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.
129 PP/01 Solaris build fix for Oracle's LDAP libraries.
130 Bugzilla 1109, patch from Stephen Usher.
132 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
134 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
137 TF/02 Fix a couple more cases where we did not log the error message
138 when unlink() failed. See also change 4.74-TF/03.
140 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
141 lock up or crash if it happened to be inside a call to libc when it
142 got a SIGUSR1 from exiwhat.
144 The SIGUSR1 handler appends the current process status to the process
145 log which is later printed by exiwhat. It used to use the general
146 purpose logging code to do this, but several functions it calls are
147 not safe for signals.
149 The new output code in the SIGUSR1 handler is specific to the process
150 log, and simple enough that it's easy to inspect for signal safety.
151 Removing some special cases also simplifies the general logging code.
152 Removing the spurious timestamps from the process log simplifies
155 TF/04 Improved ratelimit ACL condition.
157 The /noupdate option has been deprecated in favour of /readonly which
158 has clearer semantics. The /leaky, /strict, and /readonly update modes
159 are mutually exclusive. The update mode is no longer included in the
160 database key; it just determines when the database is updated. (This
161 means that when you upgrde Exim will forget old rate measurements.)
163 Exim now checks that the per_* options are used with an update mode that
164 makes sense for the current ACL. For example, when Exim is processing a
165 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
166 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
167 must specify per_mail/readonly. If you omit the update mode it defaults to
168 /leaky where that makes sense (as before) or /readonly where required.
170 The /noupdate option is now undocumented but still supported for
171 backwards compatibility. It is equivalent to /readonly except that in
172 ACLs where /readonly is required you may specify /leaky/noupdate or
173 /strict/noupdate which are treated the same as /readonly.
175 A useful new feature is the /count= option. This is a generalization
176 of the per_byte option, so that you can measure the throughput of other
177 aggregate values. For example, the per_byte option is now equivalent
178 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
180 The per_rcpt option has been generalized using the /count= mechanism
181 (though it's more complicated than the per_byte equivalence). When it is
182 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
183 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
184 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
185 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
186 Note that using per_rcpt with a non-readonly update mode in more than
187 one ACL will cause the recipients to be double-counted. (The per_mail
188 and per_byte options don't have this problem.)
190 The handling of very low rates has changed slightly. If the computed rate
191 is less than the event's count (usually one) then this event is the first
192 after a long gap. In this case the rate is set to the same as this event's
193 count, so that the first message of a spam run is counted properly.
195 The major new feature is a mechanism for counting the rate of unique
196 events. The new per_addr option counts the number of different
197 recipients that someone has sent messages to in the last time period. It
198 behaves like per_rcpt if all the recipient addresses are different, but
199 duplicate recipient addresses do not increase the measured rate. Like
200 the /count= option this is a general mechanism, so the per_addr option
201 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
202 example, measure the rate that a client uses different sender addresses
203 with the options per_mail/unique=$sender_address. There are further
204 details in the main documentation.
206 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
208 TF/06 Removed a few PCRE remnants.
210 TF/07 Automatically extract Exim's version number from tags in the git
211 repository when doing development or release builds.
213 PP/02 Raise smtp_cmd_buffer_size to 16kB.
214 Bugzilla 879. Patch from Paul Fisher.
216 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
217 Heavily based on revision 40f9a89a from Simon Arlott's tree.
220 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
222 PP/05 Variable $av_failed, true if the AV scanner deferred.
223 Bugzilla 1078. Patch from John Horne.
225 PP/06 Stop make process more reliably on build failure.
226 Bugzilla 1087. Patch from Heiko Schlittermann.
228 PP/07 Make maildir_use_size_file an _expandable_ boolean.
229 Bugzilla 1089. Patch from Heiko Schlittermann.
231 PP/08 Handle ${run} returning more data than OS pipe buffer size.
232 Bugzilla 1131. Patch from Holger Weiß.
234 PP/09 Handle IPv6 addresses with SPF.
235 Bugzilla 860. Patch from Wolfgang Breyha.
237 PP/10 GnuTLS: support TLS 1.2 & 1.1.
239 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
242 PP/11 match_* no longer expand right-hand-side by default.
243 New compile-time build option, EXPAND_LISTMATCH_RHS.
244 New expansion conditions, "inlist", "inlisti".
246 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
248 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
250 PP/14 fix log_write() format string regression from TF/03.
251 Bugzilla 1152. Patch from Dmitry Isaikin.
257 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
259 PP/02 Harmonised TLS library version reporting; only show if debugging.
260 Layout now matches that introduced for other libraries in 4.74 PP/03.
262 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
264 PP/04 New "dns_use_edns0" global option.
266 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
269 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
270 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
272 TK/01 Updated PolarSSL code to 0.14.2.
273 Bugzilla 1097. Patch from Andreas Metzler.
275 PP/07 Catch divide-by-zero in ${eval:...}.
278 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
281 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
282 format-string attack -- SECURITY: remote arbitrary code execution.
284 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
285 time unintentionally subject to list matching rules, letting the header
286 cause arbitrary Exim lookups (of items which can occur in lists, *not*
287 arbitrary string expansion). This allowed for information disclosure.
289 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
290 INT_MIN/-1 -- value coerced to INT_MAX.
296 NM/01 Workround for PCRE version dependency in version reporting
299 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
300 This fixes portability to compilers other than gcc, notably
301 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
303 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
304 makefiles for portability to HP-UX and POSIX correctness.
306 PP/01 Permit LOOKUP_foo enabling on the make command-line.
307 Also via indented variable definition in the Makefile.
308 (Debugging by Oliver Heesakkers).
310 PP/02 Restore caching of spamd results with expanded spamd_address.
311 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
313 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
314 Improves build reliability. Fix from: Frank Elsner
316 NM/02 Fix wide character breakage in the rfc2047 coding
317 Fixes bug 1064. Patch from Andrey N. Oktyabrski
319 NM/03 Allow underscore in dnslist lookups
320 Fixes bug 1026. Patch from Graeme Fowler
322 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
323 Code patches from Adam Ciarcinski of NetBSD.
325 NM/04 Fixed exiqgrep to cope with mailq missing size issue
328 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
329 is logged, to avoid truncation. Patch from John Horne.
331 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
332 Patch from Jakob Hirsch.
334 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
335 SQL string expansion failure details.
336 Patch from Andrey Oktyabrski.
338 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
339 Patch from Simon Arlott.
341 PP/09 New lookups functionality failed to compile on old gcc which rejects
342 extern declarations in function scope.
343 Patch from Oliver Fleischmann
345 PP/10 Use sig_atomic_t for flags set from signal handlers.
346 Check getgroups() return and improve debugging.
347 Fixed developed for diagnosis in bug 927 (which turned out to be
350 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
351 Patch from Mark Zealey.
353 PP/12 Bugzilla 1056: Improved spamd server selection.
354 Patch from Mark Zealey.
356 PP/13 Bugzilla 1086: Deal with maildir quota file races.
357 Based on patch from Heiko Schlittermann.
359 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
360 Patch from Uwe Doering, sign-off by Michael Haardt.
362 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
363 variable declaration deep within a block. Bug and patch from
366 PP/15 lookups-Makefile IRIX compatibilty coercion.
368 PP/16 Make DISABLE_DKIM build knob functional.
370 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
371 Patch by Simon Arlott
373 TF/03 Fix valgrind.h portability to C89 compilers that do not support
374 variable argument macros. Our copy now differs from upstream.
