. Specifically, suggested +O2 rather than +O1 for the
2303 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2305 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2306 record, the port was not correctly logged when the outgoing_port log
2307 selector was set (it logged the transort's default port).
2309 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2310 fallback_hosts, and "hosts" in the smtp transport.
2312 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2313 host errors such as "Connection refused".
2315 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2316 authentication with radiusclient 0.4.9:
2318 - Error returned from rc_read_config was caught wrongly
2319 - Username/password not passed on to radius server due to wrong length.
2321 The presumption is that some radiusclient API changes for 4.51/PH/17
2322 were not taken care of correctly. The code is still untested by me (my
2323 Linux distribution still has 0.3.2 of radiusclient), but it was
2324 contributed by a Radius user.
2326 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2327 expanding the "port" option of the smtp transport.
2329 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2330 while reading a MIME header. Thanks to Tom Hughes for a patch.
2332 PH/24 Include config.h inside local_scan.h so that configuration settings are
2335 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2336 that in an ACL for RCPT (for example), you can examine exactly what was
2339 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2340 commands, but it was not correctly comparing the address with the actual
2341 client host address. Thus, it would show the EHLO address in Received:
2342 header lines when this was not necessary.
2344 PH/27 Added the % operator to ${eval:}.
2346 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2347 it should be ignoring failures (because with -C, for example, it has lost
2348 privilege). It wasn't ignoring creation failures other than "already
2351 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2354 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2355 because an input line was too long, either on its own, or by virtue of
2356 too many continuations, the temporary file was not being removed, and the
2357 return code was incorrect.
2359 PH/31 Missing "BOOL" in function definition in filtertest.c.
2361 PH/32 Applied Sieve patches from the maintainer.
2363 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2365 PH/33 Added "verify = not_blind".
2367 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2368 Local/Makefile (with some defaults set). These are used in built scripts
2369 such as exicyclog, but they have never been used in the exim_install
2370 script (though there are many overriding facilities there). I have
2371 arranged that the exim_install script now takes note of these two
2374 PH/35 Installed configuration files for Dragonfly.
2376 PH/36 When a locally submitted message by a trusted user did not contain a
2377 From: header, and the sender address was obtained from -f or from an SMTP
2378 MAIL command, and the trusted user did not use -F to supply a sender
2379 name, $originator_name was incorrectly used when constructing a From:
2380 header. Furthermore, $originator_name was used for submission mode
2381 messages from external hosts without From: headers in a similar way,
2382 which is clearly wrong.
2384 PH/37 Added control=suppress_local_fixups.
2386 PH/38 When log_selector = +received_sender was set, and the addition of the
2387 sender made the log line's construction buffer exactly full, or one byte
2388 less than full, an overflow happened when the terminating "\n" was
2391 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2392 when the result of a list match is failure because a DNS lookup failed.
2394 PH/40 RM_COMMAND is now used in the building process.
2396 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2397 the "build-* directories that it finds.
2399 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2400 address in a domain literal was a prefix of an interface address.
2402 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2403 when verifying a sender address, unless rewrite_headers is false.
2405 PH/44 Wrote a long comment about why errors_to addresses are verified as
2406 recipients, not senders.
2408 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2409 the ratelimit ACL was added.
2411 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2413 PH/46 Added extra information about PostgreSQL errors to the error string.
2415 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2416 avoiding the need to install special zones in a real server. This is
2417 backwards compatible; if it can't find the fake resolver, it drops back.
2418 Thus, both old and new test suites can be run.
2420 TF/02 Added util/ratelimit.pl
2422 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2423 clock is set back in time.
2425 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2426 Candler <B.Candler@pobox.com>.
2428 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2429 OK for addresses that are the result of redirections.
2431 PH/48 A number of further additions for the benefit of the new test suite,
2432 including a fake gethostbyname() that interfaces to the fake DNS resolver
2435 TF/06 The fix for widen_domains has also been applied to qualify_single and
2436 search_parents which are the other dnslookup options that can cause
2439 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2442 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2444 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2445 service" if it encountered an MX record with an empty target hostname.
2446 The message is now "an MX or SRV record indicated no SMTP service".
2448 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2449 if the DNS of the sending site is misconfigured. This is quite a
2450 common situation. This change restores the behaviour of treating a
2451 helo verification defer as a failure.
2453 PH/51 If self=fail was set on a router, the bounce message did not include the
2454 actual error message.
2460 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2462 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2463 command itself ended in error, the underlying message about the transport
2464 filter timeout was being overwritten with the pipe command error. Now the
2465 underlying error message should be appended to the second error message.
2467 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2469 PH/02 Reduce the amount of output that "make" produces by default. Full output
2470 can still be requested.
2472 PH/03 The warning log line about a condition test deferring for a "warn" verb
2473 was being output only once per connection, rather than after each
2474 occurrence (because it was using the same function as for successful
2475 "warn" verbs). This seems wrong, so I have changed it.
2477 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2478 it should not have, which might have caused a crash in the right
2479 circumstances, but probably never did.
2481 PH/04 Installed a modified version of Tony Finch's patch to make submission
2482 mode fix the return path as well as the Sender: header line, and to
2483 add a /name= option so that you can make the user's friendly name appear
2486 TF/03 Added the control = fakedefer ACL modifier.
2488 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2489 Mark Lowes for thorough testing.
2491 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2493 TK/03 Merged latest SRS patch from Miles Wilton.
2495 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2496 with the definition in sysexits.h (which is #included earlier).
2497 Fortunately, Exim does not actually use EX_OK. The code used to try to
2498 preserve the sysexits.h value, by assumimg that macro definitions were
2499 scanned for macro replacements. I have been disabused of this notion,
2500 so now the code just undefines EX_OK before #including unistd.h.
2502 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2503 in the smtp transport. When a block could not be written in a single
2504 write() function, the timeout was being re-applied to each part-write.
2505 This seems wrong - if the receiver was accepting one byte at a time it
2506 would take for ever. The timeout is now adjusted when this happens. It
2507 doesn't have to be particularly precise.
2509 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2510 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2512 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2513 without a domain if the check to <postmaster@domain> fails.
2515 SC/01 Eximstats: added -xls and the ability to specify output files
2516 (patch written by Frank Heydlauf).
2518 SC/02 Eximstats: use FileHandles for outputing results.
2520 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2522 SC/04 Eximstats: fixed display of large numbers with -nvr option
2524 SC/05 Eximstats: fixed merging of reports with empty tables.
2526 SC/06 Eximstats: added the -include_original_destination flag
2528 SC/07 Eximstats: removed tabs and trailing whitespace.
2530 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2532 TK/06 MBOX spool code: Add real "From " MBOX separator line
2533 so the .eml file is really in mbox format (even though
2534 most programs do not really care). Patch from Alex Miller.
2536 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2537 The latter is generated from $received_to and is only set if the
2538 message has one envelope recipient. SA can use these headers,
2539 obviously out-of-the-box. Patch from Alex Miller.
2541 PH/08 The ${def test on a variable was returning false if the variable's
2542 value was "0", contrary to what the specification has always said!
2543 The result should be true unless the variable is empty.
2545 PH/09 The syntax error of a character other than { following "${if
2546 def:variable_name" (after optional whitespace) was not being diagnosed.
2547 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2548 accidental colon was present, for example, could give incorrect results.
2550 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2551 result is used (not including appendfile, where other changes are about
2554 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2555 This involved changing a lot of size variables from int to off_t. It
2556 should work with maildirs and everything.
2558 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2559 spamd dying while we are connected to it.
2561 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2564 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2565 function for BDB 4.3.
2567 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2569 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2570 the sources for PCRE have been re-arranged and more files are now
2573 PH/15 The code I had for printing potentially long long variables in PH/11
2574 above was not the best (it lost precision). The length of off_t variables
2575 is now inspected at build time, and an appropriate printing format (%ld
2576 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2577 to be "long long int" or "long int". This is needed for the internal
2578 formatting function string_vformat().
2580 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2581 the configuration file to be ":syslog", then the script "guesses" where
2582 the logs files are, rather than using the compiled in default. In our
2583 case the guess is not the same as the compiled default, so the script
2584 suddenly stopped working when I started to use syslog. The patch checks
2585 to see if log_file_path is "". If so, it attempts to read it from exim
2586 with no configuration file to get the compiled in version, before it
2587 falls back to the previous guessing code."
2589 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2590 implementing BATV in an Exim configuration. See NewStuff for the gory
2593 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2594 Makefile that are specific to HP-UX.
2596 PH/18 If the "use_postmaster" option was set for a recipient callout together
2597 with the "random" option, the postmaster address was used as the MAIL
2598 FROM address for the random test, but not for the subsequent recipient
2599 test. It is now used for both.
2601 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2602 patch removes a few documentation additions to RFC 3028, because the
2603 latest draft now contains them. It adds the new en;ascii-case comparator
2604 and a new error check for 8bit text in MIME parts. Comparator and
2605 require names are now matched exactly. I enabled the subaddress
2606 extension, but it is not well tested yet (read: it works for me)."
2608 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2609 rework some of the code of TK/09 above to avoid the hardwired use of
2610 "%lld" and "long long". Replaced the call to snprintf() with a call to
2613 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2614 records point to non-existent hosts", "retry timeout exceeded", and
2615 "retry time not reached for any host after a long failure period".
2617 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2618 experimental DomainKeys support:
2620 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2621 (2) On an error such as an illegally used "control", the wrong name for
2622 the control was given.
2624 These problems did NOT occur unless DomainKeys support was compiled.
2626 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2628 PH/24 Added ${if match_ip condition.
2630 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2631 databases so that it will be absolutely obvious if a crash occurs in the
2632 DB library. This is a regular occurrence (often caused by mis-matched
2635 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2636 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2637 picked up on a user's system that detects such things. There doesn't seem
2638 to be a gcc warning option for this - only an attribute that has to be
2639 put on the function's prototype. It seems that in Fedora Core 4 they have
2640 set this on a number of new functions. No doubt there will be more in due
2643 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2644 specify a transport. However, if an address that was verified by such a
2645 router was the subject of a callout, Exim crashed because it tried to
2646 read the rcpt_include_affixes from the non-existent transport. Now it
2647 just assumes that the setting of that option is false. This bug was
2648 introduced by 4.51/PH/31.
2650 PH/28 Changed -d+all to exclude +memory, because that information is very
2651 rarely of interest, but it makes the output a lot bigger. People tend to
2652 do -d+all out of habit.
2654 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2655 code in os-type was giving problems when libc.so lives in lib64, like on
2658 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2659 aren't the modern standard, and it seems that some systems' include files
2660 don't always have them. Exim was already checking for some of the newer
2661 ones like T_AAAA, and defining it itself. I've added checks for all the
2662 record types that Exim uses.
2664 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2665 not automatically generating a new one, as it is supposed to. This
2666 prevented TLS from working. If the file did exist, but contained invalid
2667 data, a new version was generated, as expected. It was only the case of a
2668 non-existent file that was broken.
2670 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2671 with a change in libdomainkeys > 0.64.
2673 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2674 from DNS. If the selector record carries the flag, it now has
2675 precedence over the domain-wide flag.
2677 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2679 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2680 the use of a port name that isn't defined in /etc/services) occurred, the
2681 message was deferred as in a normal delivery, and thus remained on the
2682 spool, instead of being failed because of the mua_wrapper setting. This
2683 is now fixed, and I tidied up some of the mua_wrapper messages at the
2686 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2687 the messages in a hash of arrays rather than using individual hashes.
2688 This is a bit cleaner and results in dramatic memory savings, albeit
2689 at a slight CPU cost.
2691 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2692 as requested by Marc Sherman.
2694 SC/10 Eximstats: added histograms for user specified patterns as requested
2697 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2699 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2700 fopen() in the content-scanning modules that did not already have it.
2706 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2707 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2709 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2711 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2713 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2714 to test Sieve filters that use "vacation".
2716 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2717 that changes the way the GnuTLS parameters are stored in the cache file.
2718 The new format can be generated externally. For backward compatibility,
2719 if the data in the cache doesn't make sense, Exim assumes it has read an
2720 old-format file, and it generates new data and writes a new file. This
2721 means that you can't go back to an older release without removing the
2724 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2725 work if there are any delivery delays because "one_time" forces the
2726 parent to be marked "delivered", so its unseen clone is never tried
2727 again. For this reason, Exim now forbids the simultaneous setting of
2730 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2731 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2732 entry if you want to know the details. The fix, however, neglected to
2733 consider the case when local delivery batching is involved. The test for
2734 "previously delivered" was not happening when checking to see if an
2735 address could be batched with a previous (undelivered) one; under
2736 certain circumstances this could lead to multiple deliveries to the same
2739 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2740 in its include files, and this causes problems building Exim.
2742 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2743 header_syntax) but Exim was just ignoring anything given after a slash.
2744 In particular, this caused confusion with an attempt to use "verify =
2745 reverse_host_lookup/defer_ok". An error is now given when options are
2746 supplied for verify items that do not have them. (Maybe reverse_host_
2747 lookup should have a defer_ok option, but that's a different point.)
2749 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2750 defined by RFC 821) to 2048, because there were problems with some AUTH
2751 commands, and RFC 1869 says the size should be increased for extended
2752 SMTP commands that take arguments.
2754 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2757 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2758 "unknown" error; now it says that the functionality isn't in the binary.
2760 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2761 an address' error message when a string expansion fails (syntax or
2762 whatever). Otherwise the password may appear in the log. Following change
2763 PH/42 below, there is no longer a chance of it appearing in a bounce
2766 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2768 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2769 of its addresses was ever tried. (Bugzilla bug #2.)
2771 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2772 the result incorrectly in the debug output. (It correctly added a newline
2773 to what was transported.)
2775 TF/01 Added $received_time.
2777 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2778 commented out examples of how to interface to a virus scanner and to
2779 SpamAssassin. Also added commented examples of av_scanner and
2780 spamd_address settings.
2782 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2783 and controls are allowed in which ACLs. There were a couple of minor
2784 errors. Some of the entries in the conditions table (which is a table of
2785 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2786 negation of where the condition IS allowed.