380 TF/01 Failure to get a lock on a hints database can have serious
381 consequences so log it to the panic log.
383 TF/02 Log LMTP confirmation messages in the same way as SMTP,
384 controlled using the smtp_confirmation log selector.
386 TF/03 Include the error message when we fail to unlink a spool file.
388 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
389 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
390 for maintaining out-of-tree patches for some time.
392 PP/01 Bugzilla 139: Documentation and portability issues.
393 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
394 Handle per-OS dynamic-module compilation flags.
396 PP/02 Let /dev/null have normal permissions.
397 The 4.73 fixes were a little too stringent and complained about the
398 permissions on /dev/null. Exempt it from some checks.
399 Reported by Andreas M. Kirchwitz.
401 PP/03 Report version information for many libraries, including
402 Exim version information for dynamically loaded libraries. Created
403 version.h, now support a version extension string for distributors
404 who patch heavily. Dynamic module ABI change.
406 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
407 privilege escalation vulnerability whereby the Exim run-time user
408 can cause root to append content of the attacker's choosing to
411 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
414 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
415 If dropping privileges for untrusted macros, we disabled normal logging
416 on the basis that it would fail; for the Exim run-time user, this is not
417 the case, and it resulted in successful deliveries going unlogged.
418 Fixed. Reported by Andreas Metzler.
424 PP/01 Date: & Message-Id: revert to normally being appended to a message,
425 only prepend for the Resent-* case. Fixes regression introduced in
426 Exim 4.70 by NM/22 for Bugzilla 607.
428 PP/02 Include check_rfc2047_length in configure.default because we're seeing
429 increasing numbers of administrators be bitten by this.
431 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
433 PP/03 Bugzilla 994: added openssl_options main configuration option.
435 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
437 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
439 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
441 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
442 without a peer certificate, leading to a segfault because of an
443 assumption that peers always have certificates. Be a little more
444 paranoid. Problem reported by Martin Tscholak.
446 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
447 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
448 NB: ClamAV planning to remove STREAM in "middle of 2010".
449 CL also introduces -bmalware, various -d+acl logging additions and
450 more caution in buffer sizes.
452 PP/09 Implemented reverse_ip expansion operator.
454 PP/10 Bugzilla 937: provide a "debug" ACL control.
456 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
458 PP/12 Bugzilla 973: Implement --version.
460 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
462 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
464 PP/15 Bugzilla 816: support multiple condition rules on Routers.
466 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
467 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
468 ignore trailing whitespace.
470 JJ/02 prevent non-panic DKIM error from being sent to paniclog
472 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
475 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
476 Notification from Dr Andrew Aitchison.
478 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
479 ExtendedDetectionInfo response format.
480 Notification from John Horne.
482 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
485 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
486 XSL and documented dependency on system catalogs, with examples of how
489 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
492 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
493 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
494 configuration file which is writeable by the Exim user or group.
496 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
497 of configuration files to cover files specified with the -C option if
498 they are going to be used with root privileges, not just the default
501 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
502 option (effectively making it always true).
504 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
505 files to be used while preserving root privileges.
507 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
508 that rogue child processes cannot use them.
510 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
511 run-time user, instead of root.
513 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
514 Exim run-time user without dropping privileges.
516 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
517 result string, instead of calling string_vformat() twice with the same
520 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
521 for other users. Others should always drop root privileges if they use
522 -C on the command line, even for a whitelisted configure file.
524 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
526 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
532 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
533 $data_path, and $header_path variables; fixed documentation bugs and
536 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
537 exipick to access non-standard spools, including the "frozen" queue
540 NM/01 Bugzilla 965: Support mysql stored procedures.
541 Patch from Alain Williams
543 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
545 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
546 Patch from Andreas Metzler
548 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
549 Patch from Kirill Miazine
551 NM/05 Bugzilla 671: Added umask to procmail example.
553 JJ/03 installed exipick 20100323.0, fixing doc bug
555 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
556 directory. Notification and patch from Dan Rosenberg.
558 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
560 TK/02 Improve log output when DKIM signing operation fails.
562 MH/01 Treat the transport option dkim_domain as a colon separated
563 list, not as a single string, and sign the message with each element,
564 omitting multiple occurences of the same signer.
566 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
567 Bugzilla 985, 986. Patch by Simon Arlott
569 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
570 Patch by Simon Arlott
572 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
573 MBX locking. Notification from Dan Rosenberg.
579 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
581 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
583 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
585 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
587 NM/04 Bugzilla 915: Flag broken perl installation during build.
593 TK/01 Added patch by Johannes Berg that expands the main option
594 "spamd_address" if it starts with a dollar sign.
596 TK/02 Write list of recipients to X-Envelope-Sender header when building
597 the mbox-format spool file for content scanning (suggested by Jakob
600 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
601 (http://www.dcc-servers.net/) support via dccifd. Activated by
602 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
604 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
605 by Mark Daniel Reidel <mr@df.eu>.
607 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
608 When building exim an external PCRE library is now needed -
609 PCRE is a system library on the majority of modern systems.
610 See entry on PCRE_LIBS in EDITME file.
612 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
613 conversation. Added nologin parameter to request.
614 Patch contributed by Kirill Miazine.
616 TF/01 Do not log submission mode rewrites if they do not change the address.
618 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
620 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
621 log files in place. Contributed by Roberto Lima.
623 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
625 TF/03 Bugzilla 615: When checking the local_parts router precondition
626 after a local_part_suffix or local_part_prefix option, Exim now
627 does not use the address's named list lookup cache, since this
628 contains cached lookups for the whole local part.
630 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
631 Robert Millan. Documentation is in experimental-spec.txt.
633 TF/04 Bugzilla 668: Fix parallel build (make -j).
635 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
637 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
638 Patch provided by Jan Srzednicki.
640 TF/05 Leading white space used to be stripped from $spam_report which
641 wrecked the formatting. Now it is preserved.
643 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
644 that they are available at delivery time.
646 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
648 TF/08 TLS error reporting now respects the incoming_interface and
649 incoming_port log selectors.
651 TF/09 Produce a more useful error message if an SMTP transport's hosts
652 setting expands to an empty string.
654 NM/06 Bugzilla 744: EXPN did not work under TLS.
655 Patch provided by Phil Pennock.
657 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
658 Patch provided by Richard Godbee.
660 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
661 acl_smtp_notquit, added index entry.
663 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
664 Patch provided by Eugene Bujak.
666 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
667 accept(). Patch provided by Maxim Dounin.
669 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
670 Patch provided by Phil Pennock.
672 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
674 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
675 Patch provided by Brad "anomie" Jorsch.
677 NM/14 Bugzilla 622: Added timeout setting to transport filter.
678 Patch provided by Dean Brooks.
680 TK/05 Add native DKIM support (does not depend on external libraries).
682 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
683 Patch provided by Graeme Fowler.
685 NM/16 Bugzilla 851: Documentation example syntax fix.
687 NM/17 Changed NOTICE file to remove references to embedded PCRE.
689 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
692 NM/19 Bugzilla 745: TLS version reporting.
693 Patch provided by Phil Pennock.
695 NM/20 Bugzilla 167: bool: condition support.
696 Patch provided by Phil Pennock.
698 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
699 clients. Patch provided by Phil Pennock.
701 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
702 Patch provided by Brad "anomie" Jorsch.
704 NM/23 Bugzilla 687: Fix misparses in eximstats.
705 Patch provided by Heiko Schlittermann.
707 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
708 Patch provided by Heiko Schlittermann.
710 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
711 plus update to original patch.
713 NM/26 Bugzilla 799: Documentation correction for ratelimit.
715 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
716 Patch provided by David Brownlee.
718 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
720 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
722 NM/30 Bugzilla 888: TLS documentation bugfixes.
724 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
726 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
727 Unlike the original bugzilla I have changed all shell scripts in src tree.
729 NM/33 Bugzilla 898: Transport filter timeout fix.
730 Patch by Todd Rinaldo.
732 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
733 Patch by Serge Demonchaux.
735 NM/35 Bugzilla 39: Base64 decode bug fixes.
736 Patch by Jakob Hirsch.
738 NM/36 Bugzilla 909: Correct connect() call in dcc code.
740 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
742 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
744 NM/39 Bugzilla 911: Fixed MakeLinks build script.
750 TK/01 Add preliminary DKIM support. Currently requires a forked version of
751 ALT-N's libdkim that I have put here:
752 http://duncanthrax.net/exim-experimental/
754 Note to Michael Haardt: I had to rename some vars in sieve.c. They
755 were called 'true' and it seems that C99 defines that as a reserved
756 keyword to be used with 'bool' variable types. That means you could
757 not include C99-style headers which use bools without triggering
758 build errors in sieve.c.
760 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
761 as mailq or other aliases. Changed the --help handling significantly
762 to do whats expected. exim_usage() emits usage/help information.
764 SC/01 Added the -bylocaldomain option to eximstats.
766 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
768 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
770 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
776 PH/01 Another patch from the Sieve maintainer.
778 PH/02 When an IPv6 address is converted to a string for single-key lookup
779 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
780 dots are used instead of colons so that keys in lsearch files need not
781 contain colons. This was done some time before quoting was made available
782 in lsearch files. However, iplsearch files do require colons in IPv6 keys
783 (notated using the quote facility) so as to distinguish them from IPv4
784 keys. This meant that lookups for IP addresses in host lists did not work
785 for iplsearch lookups.
787 This has been fixed by arranging for IPv6 addresses to be expressed with
788 colons if the lookup type is iplsearch. This is not incompatible, because
789 previously such lookups could never work.
791 The situation is now rather anomolous, since one *can* have colons in
792 ordinary lsearch keys. However, making the change in all cases is
793 incompatible and would probably break a number of configurations.
795 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
798 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
799 conversion specification without a maximum field width, thereby enabling
800 a rogue spamd server to cause a buffer overflow. While nobody in their
801 right mind would setup Exim to query an untrusted spamd server, an
802 attacker that gains access to a server running spamd could potentially
803 exploit this vulnerability to run arbitrary code as the Exim user.
805 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
806 $primary_hostname instead of what libspf2 thinks the hosts name is.
808 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
809 a directory entry by the name of the lookup key. Previously, if a
810 symlink pointed to a non-existing file or a file in a directory that
811 Exim lacked permissions to read, a lookup for a key matching that
812 symlink would fail. Now it is enough that a matching directory entry
813 exists, symlink or not. (Bugzilla 503.)
815 PH/03 The body_linecount and body_zerocount variables are now exported in the
818 PH/04 Added the $dnslist_matched variable.
820 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
821 This means they are set thereafter only if the connection becomes
824 PH/06 Added the client_condition to authenticators so that some can be skipped
825 by clients under certain conditions.
827 PH/07 The error message for a badly-placed control=no_multiline_responses left
828 "_responses" off the end of the name.
830 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
832 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
833 (without spaces) instead of just copying the configuration text.
835 PH/10 Added the /noupdate option to the ratelimit ACL condition.
837 PH/11 Added $max_received_linelength.
839 PH/12 Added +ignore_defer and +include_defer to host lists.
841 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
842 way in which PCRE > 7.0 is built.
844 PH/14 Implemented queue_only_load_latch.
846 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
847 MAIL command. The effect was to mangle the value on 64-bit systems.
849 PH/16 Another patch from the Sieve maintainer.
851 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
853 PH/18 If a system quota error occurred while trying to create the file for
854 a maildir delivery, the message "Mailbox is full" was not appended to the
855 bounce if the delivery eventually timed out. Change 4.67/27 below applied
856 only to a quota excession during the actual writing of the file.
858 PH/19 It seems that peer DN values may contain newlines (and other non-printing
859 characters?) which causes problems in log lines. The DN values are now
860 passed through string_printing() before being added to log lines.
862 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
863 and InterBase are left for another time.)
865 PH/21 Added message_body_newlines option.
867 PH/22 Guard against possible overflow in moan_check_errorcopy().
869 PH/23 POSIX allows open() to be a macro; guard against that.
871 PH/24 If the recipient of an error message contained an @ in the local part
872 (suitably quoted, of course), incorrect values were put in $domain and
873 $local_part during the evaluation of errors_copy.
879 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
880 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
883 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
884 issue a MAIL command.
886 PH/02 In an ACL statement such as
888 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
890 if a client was not listed at all, or was listed with a value other than
891 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
892 the condition was not true (as it should be), so access was not denied.
893 The bug was that the ! inversion was incorrectly passed on to the second
894 item. This has been fixed.
896 PH/03 Added additional dnslists conditions == and =& which are different from
897 = and & when the dns lookup returns more than one IP address.
899 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
900 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
902 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
903 FSYNC, which compiles an option called disable_fsync that allows for
904 bypassing fsync(). The documentation is heavily laced with warnings.
906 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
908 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
909 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
910 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
911 including adding "make clean"; (3) Added -fPIC when compiling the test
912 dynamically loaded module, to get rid of a warning.
914 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
915 message fails, move_frozen_messages = true and ignore_bounce_errors_after
916 = 0s. The bug is otherwise harmless.
918 PH/07 There was a bug in the dovecot authenticator such that the value of
919 $auth1 could be overwritten, and so not correctly preserved, after a
920 successful authentication. This usually meant that the value preserved by
921 the server_setid option was incorrect.
923 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
925 PH/09 Installed PCRE release 7.0.
927 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
928 run for batched SMTP input. It is now run at the start of every message
929 in the batch. While fixing this I discovered that the process information
930 (output by running exiwhat) was not always getting set for -bs and -bS
931 input. This is fixed, and it now also says "batched" for BSMTP.
933 PH/11 Added control=no_pipelining.
935 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
936 patch, slightly modified), and move the expansion of helo_data till after
937 the connection is made in the smtp transport (so it can use these
940 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
942 PH/14 Added log_selector = +pid.
944 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
946 PH/16 Add ${if forany and ${if forall.
948 PH/17 Added dsn_from option to vary the From: line in DSNs.
950 PH/18 Flush SMTP output before performing a callout, unless control =
951 no_callout_flush is set.
953 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
954 was true (the default) a successful delivery failed to delete the retry
955 item, thus causing premature timeout of the address. The bug is now
958 PH/20 Added hosts_avoid_pipelining to the smtp transport.
960 PH/21 Long custom messages for fakedefer and fakereject are now split up
961 into multiline reponses in the same way that messages for "deny" and
962 other ACL rejections are.