2788 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2790 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2791 header file does not have a version number, so I've had to invent a new
2792 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2793 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2794 radiusclient), but it was contributed by a Radius user.
2796 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2797 files or directories, for OpenSSL.
2799 PH/19 When an Exim process that is running as root has to create an Exim log
2800 file, it does so in a subprocess that runs as exim:exim so as to get the
2801 ownership right at creation (otherwise, other Exim processes might see
2802 the file with the wrong ownership). There was no test for failure of this
2803 fork() call, which would lead to the process getting stuck as it waited
2804 for a non-existent subprocess. Forks do occasionally fail when resources
2805 run out. I reviewed all the other calls to fork(); they all seem to check
2808 PH/20 When checking for unexpected SMTP input at connect time (before writing
2809 the banner), Exim was not dealing correctly with a non-positive return
2810 from the read() function. If the client had disconnected by this time,
2811 the result was a log entry for a synchronization error with an empty
2812 string after "input=" when read() returned zero. If read() returned -1
2813 (an event I could not check), uninitialized data bytes were printed.
2814 There were reports of junk text (parts of files, etc) appearing after
2817 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2819 PH/22 Added support for macro redefinition, and (re)definition in between
2820 driver and ACL definitions.
2822 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2823 forgetting to use the resulting value; it was using the unexpanded value.
2825 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2826 hadn't been configured. The fix is from Juergen Kreileder, who
2827 understands it better than I do:
2829 "Here's what I see happening with three configured cyrus_sasl
2830 authenticators configured (plain, login, cram-md5):
2832 On startup auth_cyrus_sasl_init() gets called for each of these.
2833 This means three calls to sasl_listmech() without a specified mech_list.
2834 => SASL tests which mechs of all available mechs actually work
2835 => three warnings about OTP not working
2836 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2838 With the patch, sasl_listmech() also gets called three times. But now
2839 SASL's mech_list option is set to the server_mech specified in the the
2840 authenticator. Or in other words, the answer from sasl_listmech()
2841 gets limited to just the mech you're testing for (which is different
2843 => the return list contains just 'plain' or 'login', 'cram-md5' or
2844 nothing depending on the value of ob->server_mech.
2846 I've just tested the patch: Authentication still works fine,
2847 unavailable mechs specified in the exim configuration are still
2848 caught, and the auth.log warnings about OTP are gone."
2850 PH/25 When debugging is enabled, the contents of the command line are added
2851 to the debugging output, even when log_selector=+arguments is not
2854 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2855 answer is "GNU", and only if the return is "GNU/something" is the answer
2858 PH/27 $acl_verify_message is now set immediately after the failure of a
2859 verification in an ACL, and so is available in subsequent modifiers. In
2860 particular, the message can be preserved by coding like this:
2862 warn !verify = sender
2863 set acl_m0 = $acl_verify_message
2865 Previously, $acl_verify_message was set only while expanding "message"
2866 and "log_message" when a very denied access.
2868 PH/28 Modified OS/os.c-Linux with
2870 -#ifndef OS_LOAD_AVERAGE
2871 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2873 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2874 nomenclature these days.)
2876 PH/29 Installed patch from the Sieve maintainer that adds the options
2877 sieve_useraddress and sieve_subaddress to the redirect router.
2879 PH/30 In these circumstances:
2880 . Two addresses routed to the same list of hosts;
2881 . First host does not offer TLS;
2882 . First host accepts first address;
2883 . First host gives temporary error to second address;
2884 . Second host offers TLS and a TLS session is established;
2885 . Second host accepts second address.
2886 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2887 and peerdn, if requested) that were in fact used only for the second
2890 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2891 attention to any local part prefix or suffix that was matched by the
2892 router that accepted the address. It now behaves in the same way as it
2893 does for delivery: the affixes are removed from the local part unless
2894 rcpt_include_affixes is set on the transport.
2896 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2897 timeout during the DATA phase of an incoming message.
2899 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2900 applied a patch sanctioned by the Sieve maintainer.
2902 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2903 the uid or gid is negative. A case of a negative gid caused this to be
2904 noticed. The fix allows for either to be negative.
2906 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2907 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2910 PH/36 The change PH/12 above was broken. Fixed it.
2912 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2913 the grounds that routing the same address twice would always produce the
2914 same answer. This might have been true once, but it is certainly no
2915 longer true now. Routing a child address may depend on the previous
2916 routing that produced that child. Some complicated redirection strategies
2917 went wrong when messages had multiple recipients, and made Exim's
2918 behaviour dependent on the order in which the addresses were given.
2920 I have moved the duplicate checking until after the routing is complete.
2921 Exim scans the addresses that are assigned to local and remote
2922 transports, and removes any duplicates. This means that more work will be
2923 done, as duplicates will always all be routed, but duplicates are
2924 presumably rare, so I don't expect this is of any significance.
2926 For deliveries to pipes, files, and autoreplies, the duplicate checking
2927 still happens during the routing process, since they are not going to be
2930 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2931 It corrects a timeout issue with spamd. This is Ian's comment: "The
2932 background is that sometimes spamd either never reads data from a
2933 connection it has accepted, or it never writes response data. The exiscan
2934 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2935 blindly assumes that writes won't block so it may never time out."
2937 PH/39 Allow G after quota size as well as K and M.
2939 PH/40 The value set for $authenticated_id in an authenticator may not contain
2940 binary zeroes or newlines because the value is written to log lines and
2941 to spool files. There was no check on this. Now the value is run through
2942 the string_printing() function so that such characters are converted to
2943 printable escape sequences.
2945 PH/41 $message_linecount is a new variable that contains the total number of
2946 lines in the message. Compare $body_linecount, which is the count for the
2949 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2950 bounce and delay warning messages, except in certain special cases, which
2953 (a) An SMTP error message from a remote host;
2954 (b) A message specified in a :fail: redirection;
2955 (c) A message specified in a "fail" command in a system filter;
2956 (d) A message specified in a FAIL return from the queryprogram router;
2957 (e) A message specified by the cannot_route_message router option.
2959 In these cases only, Exim does include the error details in bounce and
2960 warning messages. There are also a few cases where bland messages such
2961 as "unrouteable address" or "local delivery error" are given.
2963 PH/43 $value is now also set for the "else" part of a ${run expansion.
2965 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2966 being worked on, but at least Exim now implements the latest version to
2969 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2970 process to complete was treated as a delivery failure, a timeout while
2971 writing the message to the pipe was logged, but erroneously treated as a
2972 successful delivery. Such timeouts include transport filter timeouts. For
2973 consistency with the overall process timeout, these timeouts are now
2974 treated as errors, giving rise to delivery failures by default. However,
2975 there is now a new Boolean option for the pipe transport called
2976 timeout_defer, which, if set TRUE, converts the failures into defers for
2977 both kinds of timeout. A transport filter timeout is now identified in
2980 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2981 systems where "make" and "gmake" are different, calling "gmake" at top
2982 level broke things. I've arranged for the value of $(MAKE) to be passed
2983 from the Makefile to this script so that it can call the same version of
2987 A note about Exim versions 4.44 and 4.50
2988 ----------------------------------------
2990 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2991 changes of various kinds. As a consequence, a big documentation update was
2992 needed. This delayed the release for rather longer than seemed good, especially
2993 in the light of a couple of (minor) security issues. Therefore, the changes
2994 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2995 release. So 4.44 and 4.50 are in effect two different branches that both start
2998 I have left the 4.50 change log unchanged; it contains all the changes since
2999 4.43. The change log for 4.44 is below; many of its items are identical to
3000 those for 4.50. This seems to be the most sensible way to preserve the
3001 historical information.
3007 1. Minor wording change to the doc/README.SIEVE file.
3009 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3010 computation of the current number of files was incorrect.
3012 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3013 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3016 4. Give more explanation in the error message when the command for a transport
3017 filter fails to execute.
3019 5. There are several places where Exim runs a non-Exim command in a
3020 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3021 was being done only for the command run by the queryprogram router. It is
3022 now done for all such subprocesses. The other cases are: ${run, transport
3023 filters, and the commands run by the lmtp and pipe transports.
3025 6. Added CONFIGURE_GROUP build-time option.
3027 7. Some older OS have a limit of 256 on the maximum number of file
3028 descriptors. Exim was using setrlimit() to set 1000 as a large value
3029 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3030 systems. I've change it so that if it can't get 1000, it tries for 256.
3032 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3033 was an oversight, and furthermore, ever since the addition of extra
3034 controls (e.g. 4.43/32), the checks on when to allow different forms of
3035 "control" were broken. There should now be diagnostics for all cases when a
3036 control that does not make sense is encountered.
3038 9. Added the /retain_sender option to "control=submission".
3040 10. $recipients is now available in the predata ACL (oversight).
3042 11. Tidy the search cache before the fork to do a delivery from a message
3043 received from the command line. Otherwise the child will trigger a lookup
3044 failure and thereby defer the delivery if it tries to use (for example) a
3045 cached ldap connection that the parent has called unbind on.
3047 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3048 of $address_data from the recipient verification was clobbered by the
3049 sender verification.
3051 13. The value of address_data from a sender verification is now available in
3052 $sender_address_data in subsequent conditions in the ACL statement.
3054 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3056 15. Added a new option "connect=<time>" to callout options, to set a different
3059 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3060 was its contents. (It was OK if the option was not defined at all.)
3062 17. A "Completed" log line is now written for messages that are removed from
3063 the spool by the -Mrm option.
3065 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3066 information about exactly what failed.
3068 19. Added -dd to debug only the daemon process.
3070 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3071 handles timeouts, both on the server side and network timeouts. Renamed the
3072 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3074 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3075 It is now set to "smtps".
3077 22. $host_address is now set to the target address during the checking of
3078 ignore_target_hosts.
3080 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3081 being passed; this would have caused $sender_host_name to have been used if
3082 matching the list had actually called for a host name (not very likely,
3083 since this list is usually IP addresses). A host name is now passed as
3086 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3087 code that specifies a non-restarting handler (typically sigaction() in
3088 modern systems) in an attempt to fix a rare and obscure crash bug.
3090 25. Narrowed the window for a race in the daemon that could cause it to ignore
3091 SIGCHLD signals. This is not a major problem, because they are used only to
3092 wake it up if nothing else does.
3094 26. A malformed maildirsize file could cause Exim to calculate negative values
3095 for the mailbox size or file count. Odd effects could occur as a result.
3096 The maildirsize information is now recalculated if the size or filecount
3099 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3100 support for a long time. Removed HAVE_SYS_VFS_H.
3102 28. Installed the latest version of exipick from John Jetmore.
3104 29. In an address list, if the pattern was not a regular expression, an empty
3105 subject address (from a bounce message) matched only if the pattern was an
3106 empty string. Non-empty patterns were not even tested. This was the wrong
3107 because it is perfectly reasonable to use an empty address as part of a
3108 database query. An empty address is now tested by patterns that are
3109 lookups. However, all the other forms of pattern expect the subject to
3110 contain a local part and a domain, and therefore, for them, an empty
3111 address still always fails if the pattern is not itself empty.
3113 30. Exim went into a mad DNS loop when attempting to do a callout where the
3114 host was specified on an smtp transport, and looking it up yielded more
3115 than one IP address.
3117 31. Re-factored the code for checking spool and log partition space into a
3118 function that finds that data and another that does the check. The former
3119 is then used to implement four new variables: $spool_space, $log_space,
3120 $spool_inodes, and $log_inodes.
3122 32. The RFC2047 encoding function was originally intended for short strings
3123 such as real names; it was not keeping to the 75-character limit for
3124 encoded words that the RFC imposes. It now respects the limit, and
3125 generates multiple encoded words if necessary. To be on the safe side, I
3126 have increased the buffer size for the ${rfc2047: expansion operator from
3129 33. It is now permitted to omit both strings after an "if" condition; if the
3130 condition is true, the result is "true". As before, when the second string
3131 is omitted, a false condition yields an empty string. This makes it less
3132 cumbersome to write custom ACL and router conditions.
3134 34. Failure to deliver a bounce message always caused it to be frozen, even if
3135 there was an errors_to setting on the router. The errors_to setting is now
3138 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3139 canonical form (fully expanded) before being placed in
3140 $sender_host_address.
3142 36. The table in the code that translates DNS record types into text (T_A to
3143 "A" for instance) was missing entries for NS and CNAME. It is just possible
3144 that this could have caused confusion if both these types were looked up
3145 for the same domain, because the text type is used as part of Exim's
3146 per-process caching. But the chance of anyone hitting this buglet seems
3149 37. The dnsdb lookup has been extended in a number of ways.
3151 (1) There is a new type, "zns", which walks up the domain tree until it
3152 finds some nameserver records. It should be used with care.
3154 (2) There is a new type, "mxh", which is like "mx" except that it returns
3155 just the host names, not the priorities.
3157 (3) It is now possible to give a list of domains (or IP addresses) to be
3158 looked up. The behaviour when one of the lookups defers can be
3159 controlled by a keyword.
3161 (4) It is now possible to specify the separator character for use when
3162 multiple records are returned.
3164 38. The dnslists ACL condition has been extended: it is now possible to supply
3165 a list of IP addresses and/or domains to be looked up in a particular DNS
3168 39. Added log_selector=+queue_time_overall.
3170 40. When running the queue in the test harness, wait just a tad after forking a
3171 delivery process, to get repeatability of debugging output.
3173 41. Include certificate and key file names in error message when GnuTLS fails
3174 to set them up, because the GnuTLS error message doesn't include the name
3175 of the failing file when there is a problem reading it.
3177 42. Allow both -bf and -bF in the same test run.
3179 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3181 44. The "Exiscan patch" is now merged into the mainline Exim source.
3183 45. Sometimes the final signoff response after QUIT could fail to get
3184 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3185 < 0 before doing a fflush(). This bug looks as though it goes back to the
3186 introduction of TLS in release 3.20, but "sometimes" must have been rare
3187 because the tests only now provoked it.