964 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
965 with slight modification.
967 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
968 draft, changing the syntax and factoring some duplicate code".
970 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
971 for deliveries of the second and subsequent messages over the same SMTP
974 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
975 ${reduce, with only minor "tidies".
977 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
979 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
980 expansion side effects.
982 PH/27 When a message times out after an over-quota error from an Exim-imposed
983 quota, the bounce message says "mailbox is full". This message was not
984 being given when it was a system quota that was exceeded. It now should
987 MH/03 Made $recipients available in local_scan(). local_scan() already has
988 better access to the recipient list through recipients_list[], but
989 $recipients can be useful in postmaster-provided expansion strings.
991 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
992 in the case of a MAIL command with additional options following the
993 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
994 were accidentally chopped off.
996 PH/29 SMTP synchronization checks are implemented when a command is read -
997 there is a check that no more input is waiting when there shouldn't be
998 any. However, for some commands, a delay in an ACL can mean that it is
999 some time before the response is written. In this time, more input might
1000 arrive, invalidly. So now there are extra checks after an ACL has run for
1001 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1002 pipelining has not been advertised.
1004 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1006 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1007 returned in the SMTP response when smtp_return_error_details was false.
1008 This has been fixed.
1010 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1011 instead of the C I/O that was originally supplied, because problems were
1012 reported on Solaris.
1014 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1015 Exim which did not show up earlier: it was assuming that a call to
1016 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1017 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1018 was a macro that became an assignment, so it seemed to work. This has
1019 changed to a proper function call with a void return, hence the compile
1020 error. Exim's code has been fixed.
1022 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1025 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1027 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1029 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1030 because it turns out to be too big or there is a timeout) there is no
1031 "Completed" line in the log. When some messages of this type were
1032 selected by exigrep, they were listed as "not completed". Others were
1033 picked up by some special patterns. I have improved the selection
1034 criteria to be more general.
1036 PH/38 The host_find_failed option in the manualroute router can now be set
1037 to "ignore", to completely ignore a host whose IP address cannot be
1038 found. If all hosts are ignored, the behaviour is controlled by the new
1039 host_all_ignored option.
1041 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1042 homing or because of multiple MX records with /mx) generated more than
1043 one IP address, and the following item turned out to be the local host,
1044 all the secondary addresses of the first item were incorrectly removed
1045 from the list, along with the local host and any following hosts (which
1046 is what is supposed to happen).
1048 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1049 whoever called Exim into the -H file. In the case of the daemon it was
1050 behaving confusingly. When first started, it used values for whoever
1051 started the daemon, but after a SIGHUP it used the Exim user (because it
1052 calls itself on a restart). I have changed the code so that it now always
1055 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1056 message are rejected with the same error (e.g. no authentication or bad
1057 sender address), and a DATA command is nevertheless sent (as can happen
1058 with PIPELINING or a stupid MUA), the error message that was given to the
1059 RCPT commands is included in the rejection of the DATA command. This is
1060 intended to be helpful for MUAs that show only the final error to their
1063 PH/42 Another patch from the Sieve maintainer.
1065 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1066 Eximstats - Fixed some broken HTML links and added missing column headers
1068 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1069 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1071 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1073 PH/43 Yet another patch from the Sieve maintainer.
1075 PH/44 I found a way to check for a TCP/IP connection going away before sending
1076 the response to the final '.' that terminates a message, but only in the
1077 case where the client has not sent further data following the '.'
1078 (unfortunately, this is allowed). However, in many cases there won't be
1079 any further data because there won't be any more messages to send. A call
1080 to select() can be used: if it shows that the input is "ready", there is
1081 either input waiting, or the socket has been closed. An attempt to read
1082 the next input character can distinguish the two cases. Previously, Exim
1083 would have sent an OK response which the client would never have see.
1084 This could lead to message repetition. This fix should cure that, at
1085 least in a lot of common cases.
1087 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1088 advertised in response to EHLO.
1094 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1095 fixed by 4.65/MH/01 (is this a record?) are fixed:
1097 (i) An empty string was always treated as zero by the numeric comparison
1098 operators. This behaviour has been restored.
1100 (ii) It is documented that the numeric comparison operators always treat
1101 their arguments as decimal numbers. This was broken in that numbers
1102 starting with 0 were being interpreted as octal.
1104 While fixing these problems I realized that there was another issue that
1105 hadn't been noticed. Values of message_size_limit (both the global option
1106 and the transport option) were treated as octal if they started with 0.
1107 The documentation was vague. These values are now always treated as
1108 decimal, and I will make that clear in the documentation.
1114 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1115 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1118 MH/01 Don't check that the operands of numeric comparison operators are
1119 integers when their expansion is in "skipping" mode (fixes bug
1120 introduced by 4.64-PH/07).
1122 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1123 child addresses, Exim now panics and dies. Previously, because the count
1124 is held in a short int, deliveries were likely to be lost. As such a
1125 large number of recipients for a single message is ridiculous
1126 (performance will be very, very poor), I have chosen to impose a limit
1127 rather than extend the field.
1133 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1134 leftover -K file (the existence of which was triggered by #402).
1135 While we were at it, introduced process PID as part of the -K
1136 filename. This should rule out race conditions when creating
1139 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1140 processing considerably. Previous code took too long for large mails,
1141 triggering a timeout which in turn triggers #401.
1143 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1144 in the DK code in transports.c. sendfile() is not really portable,
1145 hence the _LINUX specificness.
1147 TF/01 In the add_headers option to the mail command in an Exim filter,
1148 there was a bug that Exim would claim a syntax error in any
1149 header after the first one which had an odd number of characters
1152 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1153 callout verification, Exim cached a "reject" for the entire domain. This
1154 is correct for most verifications, but it is not correct for a recipient
1155 verification with use_sender or use_postmaster set, because in that case
1156 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1157 case of MAIL FROM:<> rejection from other early rejections (e.g.
1158 rejection of HELO). When verifying a recipient using a non-null MAIL
1159 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1160 Whatever the result of the callout, the value of the domain cache is
1161 left unchanged (for any other kind of callout, getting as far as trying
1162 RCPT means that the domain itself is ok).
1164 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1167 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1168 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1169 session, the daemon ignores ECONNECT errors and logs others; it now
1170 ignores EPIPE as well.
1172 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1173 (quoted-printable decoding).
1175 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1176 later the small subsequent patch to fix an introduced bug.
1178 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1180 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1182 PH/08 An error is now given if message_size_limit is specified negative.
1184 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1185 to be given (somewhat) arbitrary names.
1187 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1190 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1191 miscellaneous code fixes
1193 PH/10 Added the log_reject_target ACL modifier to specify where to log
1196 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1197 hostname. This is wrong, because it relates to the incoming message (and
1198 probably the interface on which it is arriving) and not to the outgoing
1199 callout (which could be using a different interface). This has been
1200 changed to use the value of the helo_data option from the smtp transport
1201 instead - this is what is used when a message is actually being sent. If
1202 there is no remote transport (possible with a router that sets up host
1203 addresses), $smtp_active_hostname is used.
1205 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1206 tweaks were necessary in order to get it to work (see also 21 below):
1207 (a) The code assumed that strncpy() returns a negative number on buffer
1208 overflow, which isn't the case. Replaced with Exim's string_format()
1210 (b) There were several signed/unsigned issues. I just did the minimum
1211 hacking in of casts. There is scope for a larger refactoring.
1212 (c) The code used strcasecmp() which is not a standard C function.
1213 Replaced with Exim's strcmpic() function.
1214 (d) The code set only $1; it now sets $auth1 as well.
1215 (e) A simple test gave the error "authentication client didn't specify
1216 service in request". It would seem that Dovecot has changed its
1217 interface. Fortunately there's a specification; I followed it and
1218 changed what the client sends and it appears to be working now.
1220 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1223 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1224 address A is aliased to B and C, where B exists and C does not. Without
1225 -v the output is "A verified" because verification stops after a
1226 successful redirection if more than one address is generated. However,
1227 with -v the child addresses are also verified. Exim was outputting "A
1228 failed to verify" and then showing the successful verification for C,
1229 with its parentage. It now outputs "B failed to verify", showing B's
1230 parentage before showing the successful verification of C.
1232 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1233 look up a TXT record in a specific list after matching in a combined
1236 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1237 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1238 they consult the DNS. I had assumed they would set it the way they
1239 wanted; and indeed my experiments on Linux seem to show that in some
1240 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1241 To be on the safe side, however, I have now made the interface to
1242 host_find_byname() similar to host_find_bydns(), with an argument
1243 containing the DNS resolver options. The host_find_byname() function now
1244 sets these options at its start, just as host_find_bydns() does. The smtp
1245 transport options dns_qualify_single and dns_search_parents are passed to
1246 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1247 of host_find_byname() use the default settings of RES_DEFNAMES
1248 (qualify_single) but not RES_DNSRCH (search_parents).
1250 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1251 spool_read_header() do less string testing, by means of a preliminary
1252 switch on the second character of optional "-foo" lines. (This is
1253 overdue, caused by the large number of possibilities that now exist.
1254 Originally there were few.) While I was there, I also converted the
1255 str(n)cmp tests so they don't re-test the leading "-" and the first
1256 character, in the hope this might squeeze out yet more improvement.
1258 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1259 flag allowing group syntax was set by the header_syntax check but not
1260 turned off, possible causing trouble later; (2) The flag was not being
1261 set at all for the header_verify test, causing "group"-style headers to
1262 be rejected. I have now set it in this case, and also caused header_
1263 verify to ignore an empty address taken from a group. While doing this, I
1264 came across some other cases where the code for allowing group syntax
1265 while scanning a header line wasn't quite right (mostly, not resetting
1266 the flag correctly in the right place). These bugs could have caused
1267 trouble for malformed header lines. I hope it is now all correct.
1269 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1270 with the "reply" argument non-NULL. The code, however (which originally
1271 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1272 but it didn't always do it. This confused somebody who was copying the
1273 code for some other use. I have removed all the tests.
1275 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1276 feature that was used to support insecure browsers during the U.S. crypto
1277 embargo. It requires special client support, and Exim is probably the
1278 only MTA that supported it -- and would never use it because real RSA is
1279 always available. This code has been removed, because it had the bad
1280 effect of slowing Exim down by computing (never used) parameters for the
1281 RSA_EXPORT functionality.
1283 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1284 authenticator to fail if there's a tab character in the incoming data
1285 (there should never be unless someone is messing about, as it's supposed
1286 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1287 if the connection is using TLS or if the remote IP is the same as the
1288 local IP, and the "valid-client-cert option" if a client certificate has
1291 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1292 authenticators. This can be used for authorization after authentication
1293 succeeds. (In the case of plaintext, it servers for both authentication
1296 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1297 if any retry times were supplied.
1299 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1300 connection, where there is no client IP address to check. In this
1301 situation, the verify now always succeeds.
1303 PH/25 Applied John Jetmore's -Mset patch.
1305 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1307 PH/27 In a string expansion for a processed (not raw) header when multiple
1308 headers of the same name were present, leading whitespace was being
1309 removed from all of them, but trailing whitespace was being removed only
1310 from the last one. Now trailing whitespace is removed from each header
1311 before concatenation. Completely empty headers in a concatenation (as
1312 before) are ignored.
1314 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1315 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1317 PH/29 [Removed. This was a change that I later backed out, and forgot to
1318 correct the ChangeLog entry (that I had efficiently created) before
1319 committing the later change.]
1321 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1322 address errors (4xx response to RCPT) over the same connection as other
1323 messages routed to the same hosts. Such deliveries are always "forced",
1324 so retry times are not inspected. This resulted in far too many retries
1325 for the affected addresses. The effect occurred only when there were more
1326 hosts than the hosts_max_try setting in the smtp transport when it had
1327 the 4xx errors. Those hosts that it had tried were not added to the list
1328 of hosts for which the message was waiting, so if all were tried, there
1329 was no problem. Two fixes have been applied:
1331 (i) If there are any address or message errors in an SMTP delivery, none
1332 of the hosts (tried or untried) are now added to the list of hosts
1333 for which the message is waiting, so the message should not be a
1334 candidate for sending over the same connection that was used for a
1335 successful delivery of some other message. This seems entirely
1336 reasonable: after all the message is NOT "waiting for some host".
1337 This is so "obvious" that I'm not sure why it wasn't done
1338 previously. Hope I haven't missed anything, but it can't do any
1339 harm, as the worst effect is to miss an optimization.
1341 (ii) If, despite (i), such a delivery is accidentally attempted, the
1342 routing retry time is respected, so at least it doesn't keep
1343 hammering the server.
1345 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1346 in ${readsocket because some servers need this prod.
1348 PH/32 Added some extra debug output when updating a wait-xxx database.
1350 PH/33 The hint "could be header name not terminated by colon", which has been
1351 given for certain expansion errors for a long time, was not being given
1352 for the ${if def:h_colon_omitted{... case.
1354 PH/34 The spec says: "With one important exception, whenever a domain list is
1355 being scanned, $domain contains the subject domain." There was at least
1356 one case where this was not true.
1358 PH/35 The error "getsockname() failed: connection reset by peer" was being
1359 written to the panic log as well as the main log, but it isn't really
1360 panic-worthy as it just means the connection died rather early on. I have
1361 removed the panic log writing for the ECONNRESET error when getsockname()
1364 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1365 runs only) independently of the message's sender address. This meant
1366 that, if the 4xx error was in fact related to the sender, a different
1367 message to the same recipient with a different sender could confuse
1368 things. In particualar, this can happen when sending to a greylisting
1369 server, but other circumstances could also provoke similar problems.
1370 I have changed the default so that the retry time for these errors is now
1371 based a combination of the sender and recipient addresses. This change
1372 can be overridden by setting address_retry_include_sender=false in the
1375 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1376 remote server are returned as part of bounce messages. This was not
1377 happening for LMTP over a pipe (the lmtp transport), but now it is the
1378 same for both kinds of LMTP.
1380 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1381 in header lines that were in fact CNAMEs. This is no longer the case.
1383 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1384 and queue runs started by the daemon processed all messages. This has
1385 been fixed so that -R and -S can now usefully be given with -q<time>.
1387 PH/40 Import PCRE release 6.7 (fixes some bugs).
1389 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1391 PH/42 Give an error if -q is specified more than once.
1393 PH/43 Renamed the variables $interface_address and $interface_port as
1394 $received_ip_address and $received_port, to make it clear that these
1395 values apply to message reception, and not to the outgoing interface when
1396 a message is delivered. (The old names remain recognized, of course.)