3189 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3190 (this can affect the format of dates).
3192 47. exim_tidydb, when checking for the continued existence of a message for
3193 which it has found a message-specific retry record, was not finding
3194 messages that were in split spool directories. Consequently, it was
3195 deleting retry records that should have stayed in existence.
3197 48. Steve fixed some bugs in eximstats.
3199 49. The SPA authentication driver was not abandoning authentication and moving
3200 on to the next authenticator when an expansion was forced to fail,
3201 contradicting the general specification for all authenticators. Instead it
3202 was generating a temporary error. It now behaves as specified.
3204 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3205 (the order specifies the preference for clients). The order is now AES256,
3206 AES128, 3DES, ARCFOUR128.
3208 51. Small patch to Sieve code - explicitly set From: when generating an
3211 52. Exim crashed if a remote delivery caused a very long error message to be
3212 recorded - for instance if somebody sent an entire SpamAssassin report back
3213 as a large number of 550 error lines. This bug was coincidentally fixed by
3214 increasing the size of one of Exim's internal buffers (big_buffer) that
3215 happened as part of the Exiscan merge. However, to be on the safe side, I
3216 have made the code more robust (and fixed the comments that describe what
3219 53. Now that there can be additional text after "Completed" in log lines (if
3220 the queue_time_overall log selector is set), a one-byte patch to exigrep
3221 was needed to allow it to recognize "Completed" as not the last thing in
3224 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3225 patch that reportedly fixes this has been added. I am not expert enough to
3226 create a test for it. This is what the patch creator wrote:
3228 "I found a little strange behaviour of ldap code when working with
3229 Windows 2003 AD Domain, where users was placed in more than one
3230 Organization Units. When I tried to give exim partial DN, the exit code
3231 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3232 But simultaneously result of request was absolutely normal ldap result,
3233 so I produce this patch..."
3235 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3236 REFERENCE, so I have modified the code to exclude the patch when that macro
3239 55. Some experimental protocols are using DNS PTR records for new purposes. The
3240 keys for these records are domain names, not reversed IP addresses. The
3241 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3242 leaves it alone. Component reversal etc. now happens only for IP addresses.
3245 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3247 57. Double the size of the debug message buffer (to 2048) so that more of very
3248 long debug lines gets shown.
3250 58. The exicyclog utility now does better if the number of log files to keep
3251 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3253 59. Two changes related to the smtp_active_hostname option:
3255 (1) $smtp_active_hostname is now available as a variable.
3256 (2) The default for smtp_banner uses $smtp_active_hostname instead
3257 of $primary_hostname.
3259 60. The host_aton() function is supposed to be passed a string that is known
3260 to be a valid IP address. However, in the case of IPv6 addresses, it was
3261 not checking this. This is a hostage to fortune. Exim now panics and dies
3262 if the condition is not met. A case was found where this could be provoked
3263 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3264 components; fortuitously, this particular loophole had already been fixed
3265 by change 4.50/55 above.
3267 If there are any other similar loopholes, the new check in host_aton()
3268 itself should stop them being exploited. The report I received stated that
3269 data on the command line could provoke the exploit when Exim was running as
3270 exim, but did not say which command line option was involved. All I could
3271 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3272 running as the user.
3275 61. There was a buffer overflow vulnerability in the SPA authentication code
3276 (which came originally from the Samba project). I have added a test to the
3277 spa_base64_to_bits() function which I hope fixes it.
3280 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3281 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3283 63. The daemon start-up calls getloadavg() while still root for those OS that
3284 need the first call to be done as root, but it missed one case: when
3285 deliver_queue_load_max is set with deliver_drop_privilege. This is
3286 necessary for the benefit of the queue runner, because there is no re-exec
3287 when deliver_drop_privilege is set.
3289 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3290 This has been fixed.
3292 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3293 was in use, was not putting the data itself into the right store pool;
3294 consequently, it could be overwritten for a subsequent message in the same
3295 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3298 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3300 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3301 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3302 the function were treating the return as a boolean value, which happened to
3303 work because 0=false and not-0=true, but is not correct code.
3305 68. The host_aton() function was not handling scoped IPv6 addresses (those
3306 with, for example, "%eth0" on the end) correctly.
3308 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3309 negated items (that is, ~something) in unsigned ints. Some compilers
3310 apparently mutter when there is no cast.
3312 70. If an address verification called from an ACL failed, and did not produce a
3313 user-specific message (i.e. there was only a "system" message), nothing was
3314 put in $acl_verify_message. In this situation, it now puts the system
3317 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3318 session; change 4.31/43 added the unwanted input to the log line - except
3319 that it did not do this in the start of session case. It now does.
3321 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3322 This is wrong and can cause the other end to generate a synchronization
3323 error if it is another Exim or anything else that does the synchronization
3324 check. A QUIT command is no longer sent after a timeout.
3326 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3327 during host lookups.
3329 74. The defer_ok option of callout verification was not working if it was used
3330 when verifying addresses in header lines, that is, for this case:
3332 verify = header_sender/callout=defer_ok
3334 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3335 those file descriptors could be used for SMTP connections. If anything
3336 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3337 could be sent to the SMTP client, causing chaos. The daemon now opens
3338 stdin, stdout, and stderr to /dev/null when it puts itself into the
3341 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3342 log by default. The user can override this with suitable Perl magic.
3344 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3345 the log message when discard triggers, was not working for the DATA ACL or
3346 for the non-SMTP ACL.
3348 78. Error message wording change in sieve.c.
3350 79. If smtp_accept_max_per_host was set, the number of connections could be
3351 restricted to fewer than expected, because the daemon was trying to set up
3352 a new connection before checking whether the processes handling previous
3353 connections had finished. The check for completed processes is now done
3354 earlier. On busy systems, this bug wouldn't be noticed because something
3355 else would have woken the daemon, and it would have reaped the completed
3358 80. If a message was submitted locally by a user whose login name contained one
3359 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3360 It caused a spool format error. I have fixed the spool reading code. A
3361 related problem was that the "from" clause in the Received: line became
3362 illegal because of the space(s). It is now covered by ${quote_local_part.
3364 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3367 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3369 83. Patch from Sieve maintainer to fix unterminated string problem in
3370 "vacation" handling.
3372 84. Some minor changes to the Linux configuration files to help with other
3373 OS variants using glibc.
3375 85. One more patch for Sieve to update vacation handling to latest spec.
3378 ----------------------------------------------------
3379 See the note above about the 4.44 and 4.50 releases.
3380 ----------------------------------------------------
3386 1. Change 4.43/35 introduced a bug that caused file counts to be
3387 incorrectly computed when quota_filecount was set in an appendfile
3390 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3391 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3394 3. Give more explanation in the error message when the command for a transport
3395 filter fails to execute.
3397 4. There are several places where Exim runs a non-Exim command in a
3398 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3399 was being done only for the command run by the queryprogram router. It is
3400 now done for all such subprocesses. The other cases are: ${run, transport
3401 filters, and the commands run by the lmtp and pipe transports.
3403 5. Some older OS have a limit of 256 on the maximum number of file
3404 descriptors. Exim was using setrlimit() to set 1000 as a large value
3405 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3406 systems. I've change it so that if it can't get 1000, it tries for 256.
3408 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3409 was an oversight, and furthermore, ever since the addition of extra
3410 controls (e.g. 4.43/32), the checks on when to allow different forms of
3411 "control" were broken. There should now be diagnostics for all cases when a
3412 control that does not make sense is encountered.
3414 7. $recipients is now available in the predata ACL (oversight).
3416 8. Tidy the search cache before the fork to do a delivery from a message
3417 received from the command line. Otherwise the child will trigger a lookup
3418 failure and thereby defer the delivery if it tries to use (for example) a
3419 cached ldap connection that the parent has called unbind on.
3421 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3422 of $address_data from the recipient verification was clobbered by the
3423 sender verification.
3425 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3426 was its contents. (It was OK if the option was not defined at all.)
3428 11. A "Completed" log line is now written for messages that are removed from
3429 the spool by the -Mrm option.
3431 12. $host_address is now set to the target address during the checking of
3432 ignore_target_hosts.
3434 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3435 being passed; this would have caused $sender_host_name to have been used if
3436 matching the list had actually called for a host name (not very likely,
3437 since this list is usually IP addresses). A host name is now passed as
3440 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3441 code that specifies a non-restarting handler (typically sigaction() in
3442 modern systems) in an attempt to fix a rare and obscure crash bug.
3444 15. Narrowed the window for a race in the daemon that could cause it to ignore
3445 SIGCHLD signals. This is not a major problem, because they are used only to
3446 wake it up if nothing else does.
3448 16. A malformed maildirsize file could cause Exim to calculate negative values
3449 for the mailbox size or file count. Odd effects could occur as a result.
3450 The maildirsize information is now recalculated if the size or filecount
3453 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3454 support for a long time. Removed HAVE_SYS_VFS_H.
3456 18. Updated exipick to current release from John Jetmore.
3458 19. Allow an empty sender to be matched against a lookup in an address list.
3459 Previously the only cases considered were a regular expression, or an
3462 20. Exim went into a mad DNS lookup loop when doing a callout where the
3463 host was specified on the transport, if the DNS lookup yielded more than
3466 21. The RFC2047 encoding function was originally intended for short strings
3467 such as real names; it was not keeping to the 75-character limit for
3468 encoded words that the RFC imposes. It now respects the limit, and
3469 generates multiple encoded words if necessary. To be on the safe side, I
3470 have increased the buffer size for the ${rfc2047: expansion operator from
3473 22. Failure to deliver a bounce message always caused it to be frozen, even if
3474 there was an errors_to setting on the router. The errors_to setting is now
3477 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3478 canonical form (fully expanded) before being placed in
3479 $sender_host_address.
3481 24. Updated eximstats to version 1.33
3483 25. Include certificate and key file names in error message when GnuTLS fails
3484 to set them up, because the GnuTLS error message doesn't include the name
3485 of the failing file when there is a problem reading it.
3487 26. Expand error message when OpenSSL has problems setting up cert/key files.
3490 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3491 (this can affect the format of dates).
3493 28. exim_tidydb, when checking for the continued existence of a message for
3494 which it has found a message-specific retry record, was not finding
3495 messages that were in split spool directories. Consequently, it was
3496 deleting retry records that should have stayed in existence.
3498 29. eximstats updated to version 1.35
3499 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3500 1.35 - bugfix such that pie charts by volume are generated correctly
3502 30. The SPA authentication driver was not abandoning authentication and moving
3503 on to the next authenticator when an expansion was forced to fail,
3504 contradicting the general specification for all authenticators. Instead it
3505 was generating a temporary error. It now behaves as specified.
3507 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3508 (the order specifies the preference for clients). The order is now AES256,
3509 AES128, 3DES, ARCFOUR128.
3511 31. Small patch to Sieve code - explicitly set From: when generating an
3514 32. Exim crashed if a remote delivery caused a very long error message to be
3515 recorded - for instance if somebody sent an entire SpamAssassin report back
3516 as a large number of 550 error lines. This bug was coincidentally fixed by
3517 increasing the size of one of Exim's internal buffers (big_buffer) that
3518 happened as part of the Exiscan merge. However, to be on the safe side, I
3519 have made the code more robust (and fixed the comments that describe what
3522 33. Some experimental protocols are using DNS PTR records for new purposes. The
3523 keys for these records are domain names, not reversed IP addresses. The
3524 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3525 leaves it alone. Component reversal etc. now happens only for IP addresses.
3528 34. The host_aton() function is supposed to be passed a string that is known
3529 to be a valid IP address. However, in the case of IPv6 addresses, it was
3530 not checking this. This is a hostage to fortune. Exim now panics and dies
3531 if the condition is not met. A case was found where this could be provoked
3532 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3533 components; fortuitously, this particular loophole had already been fixed
3534 by change 4.50/55 or 4.44/33 above.
3536 If there are any other similar loopholes, the new check in host_aton()
3537 itself should stop them being exploited. The report I received stated that
3538 data on the command line could provoke the exploit when Exim was running as
3539 exim, but did not say which command line option was involved. All I could
3540 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3541 running as the user.
3544 35. There was a buffer overflow vulnerability in the SPA authentication code
3545 (which came originally from the Samba project). I have added a test to the
3546 spa_base64_to_bits() function which I hope fixes it.
3549 36. The daemon start-up calls getloadavg() while still root for those OS that
3550 need the first call to be done as root, but it missed one case: when
3551 deliver_queue_load_max is set with deliver_drop_privilege. This is
3552 necessary for the benefit of the queue runner, because there is no re-exec
3553 when deliver_drop_privilege is set.
3555 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3556 was in use, was not putting the data itself into the right store pool;
3557 consequently, it could be overwritten for a subsequent message in the same
3558 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3561 38. Sometimes the final signoff response after QUIT could fail to get
3562 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3563 < 0 before doing a fflush(). This bug looks as though it goes back to the
3564 introduction of TLS in release 3.20, but "sometimes" must have been rare
3565 because the tests only now provoked it.
3571 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3572 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3573 Now it returns an integer. A number of calls were still expecting a T/F
3574 return. Fortuitously, in all cases, the tests worked in OK situations,
3575 which is the norm. However, things would have gone wrong on any write
3576 failures on the smtp file descriptor. This function is used when sending
3577 messages over SMTP and also when doing verify callouts.
3579 2. When Exim is called to do synchronous delivery of a locally submitted
3580 message (the -odf or -odi options), it no longer closes stderr before doing
3583 3. Implemented the mua_wrapper option.
3585 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3587 5. Implemented the functions header_remove(), header_testname(),
3588 header_add_at_position(), and receive_remove_recipient(), and exported them
3591 6. If an ACL "warn" statement specified the addition of headers, Exim already
3592 inserted X-ACL-Warn: at the start if there was no header name. However, it
3593 was not making this test for the second and subsequent header lines if
3594 there were newlines in the string. This meant that an invalid header could
3595 be inserted if Exim was badly configured.
3597 7. Allow an ACL "warn" statement to add header lines at the start or after all
3598 the Received: headers, as well as at the end.
3600 8. Added the rcpt_4xx retry error code.
3602 9. Added postmaster_mailfrom=xxx to callout verification option.
3604 10. Added mailfrom=xxxx to the callout verification option, for verify=
3607 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3608 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3610 12. Inserted some casts to stop certain compilers warning when using pointer
3611 differences as field lengths or precisions in printf-type calls (mostly
3612 affecting debugging statements).
3614 13. Added optional readline() support for -be (dynamically loaded).
3616 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3617 same clock tick as a message's arrival, so that its received time was the
3618 same as the "first fail" time on the retry record, and that message
3619 remained on the queue past the ultimate address timeout, every queue runner
3620 would try a delivery (because it was past the ultimate address timeout) but
3621 after another failure, the ultimate address timeout, which should have then
3622 bounced the address, did not kick in. This was a "< instead of <=" error;
3623 in most cases the first failure would have been in the next clock tick
3624 after the received time, and all would be well.
3626 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3627 being recognized when the domain list was tested by the match_domain
3628 condition in an expansion string.
3630 16. Added the ${str2b64: operator.
3632 17. Exim was always calling setrlimit() to set a large limit for the number of
3633 processes, without checking whether the existing limit was already
3634 adequate. (It did check for the limit on file descriptors.) Furthermore,
3635 errors from getrlimit() and setrlimit() were being ignored. Now they are
3636 logged to the main and panic logs, but Exim does carry on, to try to do its
3637 job under whatever limits there are.
3639 18. Imported PCRE 5.0.
3641 19. Trivial typo in log message " temporarily refused connection" (the leading
3644 20. If the log selector return_path_on_delivery was set and an address was
3645 redirected to /dev/null, the delivery process crashed because it assumed
3646 that a return path would always be set for a "successful" delivery. In this
3647 case, the whole delivery is bypassed as an optimization, and therefore no
3650 21. Internal re-arrangement: the function for sending a challenge and reading
3651 a response while authentication was assuming a zero-terminated challenge
3652 string. It's now changed to take a pointer and a length, to allow for
3653 binary data in such strings.
3655 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3657 23. Exim was not respecting finduser_retries when seeking the login of the
3658 uid under which it was called; it was always trying 10 times. (The default
3659 setting of finduser_retries is zero.) Also, it was sleeping after the final
3660 failure, which is pointless.
3662 24. Implemented tls_on_connect_ports.
3664 25. Implemented acl_smtp_predata.
3666 26. If the domain in control=submission is set empty, Exim assumes that the
3667 authenticated id is a complete email address when it generates From: or
3668 Sender: header lines.
3670 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3671 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3672 chown and chgrp in /bin and hostname in /usr/bin.
3674 28. Exim was keeping the "process log" file open after each use, just as it
3675 does for the main log. This opens the possibility of it remaining open for
3676 long periods when the USR1 signal hits a daemon. Occasional processlog
3677 errors were reported, that could have been caused by this. Anyway, it seems
3678 much more sensible not to leave this file open at all, so that is what now
3681 29. The long-running daemon process does not normally write to the log once it
3682 has entered its main loop, and it closes the log before doing so. This is
3683 so that log files can straightforwardly be renamed and moved. However,
3684 there are a couple of unusual error situations where the daemon does write
3685 log entries, and I had neglected to close the log afterwards.
3687 30. The text of an SMTP error response that was received during a remote
3688 delivery was being truncated at 512 bytes. This is too short for some of
3689 the long messages that one sometimes sees. I've increased the limit to
3692 31. It is now possible to make retry rules that apply only when a message has a
3693 specific sender, in particular, an empty sender.
3695 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3696 it possible to be selective about when SMTP synchronization is enforced.
3698 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3700 32. Implemented hosts_connection_nolog.
3702 33. Added an ACL for QUIT.
3704 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3707 35. Added mailbox_size and mailbox_filecount to appendfile.
3709 36. Added control = no_multiline_responses to ACLs.
3711 37. There was a bug in the logic of the code that waits for the clock to tick
3712 in the case where the clock went backwards by a substantial amount such
3713 that the microsecond fraction of "now" was more than the microsecond
3714 fraction of "then" (but the whole seconds number was less).
3716 38. Added support for the libradius Radius client library this is found on
3717 FreeBSD (previously only the radiusclient library was supported).
3723 1. When certain lookups returned multiple values in the form name=value, the
3724 quoting of the values was not always being done properly. Specifically:
3725 (a) If the value started with a double quote, but contained no whitespace,
3727 (b) If the value contained whitespace other than a space character (i.e.
3728 tabs or newlines or carriage returns) it was not quoted.
3729 This fix has been applied to the mysql and pgsql lookups by writing a
3730 separate quoting function and calling it from the lookup code. The fix
3731 should probably also be applied to nisplus, ibase and oracle lookups, but
3732 since I cannot test any of those, I have not disturbed their existing code.
3734 2. A hit in the callout cache for a specific address caused a log line with no
3735 reason for rejecting RCPT. Now it says "Previous (cached) callout
3736 verification failure".
3738 3. There was an off-by-one bug in the queryprogram router. An over-long
3739 return line was truncated at 256 instead of 255 characters, thereby
3740 overflowing its buffer with the terminating zero. As well as fixing this, I
3741 have increased the buffer size to 1024 (and made a note to document this).
3743 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3744 when Exim is waiting for an SMTP response from a remote server, Exim
3745 restarts its select() call on the socket, thereby resetting its timeout.
3746 This is not a problem when such interrupts are rare. Somebody set up a cron
3747 job to run exiwhat every 2 minutes, which is less than the normal select()
3748 timeout (5 or 10 minutes). This meant that the select() timeout never
3749 kicked in because it was always reset. I have fixed this by comparing the
3750 time when an interrupt arrives with the time at the start of the first call
3751 to select(). If more time than the timeout has elapsed, the interrupt is
3752 treated as a timeout.
3754 5. Some internal re-factoring in preparation for the addition of Sieve
3755 extensions (by MH). In particular, the "personal" test is moved to a
3756 separate function, and given an option for scanning Cc: and Bcc: (which is
3757 not set for Exim filters).
3759 6. When Exim created an email address using the login of the caller as the
3760 local part (e.g. when creating a From: or Sender: header line), it was not
3761 quoting the local part when it contained special characters such as @.
3763 7. Installed new OpenBSD configuration files.
3765 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3766 try to make them clearer.
3768 9. Callout options, other than the timeout value, were being ignored when
3769 verifying sender addresses in header lines. For example, when using
3771 verify = header_sender/callout=no_cache
3773 the cache was (incorrectly) being used.
3775 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3776 only the Cygwin environment.
3778 11. When return_path_on_delivery was set as a log selector, if different remote
3779 addresses in the same message used different return paths and parallel
3780 remote delivery occurred, the wrong values would sometimes be logged.
3781 (Whenever a remote delivery process finished, the return path value from
3782 the most recently started remote delivery process was logged.)
3784 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3785 lines when AUTH and/or TLS are in use. This is the "received protocol"
3786 field. Exim used to use "asmtp" for authenticated SMTP, without any
3787 indication (in the protocol name) for TLS use. Now it follows the RFC and
3788 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3789 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3790 names appear in log lines as well as in Received: header lines.
3792 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3793 extensions, and comparison tests, and to fix some bugs.
3795 14. Changes to the "personal" filter test:
3797 (1) The test was buggy in that it was just doing the equivalent of
3798 "contains" tests on header lines. For example, if a user's address was
3799 anne@some.where, the "personal" test would incorrectly be true for
3801 To: susanne@some.where
3803 This test is now done by extracting each address from the header in turn,
3804 and checking the entire address. Other tests that are part of "personal"
3805 are now done using regular expressions (for example, to check local parts
3806 of addresses in From: header lines).
3808 (2) The list of non-personal local parts in From: addresses has been
3809 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3810 taken from the Sieve specification recommendations.
3812 (3) If the message contains any header line starting with "List-" it is
3813 treated as non-personal.
3815 (4) The test for "circular" in the Subject: header line has been removed
3816 because it now seems ill-conceived.
3818 15. Minor typos in src/EDITME comments corrected.
3820 16. Installed latest exipick from John Jetmore.
3822 17. If headers_add on a router specified a text string that was too long for
3823 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3824 of string_sprintf() is now avoided.
3826 18. $message_body_size was not set (it was always zero) when running the DATA
3827 ACL and the local_scan() function.
3829 19. For the "mail" command in an Exim filter, no default was being set for
3830 the once_repeat time, causing a random time value to be used if "once" was
3831 specified. (If the value happened to be <= 0, no repeat happened.) The
3832 default is now 0s, meaning "never repeat". The "vacation" command was OK
3833 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3834 (I found it when inspecting the code).
3836 20. There is now an overall timeout for performing a callout verification. It
3837 defaults to 4 times the callout timeout, which applies to individual SMTP
3838 commands during the callout. The overall timeout applies when there is more
3839 than one host that can be tried. The timeout is checked before trying the
3840 next host. This prevents very long delays if there are a large number of
3841 hosts and all are timing out (e.g. when the network connections are timing
3842 out). The value of the overall timeout can be changed by specifying an
3843 additional sub-option for "callout", called "maxwait". For example:
3845 verify = sender/callout=5s,maxwait=20s
3847 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3848 to the end before writing, but this should make it even safer).
3850 22. Exim was forgetting that it had advertised PIPELINING for the second and
3851 subsequent messages on an SMTP connection. It was also not resetting its
3852 memory on STARTTLS and an internal HELO.
3854 23. When Exim logs an SMTP synchronization error within a session, it now
3855 records whether PIPELINING has been advertised or not.
3857 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3858 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3859 rather than long int.
3861 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3863 26. Added the never_mail option to autoreply.
3869 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3870 crash if the getsockname() call failed; this can happen if a connection is
3871 closed very soon after it is established. The problem was simply in the
3872 order in which certain operations were done, causing Exim to try to write
3873 to the SMTP stream before it had set up the file descriptor. The bug has
3874 been fixed by making things happen in the correct order.
3880 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3881 before the connection was closed, thus losing the rejection response.
3883 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3884 some early Solaris releases, but causes trouble in current releases where
3885 socklen_t is defined.
3887 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3890 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3893 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3894 that releasing the top of it at the end releases what was used for sub-
3895 expansions (unless the block got too big). However, discard this block if
3896 the first thing is a variable or header, so that we can use its block when
3897 it is dynamic (useful for very large $message_headers, for example).
3899 6. Lookups now cache *every* query, not just the most recent. A new, separate
3900 store pool is used for this. It can be recovered when all lookup caches are
3901 flushed. Lookups now release memory at the end of their result strings.
3902 This has involved some general refactoring of the lookup sources.
3904 7. Some code has been added to the store_xxx() functions to reduce the amount
3905 of flapping under certain conditions.
3907 8. log_incoming_interface used to affect only the <= reception log lines. Now
3908 it causes the local interface and port to be added to several more SMTP log
3909 lines, for example "SMTP connection from", and rejection lines.
3911 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3913 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3915 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3916 could be overwritten at the end of the current message (or the start of a
3917 new message if it was set in a HELO ACL). The value is now preserved for
3918 the duration of the SMTP connection.
3920 12. If a transport had a headers_rewrite setting, and a matching header line
3921 contained an unqualified address, that address was qualified, even if it
3922 did not match any rewriting rules. The underlying bug was that the values
3923 of the flags that permit the existence of unqualified sender and recipient
3924 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3925 non-local messages, and by -bnq for local messages) were not being
3926 preserved with the message after it was received.
3928 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3929 "next input=" as part of the text comprising the host identity instead of
3930 the correct text. The code was using the same buffer for two different
3931 strings. However, depending on which order the printing function evaluated
3932 its arguments, the bug did not always show up. Under Linux, for example, my
3933 test suite worked just fine.
3935 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3936 doesn't work with some older versions of Perl. It has been changed to "my",
3937 which in any case is probably the better facility to use.
3939 15. A really picky compiler found some instances of statements for creating
3940 error messages that either had too many or two few arguments for the format
3943 16. The size of the buffer for calls to the DNS resolver has been increased
3944 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3945 for addresses that have a lot of PTR records. This alleviates a problem; it
3946 does not fully solve it.
3948 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3949 buffer now truncates the list and logs the incident, which is the same
3950 action as happens when Exim is looking up a host name and its aliases.
3951 Previously in this situation something unpredictable would happen;
3952 sometimes it was "internal error: store_reset failed".
3954 18. If a server dropped the connection unexpectedly when an Exim client was
3955 using GnuTLS and trying to read a response, the client delivery process
3956 crashed while trying to generate an error log message.
3958 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3959 string, for example:
3961 warn message = H1: something\nH2: something
3963 the text was added as a single header line from Exim's point of view
3964 though it ended up OK in the delivered message. However, searching for the
3965 second and subsequent header lines using $h_h2: did not work. This has been
3966 fixed. Similarly, if a system filter added multiple headers in this way,
3967 the routers could not see them.
3969 20. Expanded the error message when iplsearch is called with an invalid key to
3970 suggest using net-iplsearch in a host list.
3972 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3973 ACLs are no longer actually imposed (and a message to that effect is
3976 22. If a "gecos" field in a passwd entry contained escaped characters, in
3977 particular, if it contained a \" sequence, Exim got it wrong when building
3978 a From: or a Sender: header from that name. A second bug also caused
3979 incorrect handling when an unquoted " was present following a character
3980 that needed quoting.
3982 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3983 was not being matched caselessly.
3985 24. Arranged for all hyphens in the exim.8 source to be escaped with
3988 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3989 properly. Recipient callout cache records were still being keyed to include
3990 the sender, even when use_sender was set false. This led to far more
3991 callouts that were necessary. The sender is no longer included in the key
3992 when use_sender is false.
3994 26. Added "control = submission" modifier to ACLs.
3996 27. Added the ${base62d: operator to decode base 62 numbers.
3998 28. dnsdb lookups can now access SRV records.
4000 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4001 the configuration file.