1398 PH/44 There was no timeout on the connect() call when using a Unix domain
1399 socket in the ${readsocket expansion. There now is.
1401 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1402 be meaningful with "accept".
1404 SC/01 Eximstats V1.43
1405 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1407 SC/02 Eximstats V1.44
1408 Use a glob alias rather than an array ref in the generated
1409 parser. This improves both readability and performance.
1411 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1412 Collect SpamAssassin and rejection statistics.
1413 Don't display local sender or destination tables unless
1414 there is data to show.
1415 Added average volumes into the top table text output.
1417 SC/04 Eximstats V1.46
1418 Collect data on the number of addresses (recipients)
1419 as well as the number of messages.
1421 SC/05 Eximstats V1.47
1422 Added 'Message too big' to the list of mail rejection
1423 reasons (thanks to Marco Gaiarin).
1425 SC/06 Eximstats V1.48
1426 Mainlog lines which have GMT offsets and are too short to
1427 have a flag are now skipped.
1429 SC/07 Eximstats V1.49 (Alain Williams)
1430 Added the -emptyok flag.
1432 SC/08 Eximstats V1.50
1433 Fixes for obtaining the IP address from reject messages.
1435 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1436 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1437 whitesspace changes from 4.64-PH/27
1439 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1442 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1443 are found, allow negative numbers in numeric criteria)
1445 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1447 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1448 to match changes made in 4.64-PH/43
1450 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1452 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1453 "domain" is actually an IP address, and force a failure. This locks out
1454 those revolvers/nameservers that support "A-for-A" lookups, in
1455 contravention of the specifications.
1457 PH/48 When a host name was looked up from an IP address, and the subsequent
1458 forward lookup of the name timed out, the host name was left in
1459 $sender_host_name, contrary to the specification.
1461 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1462 restricted to single-key lookups, Exim was not diagnosing an error if
1463 * or *@ was used with a query-style lookup.
1465 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1467 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1468 long ago, but noone interested enough thought of it. Let's just say that
1469 the "1.1" means that there are some new functions that weren't there at
1470 some point in the past.
1472 PH/51 Error processing for expansion failure of helo_data from an smtp
1473 transport during callout processing was broken.
1475 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1476 tested/used via the -bh/-bhc/-bs options.
1478 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1479 bug, fixed in subsequent PCRE releases).
1481 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1482 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1484 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1490 SC/01 Use a glob alias rather than an array ref in eximstats generated
1491 parser. This improves both readability and performance.
1493 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1494 Don't display local sender or destination tables in eximstats unless
1495 there is data to show.
1496 Added average volumes into the eximstats top table text output.
1498 SC/03 Collect data on the number of addresses (recipients) as well
1499 as the number of messages in eximstats.
1501 TF/01 Correct an error in the documentation for the redirect router. Exim
1502 does (usually) call initgroups() when daemonizing.
1504 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1505 with consistent privilege compared to when running as a daemon.
1507 TF/03 Note in the spec that $authenticated_id is not set for local
1508 submissions from trusted users.
1510 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1511 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1513 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1514 by adding some example configuration directives to the default
1515 configuration file. A little bit of work is required to uncomment the
1516 directives and define how usernames and passwords are checked, but
1517 there is now a framework to start from.
1519 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1520 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1521 without this. I don't know how relevant this is to other LDAP libraries.
1523 PH/02 Add the verb name to the "unknown ACL verb" error.
1525 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1527 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1529 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1530 directory not expanded when it should be if an expanded home directory
1531 was set for the address (which is overridden by the transport).
1533 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1536 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1537 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1538 because it is too late at that time, and has no effect.
1540 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1541 security issue with \' (bugzilla #107). I could not use the
1542 PQescapeStringConn() function, because it needs a PGconn value as one of
1545 PH/08 When testing addresses using -bt, indicate those final addresses that
1546 are duplicates that would not cause an additional delivery. At least one
1547 person was confused, thinking that -bt output corresponded to deliveries.
1548 (Suppressing duplicates isn't a good idea as you lose the information
1549 about possibly different redirections that led to the duplicates.)
1551 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1552 systems where poll() doesn't work, in particular OS X.
1554 PH/10 Added more information to debugging output for retry time not reached.
1556 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1557 operations in malware.c.
1559 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1562 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1563 syslog_duplication set false, log lines that would normally be written
1564 both the the main log and to the reject log were not written to syslog at
1567 PH/14 In the default configuration, change the use of "message" in ACL warn
1568 statements to "add_header".
1570 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1571 not followed by a command (e.g. "seen endif").
1573 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1574 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1577 PH/17 Added extra conditions to the default value of delay_warning_condition
1581 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1582 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1583 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1586 The Auto-Submitted: and various List- headers are standardised, whereas I
1587 don't think Precedence: ever was.
1589 PH/18 Refactored debugging code in route_finduser() to show more information,
1590 in particular, the error code if getpwnam() issues one.
1592 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1593 This is apparently needed in addition to the PH/07 change above to avoid
1594 any possible encoding problems.
1596 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1597 but not after initializing Perl.
1599 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1600 output them only if debugging. By default they are written stderr,
1601 apparently, which is not desirable.
1603 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1606 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1609 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1611 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1612 authenticated or an ident call has been made. Suppress the default
1613 values for $authenticated_id and $authenticated_sender (but permit -oMai
1614 and -oMas) when testing with -bh.
1616 PH/24 Re-jigged the order of the tests in the default configuration so that the
1617 tests for valid domains and recipients precede the DNS black list and CSA
1618 tests, on the grounds that those ones are more expensive.
1620 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1621 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1622 This bug exists in every version of Exim that I still have, right back to
1625 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1626 However, an attempt to turn on case-sensitivity in a regex key by
1627 including (?-i) didn't work because the subject string was already
1628 lowercased, and the effects were non-intuitive. It turns out that a
1629 one-line patch can be used to allow (?-i) to work as expected.
1635 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1636 other effects) broke the use of negated acl sub-conditions.
1638 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1641 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1642 "Deny" causes Exim to reject the incoming connection with a 554 error.
1643 Unfortunately, if there is a major crisis, such as a disk failure,
1644 tcp-wrappers gives "deny", whereas what one would like would be some
1645 kind of temporary error. A kludge has been added to help with this.
1646 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1647 554 error is used if errno is still zero or contains ENOENT (which occurs
1648 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1651 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1653 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1654 errors. Otherwise a message that provokes a temporary error (when other
1655 messages do not) can cause a whole host to time out.
1657 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1658 addresses were routed directly to files or pipes from a redirect router.
1659 File deliveries just didn't batch; pipe deliveries might have suffered
1662 PH/06 A failure to get a lock for a hints database would erroneously always say
1663 "Failed to get write lock", even when it was really a read lock.
1665 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1666 of 0600. This has been changed to use the value of the lockfile_mode
1667 option (which defaults to 0600).
1669 PH/08 Applied small patch from the Sieve maintainer.
1671 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1672 folder from quota calculations, a direct delivery into this folder messed
1673 up the contents of the maildirsize file. This was because the regex was
1674 used only to exclude .Trash (or whatever) when the size of the mailbox
1675 was calculated. There was no check that a delivery was happening into an
1676 excluded directory. This bug has been fixed by ignoring all quota
1677 processing for deliveries into excluded directories.