4003 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4004 (-v) mode. This makes the output for a verbose queue run more intelligible.
4006 31. Added a use_postmaster feature to recipient callouts.
4008 32. Added the $body_zerocount variable, containing the number of binary zero
4009 bytes in the message body.
4011 33. The time of last modification of the "new" subdirectory is now used as the
4012 "mailbox time last read" when there is a quota error for a maildir
4015 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4017 35. Added +ignore_unknown as a special item in host lists.
4019 36. Code for decoding IPv6 addresses in host lists is now included, even if
4020 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4021 address was recognized as an IP address, but was then not correctly decoded
4022 into binary, causing unexpected and incorrect effects when compared with
4029 1. Very minor rewording of debugging text in manualroute to say "list of
4030 hosts" instead of "hostlist".
4032 2. If verify=header_syntax was set, and a header line with an unqualified
4033 address (no domain) and a large number of spaces between the end of the
4034 name and the colon was received, the reception process suffered a buffer
4035 overflow, and (when I tested it) crashed. This was caused by some obsolete
4036 code that should have been removed. The fix is to remove it!
4038 3. When running in the test harness, delay a bit after writing a bounce
4039 message to get a bit more predictability in the log output.
4041 4. Added a call to search_tidyup() just before forking a reception process. In
4042 theory, someone could use a lookup in the expansion of smtp_accept_max_
4043 per_host which, without the tidyup, could leave open a database connection.
4045 5. Added the variables $recipient_data and $sender_data which get set from a
4046 lookup success in an ACL "recipients" or "senders" condition, or a router
4047 "senders" option, similar to $domain_data and $local_part_data.
4049 6. Moved the writing of debug_print from before to after the "senders" test
4052 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4053 problems for message scanning, either using a data ACL, or using
4054 local_scan() because the Received: header was not generated till after they
4055 were called (in order to set the time as the time of reception completion).
4056 I have revised the way this works. The header is now generated after the
4057 body is received, but before the ACL or local_scan() are called. After they
4058 are run, the timestamp in the header is updated.
4064 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4065 before starting a queue runner without re-exec. This happened only when
4066 deliver_drop_privilege was set or when the Exim user was set to root. The
4067 effect of the bug was that timeouts during subsequent deliveries caused
4068 crashes instead of being properly handled. The handler is now left at its
4069 default (and expected) setting.
4071 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4072 message, again when deliver_drop_privilege is set or Exim is run as root.
4073 The bug described in (1) was not present in this case, but the tidying up
4074 of the other signals was missing. I have made the two cases consistent.
4076 3. The ignore_target_hosts setting on a manualroute router was being ignored
4077 for hosts that were looked up using the /MX notation.
4079 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4082 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4083 operated on the sender address. After changing the $sender_address to <>
4084 for the sender address verify, Exim was re-instated it as the original
4085 (before rewriting) address, but remembering that it had rewritten it, so it
4086 wasn't rewriting it again. This bug also had the effect of breaking the
4087 sender address verification caching when the sender address was rewritten.
4089 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4090 This has been changed so that if the ip literal address matches
4091 ignore_target_hosts, the router declines.
4093 7. Added expansion conditions match_domain, match_address, and match_local_
4094 part (NOT match_host).
4096 8. The placeholder for the Received: header didn't have a length field set.
4098 9. Added code to Exim itself and to exim_lock to test for a specific race
4099 condition that could lead to file corruption when using MBX delivery. The
4100 issue is with the lockfile that is created in /tmp. If this file is removed
4101 after a process has opened it but before that process has acquired a lock,
4102 there is the potential for a second process to recreate the file and also
4103 acquire a lock. This could lead to two Exim processes writing to the file
4104 at the same time. The added code performs the same test as UW imapd; it
4105 checks after acquiring the lock that its file descriptor still refers to
4106 the same named file.
4108 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4109 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4112 11. Added the smtp_active_hostname option. If used, this will typically be made
4113 to depend on the incoming interface address. Because $interface_address is
4114 not set up until the daemon has forked a reception process, error responses
4115 that can happen earlier (such as "too many connections") no longer contain
4118 12. If an expansion in a condition on a "warn" statement fails because a lookup
4119 defers, the "warn" statement is abandoned, and the next ACL statement is
4120 processed. Previously this caused the whole ACL to be aborted.
4122 13. Added the iplsearch lookup type.
4124 14. Added ident_timeout as a log selector.
4126 15. Added tls_certificate_verified as a log selector.
4128 16. Added a global option tls_require_ciphers (compare the smtp transport
4129 option of the same name). This controls incoming TLS connections.
4131 17. I finally figured out how to make tls_require_ciphers do a similar thing
4132 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4133 before starting the TLS session.
4135 18. Tabs are now shown as \t in -bP output.
4137 19. If the log selector return_path_on_delivery was set, Exim crashed when
4138 bouncing a message because it had too many Received: header lines.
4140 20. If two routers both had headers_remove settings, and the first one included
4141 a superfluous trailing colon, the final name in the first list and the
4142 first name in the second list were incorrectly joined into one item (with a
4143 colon in the middle).
4149 1. Added -C and -D options to the exinext utility, mainly to make it easier
4150 to include in the automated testing, but these could be helpful when
4151 multiple configurations are in use.
4153 2. The exinext utility was not formatting the output nicely when there was
4154 an alternate port involved in the retry record key, nor when there was a
4155 message id as well (for retries that were specific to a specific message
4156 and a specific host). It was also confused by IPv6 addresses, because of
4157 the additional colons they contain. I have fixed the IPv4 problem, and
4158 patched it up to do a reasonable job for IPv6.
4160 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4161 delivery, the log line now contains "pipelined" if PIPELINING was used.
4163 4. An SMTP transport process used to panic and die if the bind() call to set
4164 an explicit outgoing interface failed. This has been changed; it is now
4165 treated in the same way as a connect() failure.
4167 5. A reference to $sender_host_name in the part of a conditional expansion
4168 that was being skipped was still causing a DNS lookup. This no longer
4171 6. The def: expansion condition was not recognizing references to header lines
4172 that used bh_ and bheader_.
4174 7. Added the _cache feature to named lists.
4176 8. The code for checking quota_filecount in the appendfile transport was
4177 allowing one more file than it should have been.
4179 9. For compatibility with Sendmail, the command line option
4187 and sets the incoming protocol and host name (for trusted callers). The
4188 host name and its colon can be omitted when only the protocol is to be set.
4189 Note the Exim already has two private options, -pd and -ps, that refer to
4190 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4191 "s", but I don't think that's a major issue.
4193 10. A number of refactoring changes to the code, none of which should affect
4196 (a) The number of logging options was getting close to filling up the
4197 32-bit word that was used as a bit map. I have split them into two classes:
4198 those that are passed in the argument to log_write(), and those that are
4199 only ever tested independently outside of that function. These are now in
4200 separate 32-bit words, so there is plenty of room for expansion again.
4201 There is no change in the user interface or the logging behaviour.
4203 (b) When building, for example, log lines, the code previously used a
4204 macro that called string_cat() twice, in order to add two strings. This is
4205 not really sufficiently general. Furthermore, there was one instance where
4206 it was actually wrong because one of the argument was used twice, and in
4207 one call a function was used. (As it happened, calling the function twice
4208 did not affect the overall behaviour.) The macro has been replaced by a
4209 function that can join an arbitrary number of extra strings onto a growing
4212 (c) The code for expansion conditions now uses a table and a binary chop
4213 instead of a serial search (which was left over from when there were very
4214 few conditions). Also, it now recognizes conditions like "pam" even when
4215 the relevant support is not compiled in: a suitably worded error message is
4216 given if an attempt is made to use such a condition.
4218 11. Added ${time_interval:xxxxx}.
4220 12. A bug was causing one of the ddress fields not to be passed back correctly
4221 from remote delivery subprocesses. The field in question was not being
4222 subsequently used, so this caused to problems in practice.
4224 13. Added new log selectors queue_time and deliver_time.
4226 14. Might have fixed a bug in maildirsizefile handling that threw up
4227 "unexpected character" debug warnings, and recalculated the data
4228 unnecessarily. In any case, I expanded the warning message to give more
4231 15. Added the message "Restricted characters in address" to the statements in
4232 the default ACL that block characters like @ and % in local parts.
4234 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4235 Three changes have been made:
4237 (a) There was a serious bug; a negative response to MAIL caused the whole
4238 recipient domain to be cached as invalid, thereby blocking all messages
4239 to all local parts at the same domain, from all senders. This bug has
4240 been fixed. The domain is no longer cached after a negative response to
4241 MAIL if the sender used is not empty.
4243 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4246 (c) A new callout option, "use_sender" has been added for people who want
4247 the modified behaviour.
4253 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4256 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4257 indeed breaks things for older releases.
4259 3. Added additional logging to the case where there is a problem reading data
4260 from a filter that is running in a subprocess using a pipe, in order to
4261 try to track down a specific problem.
4263 4. Testing facility fudge: when running in the test harness and attempting
4264 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4265 getting "No route to host". Convert this to a timeout.
4267 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4270 6. Some OS don't have socklen_t but use size_t instead. This affects the
4271 fifth argument of getsockopt() amongst other things. This is now
4272 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4273 can be set for individual OS. I have set it for SunOS5, OSF1, and
4274 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4275 some earlier ones do not.
4277 7. Change 4.30/15 was not doing the test caselessly.
4279 8. The standard form for an IPv6 address literal was being rejected by address
4280 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4281 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4282 this, as well as the form without the "IPv6" on the front (but only when
4283 address literals are enabled, of course).
4285 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4287 10. Exim crashed if a message with an empty sender address specified by -f
4288 encountered a router with an errors_to setting. This could be provoked only
4289 by a command such as
4293 where an empty string was supplied; "<>" did not hit this bug.
4295 11. Installed PCRE release 4.5.
4297 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4298 remained set. It is now erased.
4300 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4301 times from message ids (which are base 36 rather than the normal 62).
4303 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4304 were being counted as actual protocol errors, and logged if the log
4305 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4306 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4307 and DATA following a set of rejected RCPTs do not count as protocol errors.
4308 In other words, Exim assumes they were pipelined, though this may not
4309 actually be the case. Of course, in all cases the client gets an
4310 appropriate error code.
4312 15. If a lookup fails in an ACL condition, a message about the failure may
4313 be available; it is used if testing the ACL cannot continue, because most
4314 such messages specify what the cause of the deferral is. However, some
4315 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4316 that caused an old message to be retained and used if a later statement
4317 caused a defer, replacing the real cause of the deferral.
4319 16. If an IP address had so many PTR records that the DNS lookup buffer
4320 was not large enough to hold them, Exim could crash while trying to process
4321 the truncated data. It now detects and logs this case.
4323 17. Further to 4.21/58, another change has been made: if (and only if) the
4324 first line of a message (the first header line) ends with CRLF, a bare LF
4325 in a subsequent header line has a space inserted after it, so as not to
4326 terminate the header.
4328 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4329 unnecessarily, used atoi() instead of strtol(), and didn't check the
4330 termination when getting file sizes from file names by regex.
4332 19. Completely re-implemented the support for maildirsize files, in the light
4333 of a number of problems with the previous contributed implementation
4334 (4.30/29). In particular:
4336 . If the quota is zero, the maildirsize file is maintained, but no quota is
4339 . If the maildir directory does not exist, it is created before any attempt
4340 to write a maildirsize file.
4342 . The quota value in the file is just a cache; if the quota is changed in
4343 the transport, the new value overrides.
4345 . A regular expression is available for excluding directories from the
4348 20. The autoreply transport checks the characters in options that define the
4349 message's headers; it allows continued headers, but it was checking with
4350 isspace() after an embedded newline instead of explicitly looking for a
4353 21. If all the "regular" hosts to which an address was routed had passed their
4354 expiry times, and had not reached their retry times, the address was
4355 bounced, even if fallback hosts were defined. Now Exim should go on to try
4358 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4359 equivalent code in the SMTP transport. Some hosts send humungous responses
4360 to HELO/EHLO, more than 1024 it seems.
4362 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4363 gives compiler warnings in some environments. I've now done it "properly",
4366 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4367 (because of problems with the built-in one) was declared to return uschar *
4368 instead of char *, causing compiler failure.
4370 25. Fixed a file descriptor leak when processing alias/forward files.
4372 26. Fixed a minor format string issue in dbfn.c.
4374 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4376 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4377 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4380 29. When a sender address is verified, it is cached, to save repeating the test
4381 when there is more than one recipient in a message. However, when the
4382 verification involves a callout, it is possible for different callout
4383 options to be set for different recipients. It is too complicated to keep
4384 track of this in the cache, so now Exim always runs a verification when a
4385 callout is required, relying on the callout cache for the optimization.
4386 The overhead is duplication of the address routing, but this should not be
4389 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4390 to be verified with callout=postmaster, and the main callout worked but the
4391 postmaster check failed, the verification correctly failed. However, if a
4392 subsequent RCPT command asked for sender verification *without* the
4393 postmaster check, incorrect caching caused this verification also to fail,
4396 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4397 it was not caching the DNS options (qualify_single, search_parents) that
4398 were used when the lookup failed. A subsequent lookup with different
4399 options therefore always gave the same answer, though there were cases
4400 where it should not have. (Example: a "domains = !$mx_any" option on a
4401 dnslookup router: the "domains" option is always processed without any
4402 widening, but the router might have qualify_single set.) Now Exim uses the
4403 cached value only when the same options are set.
4405 32. Added John Jetmore's "exipick" utility to the distribution.
4407 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4408 than a timeout (e.g. a certificate is required, and is not provided), an
4409 Exim server now closes the connection immediately. Previously it waited for
4410 the client to close - but if the client is SSL, it seems that they each
4411 wait for each other, leading to a delay before one of them times out.
4413 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4414 maintained 0.8.x compatibility because I don't think many are using it, and
4415 it is clearly obsolete.
4417 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4420 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4421 client certificate was expired. A simple patch fixes this, though I don't
4422 understand the full logic of why the verify callback is called multiple
4425 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4426 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4427 which causes problems with some clients (such as the Certicom SSL Plus
4428 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4429 disables the coutermeasure allowing Eudora to connect."
4431 38. Exim was not checking that a write() to a log file succeeded. This could
4432 lead to Bad Things if a log got too big, in particular if it hit a file
4433 size limit. Exim now panics and dies if it cannot write to a log file, just
4434 as it does if it cannot open a log file.
4436 39. Modified OS/Makefile-Linux so that it now contains
4438 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4440 The two -D definitions ensure that Exim is compiled with large file
4441 support, which makes it possible to handle log files that are bigger than
4444 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4445 instance) a domain was checked against a named list that involved a lookup,
4446 causing $domain_data to be set, then another domain was checked against the
4447 same list, then the first domain was re-checked, the value of $domain_data
4448 after the final check could be wrong. In particular, if the second check
4449 failed, it could be set empty. This bug probably also applied to
4452 41. The strip_trailing_dot option was not being applied to the address given
4453 with the -f command-line option.
4455 42. The code for reading a message's header from the spool was incrementing
4456 $received_count, but never initializing it. This meant that the value was
4457 incorrect (doubled) while delivering a message in the same process in which
4458 it was received. In the most common configuration of Exim, this never
4459 happens - a fresh exec is done - but it can happen when
4460 deliver_drop_privilege is set.
4462 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4463 it now includes up to 150 characters of the unexpected data in the log
4466 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4467 and building data strings. The size of both of these buffers was 10 000
4468 bytes - far larger than anybody would *ever* want, thought I. Needless to
4469 say, somebody hit the limit. I have increased the maximum line length to
4470 20 000 and the maximum data length of concatenated lines to 100 000. I have
4471 also fixed two bugs, because there was no checking on these buffers. Tsk,
4472 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4473 buffer is too small.
4475 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4476 lsearch lookups. Now it does.
4478 46. When parsing a route_list item in a manualroute router, a fixed-length
4479 buffer was used for the list of hosts. I made this 1024 bytes long,
4480 thinking that nobody would ever have a list of hosts that long. Wrong.
4481 Somebody had a whole pile of complicated expansion conditions, and the
4482 string was silently truncated, leading to an expansion error. It turns out
4483 that it is easier to change to an unlimited length (owing to other changes
4484 that have happened since this code was originally written) than to build
4485 structure for giving a limitation error. The length of the item that
4486 expands into the list of hosts is now unlimited.
4488 47. The lsearch lookup could not handle data where the length of text line was
4489 more than 4095 characters. Such lines were truncated, leading to shortened
4490 data being returned. It should now handle lines of any length.
4492 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4493 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4496 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4497 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4498 no longer happens in comment lines. A list of replacements is now placed
4499 at the head of all of the source files, except those whose only change is
4500 to replace PERL_COMMAND in the very first #! line.
4502 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4503 messages on the queue, with a bottom-up merge sort, using code contributed
4504 by Michael Haardt. This should make operations like -bp somewhat faster on
4505 large queues. It won't affect queue runners, except when queue_run_in_order
4508 51. Installed eximstats 1.31 in the distribution.
4510 52. Added support for SRV lookups to the dnslookup router.
4512 53. If an ACL referred to $message_body or $message_body_end, the value was not
4513 reset for any messages that followed in the same SMTP session.
4515 54. The store-handling optimization for building very long strings was not
4516 differentiating between the different store pools. I don't think this
4517 actually made any difference in practice, but I've tidied it.
4519 55. While running the routers to verify a sender address, $sender_address
4520 was still set to the sender address. This is wrong, because when routing to
4521 send a bounce to the sender, it would be empty. Therefore, I have changed
4522 it so that, while verifying a sender address, $sender_address is set to <>.
4523 (There is no change to what happens when verifying a recipient address.)
4525 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4526 target A or AAAA records (if not already returned) without resetting the
4527 qualify_single or search_parents options of the DNS resolver. These are
4528 inappropriate in this case because the targets of MX and SRV records must
4529 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4530 target that, when qualified, matched something in the local domain. These
4531 two options are now turned off when doing these lookups.
4533 57. It seems that at least some releases of Reiserfs (which does not have the
4534 concept of a fixed number of inodes) returns zero and not -1 for the
4535 number of available inodes. This interacted badly with check_spool_inodes,
4536 which assumed that -1 was the "no such thing" setting. What I have done is
4537 to check that the total number of inodes is greater than zero before doing
4538 the test of how many are available.
4540 58. When a "warn" ACL statement has a log_message modifier, the message is
4541 remembered, and not repeated. This is to avoid a lot of repetition when a
4542 message has many recipients that cause the same warning to be written.
4543 Howewer, Exim was preserving the list of already written lines for an
4544 entire SMTP session, which doesn't seem right. The memory is now reset if a
4545 new message is started.
4547 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4548 debugging output unless log_rewrite was also set.
4550 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4551 of (int)(handle) when we know that handle contains (void *)(-1).
4553 61. The Exim daemon panic-logs an error return when it closes the incoming
4554 connection. However "connection reset by peer" seems to be common, and
4555 isn't really an error worthy of noting specially, so that particular error
4558 62. When Exim is trying to find all the local interfaces, it used to panic and
4559 die if the ioctl to get the interface flags failed. However, it seems that
4560 on at least one OS (Solaris 9) it is possible to have an interface that is
4561 included in the list of interfaces, but for which you get a failure error
4562 for this call. This happens when the interface is not "plumbed" into a
4563 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4564 failure of the "get flags" call assumes that the interface is down.
4566 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4567 makes life easier for people who are doing arithmetic on fields extracted
4568 from dates, where you often get leading zeros that should not be
4569 interpreted as octal.
4571 64. Added qualify_domain to the redirect router, to override the global
4574 65. If a pathologically long header line contained very many addresses (the
4575 report of this problem mentioned 10 000) and each of them was rewritten,
4576 Exim could use up a very large amount of memory. (It kept on making new
4577 copies of the header line as it rewrote, and never released the old ones.)
4578 At the expense of a bit more processing, the header rewriting function has
4579 been changed so that it no longer eats memory in this way.
4581 66. The generation of the Received: header has been moved from the time that a
4582 message starts to be received, to the time that it finishes. The timestamp
4583 in the Received: header should now be very close to that of the <= log
4584 line. There are two side-effects of this change:
4586 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4587 logged header lines no longer include the local Received: line, because
4588 it has not yet been created. The same applies to a copy of the message
4589 that is returned to a non-SMTP sender when a message is rejected.
4591 (b) When a filter file is tested using -bf, no additional Received: header
4592 is added to the test message. After some thought, I decided that this
4595 This change does not affect the value of $received_for. It is still set
4596 after address rewriting, but before local_scan() is called.
4598 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4600 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4601 gave an unhelpful panic error message, and a defer error. I have managed to
4602 change this behaviour so that it now rejects any supplied certificate,
4603 which seems right, as the list of acceptable certificates is empty.
4605 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4606 gave an unhelpful defer error. I have not managed to make this reject any
4607 supplied certificates, but the error message it gives is "no certificate
4608 supplied", which is not helpful.
4610 70. exigrep's output now also includes lines that are not associated with any
4611 message, but which match the given pattern. Implemented by a patch from
4612 Martin Sluka, which also tidied up the Perl a bit.
4614 71. Recipient callout verification, like sender verification, was using <> in
4615 the MAIL FROM command. This isn't really the right thing, since the actual
4616 sender may affect whether the remote host accepts the recipient or not. I
4617 have changed it to use the actual sender in the callout; this means that
4618 the cache record is now keyed on a recipient/sender pair, not just the
4619 recipient address. There doesn't seem to be a real danger of callout loops,
4620 since a callout by the remote host to check the sender would use <>.
4621 [SEE ABOVE: changed after hitting problems.]
4623 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4624 temporary errors. However, in the case of such a code being given after
4625 the end of a data transmission (i.e. after ".") Exim was failing to write
4626 a retry record for the message. (Yes, there was some broken host that was
4627 actually sending 8xx at this point.)
4629 73. An unknown lookup type in a host list could cause Exim to panic-die when
4630 the list was checked. (An example that provoked this was putting <; in the
4631 middle of a list instead of at the start.) If this happened during a DATA
4632 ACL check, a -D file could be left lying around. This kind of configuration
4633 error no longer causes Exim to die; instead it causes a defer errror. The
4634 incident is still logged to the main and panic logs.
4636 74. Buglet left over from Exim 3 conversion. The message "too many messages
4637 in one connection" was written to the rejectlog but not the mainlog, except
4638 when address rewriting (yes!) was being logged.
4640 75. Added write_rejectlog option.
4642 76. When a system filter was run not as root (that is, when system_filter_user
4643 was set), the values of the $n variables were not being returned to the
4644 main process; thus, they were not subsequently available in the $sn
4647 77. Added +return_path_on_delivery log selector.
4649 78. A connection timeout was being treated differently from recipients deferred
4650 when testing hosts_max_try with a message that was older than the host's
4651 retry timeout. (The host should not be counted, thus allowing all hosts to
4652 be tried at least once before bouncing.) This may have been the cause of an
4653 occasionally reported bug whereby a message would remain on the queue
4654 longer than the retry timeout, but would be bounced if a delivery was
4655 forced. I say "may" because I never totally pinned down the problem;
4656 setting up timeout/retry tests is difficult. See also the next item.
4658 79. The ultimate address timeout was not being applied to errors that involved
4659 a combination of host plus message (for example, a timeout on a MAIL
4660 command). When an address resolved to a number of possible hosts, and they
4661 were not all tried for each delivery (e.g. because of hosts_max_try), a
4662 message could remain on the queue longer than the retry timeout.
4664 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4667 81. Fixed an obscure SMTP outgoing bug which required at least the following
4668 conditions: (a) there was another message waiting for the same server;
4669 (b) the server returned 5xx to all RCPT commands in the first message so
4670 that the message was not completed; (c) the server dropped the connection
4671 or gave a negative response to the RSET that Exim sends to abort the
4672 transaction. The observed case was a dropped connection after DATA that had
4673 been sent in pipelining mode. That is, the server had advertised PIPELINING
4674 but was not implementing it correctly. The effect of the bug was incorrect
4675 behaviour, such as trying another host, and this could lead to a crash.
4681 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4682 and daemon.c were passed as pointers to ints; they should have been
4683 pointers to socklen_t variables (which are typically unsigned ints).
4685 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4688 3. Fixed a really odd bug that affected only the testing scheme; patching a
4689 certain fixed string in the binary changed the value of another string that
4690 happened to be identical to the end of the original first string.
4692 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4693 name", it returns that address as the IP address. On some operating
4694 systems (e.g. Solaris), it also passes back the IP address string as the
4695 "host name". However, on others (e.g. Linux), it passes back an empty
4696 string. Exim wasn't checking for this, and was changing the host name to an
4697 empty string, assuming it had been canonicized.
4699 5. Although rare, it is permitted to have more than one PTR record for a given
4700 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4701 all the names associated with an address, because they do in Solaris.
4702 However, it seems that they do not in Linux for data that comes from the
4703 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4704 I found this out when I moved to a new Linux workstation and tried to run
4705 the Exim test suite.
4707 To get round this problem I have changed the code so that it now does its
4708 own call to the DNS to look up PTR records when searching for a host name.
4709 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4710 addresses that are only in /etc/hosts are still found.
4712 This behaviour is, however, controlled by an option called host_lookup_
4713 order, which defaults to "bydns:byaddr". If people want to use the other
4714 order, or indeed, just use one or the other means of lookup, they can
4715 specify it in this variable.
4717 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4718 some operating systems, this comes back from gethostbyaddr() as an empty
4719 string, and this is what Exim used to test for. However, it seems that in
4720 other systems, "." is yielded. Exim now tests for this case too.
4722 7. The values of check_spool_space and check_log_space are now held internally
4723 as a number of kilobytes instead of an absolute number of bytes. If a
4724 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4725 kilobyte. This means that much larger values can be stored.
4727 8. Exim monitor: an attempt to get the action menu when not actually pointing
4728 at a message produces an empty menu entitled "No message selected". This
4729 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4730 no entries in it ("Shell widget menu has zero width and/or height"). So I
4731 have added a single, blank menu entry in this case.
4733 9. Added ${quote_local_part.
4735 10. MIME decoding is now applied to the contents of Subject: header lines when
4738 11. Now that a reference to $sender_host_address automatically causes a reverse
4739 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4740 host lookup before query-style lookups in lists that might use this
4741 variable. This has therefore been abolished, and the "net-" prefix is no
4742 longer necessary for query-style lookups.
4744 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4745 have been a typo for LFLAGS, so it has been changed.
4747 13. The install script calls Exim with "-C /dev/null" in order to find the
4748 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4749 to be output. Howeve, since Exim outputs its version number before the
4750 error, it didn't break the script. It just looked ugly. I fixed this by
4751 always allowing "-C /dev/null" if the caller is root.
4753 14. Ignore overlarge ACL variable number when reading spool file - insurance
4754 against a later release with more variables having written the file.
4756 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4757 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4758 this, as well as the form without the "IPv6" on the front.
4760 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4761 OS/Makefile-Darwin file.
4763 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4764 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4766 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4767 variable. However, some OS use ints for inodes. I've added cast to long int
4768 to get rid of the compiler warning.
4770 19. I had forgotten to lock out "/../" in configuration file names when
4771 ALT_CONFIG_PREFIX was set.
4773 20. Routers used for verification do not need to specify transports. However,
4774 if such a router generated a host list, and callout was configured, Exim
4775 crashed, because it could not find a port number from the (non-existent)
4776 transport. It now assumes port 25 in this circumstance.
4778 21. Added the -t option to exigrep.
4780 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4781 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4782 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4783 them is set, LOOKUP_LSEARCH is forced.
4785 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4787 24. Added sender and host information to the "rejected by local_scan()" log
4788 line; previously there was no indication of these.
4790 25. Added .include_if_exists.