1679 PH/10 Added the maildirfolder_create_regex option to appendfile.
1685 PH/01 The code for finding all the local interface addresses on a FreeBSD
1686 system running IPv6 was broken. This may well have applied to all BSD
1687 systems, as well as to others that have similar system calls. The broken
1688 code found IPv4 interfaces correctly, but gave incorrect values for the
1689 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1690 that it would not match correctly against @[] and not recognize the IPv6
1693 PH/02 The ipliteral router was not recognizing addresses of the form user@
1694 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1696 PH/03 Added disable_ipv6.
1698 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1699 decoded form, because it is most often used to construct To: headers
1700 lines in autoreplies, and the decoded form may well be syntactically
1701 invalid. However, $reply_address has leading white space removed, and all
1702 newlines turned into spaces so that the autoreply transport does not
1705 PH/05 If group was specified without a user on a router, and no group or user
1706 was specified on a transport, the group from the router was ignored.
1708 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1709 for visible compile-time settings that can be used to change these
1710 numbers, for those that want even more. Backwards compatibility with old
1711 spool files has been maintained. However, going back to a previous Exim
1712 release will lost any variables that are in spool files.
1714 PH/07 Two small changes when running in the test harness: increase delay when
1715 passing a TCP/IP connection to a new process, in case the original
1716 process has to generate a bounce, and remove special handling of
1717 127.0.0.2 (sic), which is no longer necessary.
1719 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1720 be the same on different OS.
1722 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1725 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1726 whether --show-vars was specified or not
1728 JJ/02 exipick: Added support for new ACL variable spool format introduced
1731 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1732 syntactically invalid From: or Reply-to: line, and a filter used this to
1733 generate an autoreply, and therefore failed to obtain an address for the
1734 autoreply, Exim could try to deliver to a non-existent relative file
1735 name, causing unrelated and misleading errors. What now happens is that
1736 it logs this as a hard delivery error, but does not attempt to create a
1739 PH/11 The exinext utility has a -C option for testing purposes, but although
1740 the given file was scanned by exinext itself; it wasn't being passed on
1741 when Exim was called.
1743 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1744 an end-of-file indication when reading a command response.
1746 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1747 compiled. In many other places in Exim, IPv6 addresses are always
1748 recognized, so I have changed this. It also means that IPv4 domain
1749 literals of the form [IPV4:n.n.n.n] are now always recognized.
1751 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1752 used if the router is not running as root, for example, when verifying at
1753 ACL time, or when using -bh. The debugging output from this situation was
1754 non-existent - all you got was a failure to exec. I have made two
1757 (a) Failures to set uid/gid, the current directory, or a process leader
1758 in a subprocess such as that created by queryprogram now generate
1759 suitable debugging ouput when -d is set.
1761 (b) The queryprogram router detects when it is not running as root,
1762 outputs suitable debugging information if -d is set, and then runs
1763 the subprocess without attempting to change uid/gid.
1765 PH/15 Minor change to Makefile for building test_host (undocumented testing
1768 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1769 additional section of a DNS packet that returns MX or SRV records.
1770 Instead, it always explicitly searches for A/AAAA records. This avoids
1771 major problems that occur when a DNS server includes only records of one
1772 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1773 fixed another bug: if SRV records were looked up and the corresponding
1774 address records were *not* found in the additional section, the port
1775 values from the SRV records were lost.
1777 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1778 using the correct key (the original address) when searching the retry
1779 rules in order to find which one to use for generating the retry hint.
1781 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1782 adding the default one. Similarly, if it contains a Reply-To: header, the
1783 errors_reply_to option, if set, is not used.
1785 PH/19 When calculating a retry time, Exim used to measure the "time since
1786 failure" by looking at the "first failed" field in the retry record. Now
1787 it does not use this if it is later than than the arrival time of the
1788 message. Instead it uses the arrival time. This makes for better
1789 behaviour in cases where some deliveries succeed, thus re-setting the
1790 "first failed" field. An example is a quota failure for a huge message
1791 when small messages continue to be delivered. Without this change, the
1792 "time since failure" will always be short, possible causing more frequent
1793 delivery attempts for the huge message than are intended.
1794 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1796 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1797 $1, $2, $3) because the numerical variables can be reset during some
1798 expansion items (e.g. "match"), thereby losing the authentication data.
1800 PH/21 Make -bV show the size of off_t variables so that the test suite can
1801 decide whether to run tests for quotas > 2G.
1803 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1804 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1805 filecount value is greater than 2G or if a quota value is greater than 2G
1806 on a system where the size of off_t is not greater than 4, a panic error
1809 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1810 never match. The debug and -bh output now contains an explicit error
1811 message indicating a malformed IPv4 address or mask.
1813 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1814 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1815 PH/23 above applies.
1817 PH/25 Do not write to syslog when running in the test harness. The only
1818 occasion when this arises is a failure to open the main or panic logs
1819 (for which there is an explicit test).
1821 PH/26 Added the /no_tell option to "control=freeze".
1823 PH/27 If a host name lookup failed very early in a connection, for example, if
1824 the IP address matched host_lookup and the reverse lookup yielded a name
1825 that did not have a forward lookup, an error message of the form "no IP
1826 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1827 could be logged. Now it outputs the IP address instead of "NULL".
1829 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1830 allows the sender and the authenticated sender to be set when
1831 submitting a message from within Exim. Since child_open_exim() is
1832 documented for local_scan(), the new function should be too.
1834 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1835 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1836 results in an empty string is now treated as unset.
1838 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1840 PH/31 Added sender_verify_fail logging option.
1842 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1843 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1844 tidied the source and removed it altogether.
1846 PH/33 When a queue run was abandoned because the load average was too high, a
1847 log line was always written; now it is written only if the queue_run log
1848 selector is set. In addition, the log line for abandonment now contains
1849 information about the queue run such as the pid. This is always present
1850 in "start" and "stop" lines but was omitted from the "abandon" line.
1852 PH/34 Omit spaces between a header name and the colon in the error message that
1853 is given when verify = headers_syntax fails (if there are lots of them,
1854 the message gets confusing).
1856 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1857 names, as there are now some PTR records that contain slashes. This check
1858 is only to protect against broken name servers that fall over on strange
1859 characters, so the fact that it applies to all lookups doesn't matter.
1861 PH/36 Now that the new test suite is complete, we can remove some of the
1862 special code in Exim that was needed for the old test suite. For example,
1863 sorting DNS records because real resolvers return them in an arbitrary
1864 order. The new test suite's fake resolver always returns records in the
1867 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1868 bounces) except when queue_only is set, to avoid logging races between
1869 the different processes.
1871 PH/38 Panic-die if .include specifies a non-absolute path.
1873 PH/39 A tweak to the "H" retry rule from its user.
1875 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1876 a label. They prevented compilation on older perls.
1878 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1879 a warning to be raised on newish perls.
1881 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1882 on queue. Changes to match documented behaviour of showing count of
1883 messages matching specified criteria.
1885 PH/40 Changed the default ident timeout from 30s to 5s.
1887 PH/41 Added support for the use of login_cap features, on those BSD systems
1888 that have them, for controlling the resources used by pipe deliveries.