4792 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4793 Linux. It turns out that not all file systems support this. Apparently some
4794 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4795 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4796 not supported on the file descriptor, is now ignored when Exim is trying to
4797 sync a directory. This applies only to Linux.
4799 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4801 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4802 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4805 29. Added support for maildirsize files from supplied patch (modified a bit).
4807 30. The use of :fail: followed by an empty string could lead Exim to respond to
4808 sender verification failures with (e.g.):
4810 550 Verification failed for <xxx>
4811 550 Sender verify failed
4813 where the first response line was missing the '-' that indicates it is not
4814 the final line of the response.
4816 31. The loop for finding the name of the user that called Exim had a hardwired
4817 limit of 10; it now uses the value of finduser_retries, which is used for
4818 all other user lookups.
4820 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4823 33. Exim was neglecting to zero errno before one call of strtol() when
4824 expanding a string and expecting an integer value. On some systems this
4825 resulted in spurious "integer overflow" errors. Also, it was casting the
4826 result into an int without checking.
4828 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4829 did not work. The code looks as if it has *never* worked, though it appears
4830 to have been documented since at least releast 1.62. I have made it work.
4832 35. The "timeout_DNS" error in retry rules, also documented since at least
4833 1.62, also never worked. As it isn't clear exactly what this means, and
4834 clearly it isn't a major issue, I have abolished the feature by treating it
4835 as "timeout", and writing a warning to the main and panic logs.
4837 36. The display of retry rules for -brt wasn't always showing the error code
4840 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4841 timeout_connect_A, timeout_connect_MX.
4843 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4844 to the empty sender.
4846 39. The daemon was not analysing the content of -oX till after it had closed
4847 stderr and disconnected from the controlling terminal. This meant that any
4848 syntax errors were only noted on the panic log, and the return code from
4849 the command was 0. By re-arranging the code a little, I've made the
4850 decoding happen first, so such errors now appear on stderr, and the return
4851 code is 1. However, the actual setting up of the sockets still happens in
4852 the disconnected process, so errors there are still only recorded on the
4855 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4856 connections (as happens on some IP stacks) was logged at start up time as
4857 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4858 it from "IPv6 and IPv4", which means that two separate sockets are being
4861 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4862 says whether AF_INET or AF_INET6 was passed as an argument.
4864 42. Exiwhat output was messed up when time zones were included in log
4867 43. Exiwhat now gives more information about the daemon's listening ports,
4868 and whether -tls-on-connect was used.
4870 44. The "port" option of the smtp transport is now expanded.
4872 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4873 silently ignored. Now an error message is written to the main and panic
4876 46. There's a new ACL modifier called "logwrite" which writes to a log file
4877 as soon as it is encountered.
4879 47. Added $local_user_uid and $local_user_gid at routing time.
4881 48. Exim crashed when trying to verify a sender address that was being
4884 49. Exim was recognizing only a space character after ".include". It now also
4885 recognizes a tab character.
4887 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4888 extracting the relevant information from the specification. The man page no
4889 longer contains scrambled data for the -d option, and I've added a section
4890 at the front about calling Exim under different names.
4892 51. Added "extra_headers" argument to the "mail" command in filter files.
4894 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4897 53. Installed eximstats 1.29.
4899 54. Added transport_filter_timeout as a generic transport option.
4901 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4902 Cc: header lines. This was required by RFC 822, but it not required by RFC
4905 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4906 incoming messages that did not have them. Now it does so only if the
4907 message originates locally, that is, if there is no associated remote host
4908 address. When Resent- header lines are present, this applies to the Resent-
4909 lines rather than the non-Resent- lines.
4911 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4912 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4914 58. Messages for configuration errors now include the name of the main
4915 configuration file - useful now that there may be more than one file in a
4916 list (.included file names were always shown).
4918 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4919 for those rare installations that do not start the daemon as root or run it
4920 setuid root. I've cut out the call to initgroups() if the daemon is not
4923 60. The Exim user and group can now be bound into the binary as text strings
4924 that are looked up at the start of Exim's processing.
4926 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4928 62. Added $mailstore_basename variable.
4930 63. Installed patch to sieve.c from Michael Haardt.
4932 64. When Exim failed to open the panic log after failing to open the main log,
4933 the original message it was trying to log was written to stderr and debug
4934 output, but if they were not available (the usual case in production), it
4935 was lost. Now it is written to syslog before the two lines that record the
4936 failures to open the logs.
4938 65. Users' Exim filters run in subprocesses under the user's uid. It is
4939 possible for a "deliver" command or an alias in a "personal" command to
4940 provoke an address rewrite. If logging of address rewriting is configured,
4941 this fails because the process is not running as root or exim. There may be
4942 a better way of dealing with this, but for the moment (because 4.30 needs
4943 to be released), I have disabled address rewrite logging when running a
4944 filter in a non-root, non-exim process.
4950 1. The buildconfig auxiliary program wasn't quoting the value set for
4951 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4952 not defined. This bug was masked in 4.22 by the effect that was fixed in
4955 2. Some messages that were rejected after a message id was allocated were
4956 shown as "incomplete" by exigrep. It no longer does this for messages that
4957 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4959 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4960 have allow_domain_literals set, the ID did not get logged in the <= line.
4961 Domain literals are now always recognized in Message-ID: header lines.
4963 4. The first argument for a ${extract expansion item is the key name or field
4964 number. Leading and trailing spaces in this item were not being ignored,
4965 causing some misleading effects.
4967 5. When deliver_drop_privilege was set, single queue runner processes started
4968 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4969 same command in the process it spins off) were not dropping privilege.
4971 6. When the daemon running as "exim" started a queue runner, it always
4972 re-executed Exim in the spun-off process. This is a waste of effort when
4973 deliver_drop_privilege is set. The new process now just calls the
4974 queue-runner function directly.
4980 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4983 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4984 ignored. Though the use of -oP was forcing the writing of a pid file, it
4985 was always written to the default place.
4987 3. If the message "no IP address found for host xxxx" is generated during
4988 incoming verification, it is now followed by identification of the incoming
4989 connection (so you can more easily find what provoked it).
4991 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4993 5. Added some features to "harden" Exim a bit more against certain attacks:
4995 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4996 be put in Local/Makefile. This is like the never_users runtime option,
4997 but it cannot be overridden. The default setting is "root".
4999 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5000 prefix string with which any file named in a -C command line option
5003 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5004 is retained for -C and -D only if the caller of Exim is root. Without
5005 it, the exim user may also use -C and -D and retain privilege.
5007 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5008 command line option is disabled.
5010 6. Macro names set by the -D option must start with an upper case letter, just
5011 like macro names defined in the configuration file.
5013 7. Added "dereference=" facility to LDAP.
5015 8. Two instances of the typo "uknown" in the source files are fixed.
5017 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5018 the Configure-Makefile script screwed up while processing it.
5020 10. Incorporated PCRE 4.4.
5022 11. The SMTP synchronization check was not operating right at the start of an
5023 SMTP session. For example, it could not catch a HELO sent before the client
5024 waited for the greeting. There is now a check for outstanding input at the
5025 point when the greeting is written. Because of the duplex, asynchronous
5026 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5027 way, but not yet received, when the check is performed.
5029 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5030 on TCP/IP sockets, because this apparently causes some broken clients to
5033 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5034 unchanged) from the Cygwin maintainer.
5036 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5037 was supported instead of testing for mbx. Effectively a typo.
5039 15. The spa authenticator server code was not checking that the input it
5040 received was valid base64.
5042 16. The debug output line for the "set" modifier in ACLs was not showing the
5043 name of the variable that was being set.
5045 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5047 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5048 lookup on its own. Something else was needed to trigger the lookup. For
5049 example, a match in host_lookup or the need for a host name in a host list.
5050 Now, if $sender_host_name is referenced and the host name has not yet been
5051 looked up, a lookup is performed. If the lookup fails, the variable remains
5052 empty, and $host_lookup_failed is set to "1".
5054 19. Added "eqi" as a case-independent comparison operator.
5056 20. The saslauthd authentication condition could segfault if neither service
5057 nor realm was specified.
5059 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5060 error message that was logged was misleading, and incoming SMTP
5061 connections were dropped. The message is now more accurate, and temporary
5062 errors are given to SMTP connections.
5064 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5065 (except RSET) with a 421 error, until QUIT is received. However, it was
5066 failing to send a response to QUIT.
5068 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5069 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5070 ACL is now run after the helo_try_verify_hosts code.
5072 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5073 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5074 case-independent, so other case variants are also recognized). Apparently
5075 some systems use these upper case variants.
5077 25. If more than two messages were waiting for the same host, and a transport
5078 filter was specified for the transport, Exim sent two messages over the
5079 same TCP/IP connection, and then failed with "socket operation on non-
5080 socket" when it tried to send the third.
5082 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5084 27. The extern definition of crypt16() in expand.c was not being excluded when
5085 the OS had its own crypt16() function.
5087 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5088 as a preferred synonym for return_size_limit, both as an option and as an
5091 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5093 30. Changed the default configuration ACL to relax the local part checking rule
5094 for addresses that are not in any local domains. For these addresses,
5095 slashes and pipe symbols are allowed within local parts, but the sequence
5096 /../ is explicitly forbidden.
5098 31. SPA server authentication was not clearing the challenge buffer before
5101 32. log_message in a "warn" ACL statement was writing to the reject log as
5102 well as to the main log, which contradicts the documentation and doesn't
5103 seem right (because no rejection is happening). So I have stopped it.
5105 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5106 However, I am unable to do any testing of this.
5108 34. Fixed an infelicity in the appendfile transport. When checking directories
5109 for a mailbox, to see if any needed to be created, it was accidentally
5110 using path names with one or more superfluous leading slashes; tracing
5111 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5113 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5114 contents are added to the log line that is written for every discarded
5115 recipient. (Previously a log_message setting was ignored.)
5117 36. The ${quote: operator now quotes the string if it is empty.
5119 37. The install script runs exim in order to find its version number. If for
5120 some reason other than non-existence or emptiness, which it checks, it
5121 could not run './exim', it was installing it with an empty version number,
5122 i.e. as "exim-". This error state is now caught, and the installation is
5125 38. An argument was missing from the function that creates an error message
5126 when Exim fails to connect to the socket for saslauthd authentication.
5127 This could cause Exim to crash, or give a corrupted message.
5129 39. Added isip, isip4, and isip6 to ${if conditions.
5131 40. The ACL variables $acl_xx are now saved with the message, and can be
5132 accessed later in routers, transports, and filters.
5134 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5135 strings in the file are not string-expanded.
5137 42. If a MAIL command specified a SIZE value that was too large to fit into an
5138 int variable, the check against message_size_limit failed. Such values are
5139 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5140 day this will have to be increased, but I don't think I want to be around
5141 when emails are that large.
5148 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5149 iconv() is not standard in FreeBSD.
5151 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5152 IPv6 enabled. The observed symptom was a segmentation fault on return from
5153 the function os_common_find_running_interfaces() in src/os.c.
5155 3. In the check_special_case() function in daemon.c I had used "errno" as an
5156 argument name, which causes warnings on some systems. This was basically a
5157 typo, since it was named "eno" in the comments!
5159 4. The code that waits for the clock to tick (at a resolution of some fraction
5160 of a second) so as to ensure message-id uniqueness was always waiting for
5161 at least one whole tick, when it could have waited for less. [This is
5162 almost certainly not relevant at current processor speeds, where it is
5163 unlikely to ever wait at all. But we try to future-proof.]
5165 5. The function that sleeps for a time interval that includes fractions of a
5166 second contained a race. It did not block SIGALRM between setting the
5167 timer, and suspending (a couple of lines later). If the interval was short
5168 and the sigsuspend() was delayed until after it had expired, the suspension
5169 never ended. On busy systems this could lead to processes getting stuck for
5172 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5173 process, before it forks any delivery processes. The open lookup caching
5174 mechanism meant that the open file or database connection was passed into
5175 the delivery process. The problem was that delivery processes always tidy
5176 up cached lookup data. This could cause a problem for the next delivery
5177 process started by the queue runner, because the external queue runner
5178 process does not know about the closure. So the next delivery process
5179 still has data in the lookup cache. In the case of a file lookup, there was
5180 no problem because closing a file descriptor in a subprocess doesn't affect
5181 the parent. However, if the lookup was caching a connection to a database,
5182 the connection was closed, and the second delivery process was likely to
5183 see errors such as "PGSQL: query failed: server closed the connection
5184 unexpectedly". The problem has been fixed by closing all cached lookups
5185 in a queue runner before running a delivery process.
5187 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5188 seem to have the "const" qualifier which it has on other OS. I've
5191 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5192 *of the same type* (client or server) with the same public name that an
5193 error should be diagnosed.
5195 9. When Exim looked up a host name for an IP address, but failed to find the
5196 original IP address when looking up the host name (a safety check), it
5197 output the message "<ip address> does not match any IP for NULL", which was
5198 confusing, to say the least. The bug was that the host name should have
5199 appeared instead of "NULL".
5201 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5202 user that is built into the binary, and the -C or -D options is used, root
5203 privilege is dropped before the configuration file is read. In addition,
5204 logging is switched to stderr instead of the normal log files. If the
5205 configuration then re-defines the Exim user, the unprivileged environment
5206 is probably not what is expected, so Exim logs a panic warning message (but
5209 However, if deliver_drop_privilege is set, the unprivileged state may well
5210 be exactly what is intended, so the warning has been cut out in that case,
5211 and Exim is allowed to try to write to its normal log files.
5217 1. smtp_return_error_details was not giving details for temporary sender
5218 or receiver verification errors.
5220 2. Diagnose a configuration error if two authenticators have the same public
5223 3. Exim used not to create the message log file for a message until the first
5224 delivery attempt. This could be confusing when incoming messages were held
5225 for policy or load reasons. The message log file is now created at the time
5226 the message is received, and an initial "Received" line is written to it.
5228 4. The automatically generated man page for command line options had a minor
5229 bug that caused no ill effects; however, a more serious problem was that
5230 the procedure for building the man page automatically didn't always
5231 operate. Consequently, release 4.20 contains an out-of-date version. This
5232 shouldn't happen again.