1890 PH/42 The content-scanning code uses fopen() to create files in which to put
1891 message data. Previously it was not paying any attention to the mode of
1892 the files. Exim runs with umask(0) because the rest of the code creates
1893 files with open(), and sets the required mode explicitly. Thus, these
1894 files were ending up world-writeable. This was not a big issue, because,
1895 being within the spool directory, they were not world-accessible. I have
1896 created a function called modefopen, which takes an additional mode
1897 argument. It sets umask(777), creates the file, chmods it to the required
1898 mode, then resets the umask. All the relevant calls to fopen() in the
1899 content scanning code have been changed to use this function.
1901 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1902 to 24 hours. This avoids potential overflow problems when processing G
1903 and H retry rules. I suspect nobody ever tinkers with this value.
1905 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1907 PH/45 When the plaintext authenticator is running as a client, the server's
1908 challenges are checked to ensure they are valid base64 strings. By
1909 default, the authentication attempt is cancelled if an invalid string is
1910 received. Setting client_ignore_invalid_base64 true ignores these errors.
1911 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1912 they are received. Thus, the responses can be made to depend on the
1913 challenges. If an invalid string is ignored, an empty string is placed in
1916 PH/46 Messages that are created by the autoreply transport now contains a
1917 References: header, in accordance with RFCs 2822 and 3834.
1919 PH/47 Added authenticated_sender_force to the smtp transport.
1921 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1923 PH/49 Installed latest patch from the Sieve maintainer.
1925 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1926 was also set, the appendfile transport was unnecessarily scanning a
1927 directory of message files (e.g. for maildir delivery) to find the count
1928 of files (along with the size), even though it did not need this
1929 information. It now does the scan only if it needs to find either the
1930 size of the count of files.
1932 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1934 PH/52 Two bugs concerned with error handling when the smtp transport is
1937 (i) Exim was not creating retry information for temporary errors given
1938 for individual recipients after the DATA command when the smtp transport
1939 was used in LMTP mode. This meant that they could be retried too
1940 frequently, and not timed out correctly.
1942 (ii) Exim was setting the flag that allows error details to be returned
1943 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1944 recipients that were returned after the DATA command.
1946 PH/53 This is related to PH/52, but is more general: for any failing address,
1947 when detailed error information was permitted to be returned to the
1948 sender, but the error was temporary, then after the final timeout, only
1949 "retry timeout exceeded" was returned. Now it returns the full error as
1950 well as "retry timeout exceeded".
1952 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1953 do this, and (what is worse) MTAs that accept it.
1955 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1956 will now be deprecated.
1958 PH/56 New os.c-cygwin from the Cygwin maintainer.
1960 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1961 formats (previously only available in exim formats via -bpr, -bpru,
1962 and -bpra. Now also available in native and exiqgrep formats)
1964 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1965 with very large, slow to parse queues
1967 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1969 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1971 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1972 responses to authentication challenges, though it was showing the
1973 challenges; (ii) I've removed the CR characters from the debug output for
1976 PH/58 Allow for the insertion of a newline as well as a space when a string
1977 is turned into more than one encoded-word during RFC 2047 encoding. The
1978 Sieve code now uses this.
1980 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1981 data_4xx, lost_connection, tls_required.
1983 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1984 message was being sent as an SMTP response.
1986 PH/61 Add -l and -k options to exicyclog.
1988 PH/62 When verifying, if an address was redirected to one new address, so that
1989 verification continued, and the new address failed or deferred after
1990 having set something in $address_data, the value of $address_data was not
1991 passed back to the ACL. This was different to the case when no
1992 redirection occurred. The value is now passed back in both cases.
1994 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1995 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1996 use login_cap.h, so on its own it isn't the distinguishing feature. The
1997 new name refers directly to the setclassresources() function.
1999 PH/65 Added configuration files for NetBSD3.
2001 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2003 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2004 is preferred over IPv4.
2006 PH/68 The bounce_return_message and bounce_return_body options were not being
2007 honoured for bounces generated during the reception of non-SMTP messages.
2008 In particular, this applied to messages rejected by the ACL. This bug has
2009 been fixed. However, if bounce_return_message is true and bounce_return_
2010 body is false, the headers that are returned for a non-SMTP message
2011 include only those that have been read before the error was detected.
2012 (In the case of an ACL rejection, they have all been read.)
2014 PH/69 The HTML version of the specification is now built in a directory called
2015 spec_html instead of spec.html, because the latter looks like a path with
2016 a MIME-type, and this confuses some software.
2018 PH/70 Catch two compiler warnings in sieve.c.
2020 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2021 function verify_get_ident() calls ip_connect() to connect a socket, but
2022 if the "connect()" function timed out, ip_connect() used to close the
2023 socket. However, verify_get_ident() also closes the socket later, and in
2024 between Exim writes to the log, which may get opened at this point. When
2025 the socket was closed in ip_connect(), the log could get the same file
2026 descriptor number as the socket. This naturally causes chaos. The fix is
2027 not to close the socket in ip_connect(); the socket should be closed by
2028 the function that creates it. There was only one place in the code where
2029 this was missing, in the iplookup router, which I don't think anybody now
2030 uses, but I've fixed it anyway.
2032 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2033 well as to direct DNS lookups. Otherwise the handling of names in host
2034 lists is inconsistent and therefore confusing.
2040 PH/01 Two changes to the default runtime configuration:
2042 (1) Move the checks for relay_from_hosts and authenticated clients from
2043 after to before the (commented out) DNS black list checks.
2045 (2) Add control=submission to the relay_from_hosts and authenticated
2046 clients checks, on the grounds that messages accepted by these
2047 statements are most likely to be submissions.
2049 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2051 (1) Generate an error if the third argument for the ${prvs expansion is
2054 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2057 (3) Reset the variables that are obtained from the first argument of
2058 ${prvscheck and used in the second argument before leaving the code,
2059 because their memory is reclaimed, so using them afterwards may do
2062 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2063 one (it's much easier than Tom thought :-).
2065 (5) Because of (4), we can now allow for the use of $prvscheck_result
2066 inside the third argument.
2068 PH/03 For some reason, the default setting of PATH when running a command from
2069 a pipe transport was just "/usr/bin". I have changed it to
2072 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2073 anything to be listed in the output from -bV.
2075 PH/05 When a filter generated an autoreply, the entire To: header line was
2076 quoted in the delivery log line, like this:
2078 => >A.N.Other <ano@some.domain> <original@ddress> ...
2080 This has been changed so that it extracts the operative address. There
2081 may be more than one such address. If so, they are comma-separated, like
2084 => >ano@some.domain,ona@other.domain <original@ddress> ...
2086 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2087 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2088 not being looked up in the rDNS to get a host name, Exim was showing the
2089 IP address twice in Received: lines, even though the IP addresses were
2090 identical. For example:
2092 Received: from [1.2.3.4] (helo=[1.2.3.4])
2094 However, if the real host name was known, it was omitting the HELO data
2095 if it matched the actual IP address. This has been tidied up so that it
2096 doesn't show the same IP address twice.
2098 PH/07 When both +timestamp and +memory debugging was on, the value given by
2099 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2100 called by the debug printing, thereby overwriting the timestamp buffer.
2101 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2103 PH/08 When the original message was included in an autoreply transport, it
2104 always said "this is a copy of the message, including all the headers",
2105 even if body_only or headers_only was set. It now gives an appropriate
2108 PH/09 Applied a patch from the Sieve maintainer which:
2110 o fixes some comments
2111