5234 5. When building Exim with embedded Perl support, the script that builds the
5235 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5236 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5238 6. The freeze_tell option was not being used for messages that were frozen on
5239 arrival, either by an ACL or by local_scan().
5241 7. Added the smtp_incomplete_transaction log selector.
5243 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5244 was accepting AUTH without a new EHLO.
5246 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5247 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5250 10. Logging of TCP/IP connections (when configured) now happens in the main
5251 daemon process instead of the child process, so that the TCP/IP connection
5252 count is more accurate (but it can never be perfect).
5254 11. The use of "drop" in a nested ACL was not being handled correctly in the
5255 outer ACL. Now, if condition failure induced by the nested "drop" causes
5256 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5257 or "require"), the connection is dropped.
5259 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5260 that yield "discard" can now be used with an "accept" or a "discard" verb,
5261 but an error is generated for any others (because I can't see a useful way
5262 to define what should happen).
5264 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5265 are now processed in the same way as lines in the Exim configuration file.
5266 In particular, continuation lines are supported.
5268 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5270 15. Added -ti meaning -t -i.
5272 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5273 domains, and warn by logging if others are found.
5275 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5276 returned from GIFCONF when Exim is trying to find the list of interfaces on
5277 a host. The code in os.c has been modified to copy each ifreq to an aligned
5278 structure in all cases.
5280 Also, in some cases, the returned ifreq's were being copied to a 'struct
5281 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5282 means the last couple of bytes of an IPv6 address could be chopped if the
5283 ifreq contained only a normal sockaddr (14 bytes storage).
5285 18. Named domain lists were not supported in the hosts_treat_as_local option.
5286 An entry such as +xxxx was not recognized, and was treated as a literal
5289 19. Ensure that header lines added by a DATA ACL are included in the reject log
5290 if the ACL subsequently rejects the message.
5292 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5293 MD5 (which is deprecated).
5295 21. When testing a filter file using -bf, Exim was writing a message when it
5296 took the sender from a "From " line in the message, but it was not doing so
5297 when it took $return_path from a Return-Path: header line. It now does.
5299 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5300 with a valid header line field name (a series of printing characters
5301 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5303 23. Changed "disc" in the source to "disk" to conform to the documentation and
5304 the book and for uniformity.
5306 24. Ignore Sendmail's -Ooption=value command line item.
5308 25. When execve() failed while trying to run a command in a pipe transport,
5309 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5310 could be confused with a return value of 69 from the command itself. This
5311 has been changed to 127, the value the shell returns if it is asked to run
5312 a non-existent command. The wording for the related log line suggests a
5313 non-existent command as the problem.
5315 26. If received_header_text expands to an empty string, do not add a Received:
5316 header line to the message. (Well, it adds a token one on the spool, but
5317 marks it "old" so that it doesn't get used or transmitted.)
5319 27. Installed eximstats 1.28 (addition of -nt option).
5321 28. There was no check for failure on the call to getsockname() in the daemon
5322 code. This can fail if there is a shortage of resources on the system, with
5323 ENOMEM, for example. A temporary error is now given on failure.
5325 29. Contrary to the C standard, it seems that in some environments, the
5326 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5327 program. Exim now does this explicitly; it affects the formatting of
5328 timestamps using strftime().
5330 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5331 complaints. I've now initialized all the variables, to avoid this.
5333 32. Header lines added by a system filter were not being "seen" during
5334 transport-time rewrites.
5336 33. The info_callback() function passed to OpenSSL is set up with type void
5337 (*)(SSL *, int, int), as described somewhere. However, when calling the
5338 function (actually a macro) that sets it up, the type void(*)() is
5339 expected. I've put in a cast to prevent warnings from picky compilers.
5341 34. If a DNS black list lookup found a CNAME record, but there were no A
5342 records associated with the domain it pointed at, Exim crashed.
5344 35. If a DNS black list lookup returned more than one A record, Exim ignored
5345 all but the first. It now scans all returned addresses if a particular IP
5346 value is being sought. In this situation, the contents of the
5347 $dnslist_value variable are a list of all the addresses, separated by a
5350 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5351 to accept a host name and all its aliases if the forward lookup for any of
5352 them yielded the IP address of the incoming connection. Now it accepts only
5353 those names whose forward lookup yields the correct IP address. Any other
5354 names are discarded. This closes a loophole whereby a rogue DNS
5355 administrator could create reverse DNS records to break through a
5356 wildcarded host restriction in an ACL.
5358 37. If a user filter or a system filter that ran in a subprocess used any of
5359 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5360 the wrong values were passed to the pipe command ($thisaddress had the
5361 value of $0, $0 had the value of $1, etc). This bug was introduced by
5362 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5364 38. Improved the line breaking for long SMTP error messages from ACLs.
5365 Previously, if there was no break point between 40 and 75 characters, Exim
5366 left the rest of the message alone. Two changes have been made: (a) I've
5367 reduced the minimum length to 35 characters; (b) if it can't find a break
5368 point between 35 and 75 characters, it looks ahead and uses the first one
5369 that it finds. This may give the occasional overlong line, but at least the
5370 remaining text gets split now.
5372 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5373 file descriptors might be low, and that setting 1000 would always raise it.
5374 It turns out that in some environments, the limit is already over 1000 and
5375 that lowering it causes trouble. So now Exim takes care not to decrease it.
5377 40. When delivering a message, the value of $return_path is set to $sender_
5378 address at the start of routing (routers may change the value). By an
5379 oversight, this default was not being set up when an address was tested by
5380 -bt or -bv, which affected the outcome if any router or filter referred to
5383 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5384 log only when "message" or "log_message" are set. However, if one of the
5385 conditions was an address verification, or a call to a nested ACL, the
5386 messages generated by the underlying test were being passed through. This
5387 no longer happens. The underlying message is available in $acl_verify_
5388 message for both "message" and "log_message" expansions, so it can be
5389 passed through if needed.
5391 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5392 new expansion $bh_ to give the encoded byte string without charset
5393 translation. Translation happens only if iconv() is available; HAVE_ICONV
5394 indicates this at build time. HEADERS_CHARSET gives the charset to
5395 translate to; headers_charset can change it in the configuration, and
5396 "headers charset" can change it in an individual filter file.
5398 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5399 that creates RFC 2047 encoded "words" labels them as that charset instead
5400 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5401 expansion operator; (ii) when Exim creates a From: line for a local
5402 message; (iii) when a header line is rewritten to include a "phrase" part.
5404 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5405 buggy, causing it to skip the first lines of messages whose message ID
5406 ended in 'D'. This would not have bitten before Exim release 4.14, because
5407 message IDs were unlikely to end in 'D' before then. The effect was to have
5408 incorrect size information for certain domains.
5410 45. #include "config.h" was missing at the start of the crypt16.c module. This
5411 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5414 46. If there was a timeout during a "random" callout check, Exim treated it as
5415 a failure of the random address, and carried on sending RSET and the real
5416 address. If the delay was just some slowness somewhere, the response to the
5417 original RCPT would be taken as a response to RSET and so on, causing
5418 mayhem of various kinds.
5420 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5421 when I implemented it. It didn't allow for the fact that some option values
5422 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5423 the right test for positive values.
5425 48. Domain names in DNS records are case-independent. Exim always looks them up
5426 in lower case. Some resolvers return domain names in exactly the case they
5427 appear in the zone file, that is, they may contain uppercase letters. Not
5428 all resolvers do this - some return always lower case. Exim was treating a
5429 change of case by a resolver as a change of domain, similar to a widening
5430 of a domain abbreviation. This triggered its re-routing code and so it was
5431 trying to route what was effectively the same domain again. This normally
5432 caused routing to fail (because the router wouldn't handle the domain
5433 twice). Now Exim checks for this case specially, and just changes the
5434 casing of the domain that it ultimately uses when it transmits the message
5437 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5440 50. If a filter generated a file delivery with a non-absolute name (possible if
5441 no home directory exists for the router), the forbid_file option was not
5444 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5445 the existing equality matching.
5447 52. Exim was using ints instead of ino_t variables in some places where it was
5448 dealing with inode numbers.
5450 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5451 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5452 called TMPDIR, and if it finds it is different, it changes its value.
5454 54. The smtp_printf() function is now made available to local_scan() so
5455 additional output lines can be written before returning. There is also an
5456 smtp_fflush() function to enable the detection of a dropped connection.
5457 The variables smtp_input and smtp_batched_input are exported to
5460 55. Changed the default runtime configuration: the message "Unknown user"
5461 has been removed from the ACL, and instead placed on the localuser router,
5462 using the cannot_route_message feature. This means that any verification
5463 failures that generate their own messages won't get overridden. Similarly,
5464 the "Unrouteable address" message that was in the ACL for unverifiable
5465 relay addresses has also been removed.
5467 56. Added hosts_avoid_esmtp to the smtp transport.
5469 57. The exicyclog script was not checking for the esoteric option
5470 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5471 will work only if exicyclog is run under the appropriate euid.
5473 58. Following a discussion on the list, the rules by which Exim recognises line
5474 endings on incoming messages have been changed. The -dropcr and drop_cr
5475 options are now no-ops, retained only for backwards compatibility. The
5476 following line terminators are recognized: LF CRLF CR. However, special
5477 processing applies to CR:
5479 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5480 nor a local message in the state where . is a terminator.
5482 (ii) If a bare CR is encountered in a header line, an extra space is added
5483 after the line terminator so as not to end the header. The reasoning
5484 behind this is that bare CRs in header lines are most likely either
5485 to be mistakes, or people trying to play silly games.
5487 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5488 was being incorrectly given as 18 bytes larger than it should have been.
5489 This is a VOB (very old bug).
5491 60. This may never have affected anything current, but just in case it has:
5492 When the local host is found other than at the start of a list of hosts,
5493 the local host, those with the same MX, and any that follow, are discarded.
5494 When the list in question was part of a longer list of hosts, the following
5495 hosts (not currently being processed) were also being discarded. This no
5496 longer happens. I'm not sure if this situation could ever has previously
5499 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5502 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5503 header. This is something that is recommended in a new Internet Draft, and
5504 is something that is documented as being done by Sendmail. There are two
5505 possible values. For messages generated by the autoreply transport, Exim
5508 Auto-Submitted: auto-replied
5510 whereas for all other generated messages (e.g. bounces) it adds
5512 Auto-Submitted: auto-generated
5514 63. The "personal" condition in filters now includes a test for the
5515 Auto-Submitted: header. If it contains the string "auto-" the message it
5516 not considered personal.
5518 64. Added rcpt_include_affixes as a generic transport option.
5520 65. Added queue_only_override (default true).
5522 66. Added the syslog_duplication option.
5524 67. If what should have been the first header line of a message consisted of
5525 a space followed by a colon, Exim was mis-interpreting it as a header line.
5526 It isn't of course - it is syntactically invalid and should therefore be
5527 treated as the start of the message body. The misbehaviour could have
5528 caused a number of strange effects, including loss of data in subsequent
5529 header lines, and spool format errors.
5531 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5532 client host had authenticated. This control can now be exercised by an ACL
5533 for more flexibility.
5535 69. By default, callouts do not happen when testing with -bh. There is now a
5536 variant, -bhc, which does actually run the callout code, including
5537 consulting and updating the callout cache.
5539 70. Added support for saslauthd authentication, courtesy of Alexander
5542 71. If statvfs() failed on the spool or log directories while checking their
5543 size for availability, Exim confusingly gave the error "space shortage".
5544 Furthermore, in debugging mode it crashed with a floating point exception.
5545 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5546 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5547 serious problem, Exim now writes to the main and panic logs when this
5548 happens, with details of the failure. It then refuses to accept the
5549 incoming message, giving the message "spool directory problem" or "log
5550 directory problem" with a 421 code for SMTP messages.
5552 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5553 0, 1, and 2 exist, because some OS complain for execs without them (see
5554 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5555 descriptors. However, the code omitted to check that the open succeeded,
5556 causing mysterious errors if for some reason the permissions on /dev/null
5557 got screwed. Now Exim writes a message to the main and panic logs, and
5558 bombs out if it can't open /dev/null.
5560 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5561 interact so that it is all more flexible. It is supposed to remain
5562 backwards compatible. Also added extra_local_interfaces.
5564 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5565 to bomb out with an assertion failure - to the client this appears as a
5566 connection drop. This problem occurs in the part of the code that was taken
5567 from the Samba project. Fortunately, the assertion is in a very simple
5568 function, so I have fixed this by reproducing the function inline in the
5569 one place where it is called, and arranging for authentication to fail
5570 instead of killing the process with assert().
5572 75. The SPA client code was not working when the server requested OEM rather
5573 than Unicode encoding.
5575 76. Added code to make require_files with a specific uid setting more usable in
5576 the case where statting the file as root fails - usually a non-root-mounted
5577 NFS file system. When this happens and the failure is EACCES, Exim now
5578 forks a subprocess and does the per-uid checking as the relevant uid.
5580 77. Added process_log_path.
5582 78. If log_file_path was not explicitly set, a setting of check_log_space or
5583 check_log_inodes was ignored.
5585 79. If a space check for the spool or log partitions fails, the incident is now
5586 logged. Of course, in the latter case the data may get lost...
5588 80. Added the %p formatting code to string_format() so that it can be used to
5589 print addresses in debug_print(). Adjusted all the address printing in the
5590 debugging in store.c to use %p rather than %d.
5592 81. There was a concern that a line of code in smtp_in.c could overflow a
5593 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5594 initially expressed, the concern was not well-founded, because trailing
5595 spaces are removed early. However, if the trailing spaces were followed by
5596 a NULL, they did not get removed, so the overflow was possible. Two fixes
5599 (a) I re-wrote the offending code in a cleaner fashion.
5600 (b) If an incoming SMTP command contains a NULL character, it is rejected
5603 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5604 runs initgroups(), so that if the Exim user is in any additional groups,
5605 they will be used during message reception.
5611 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob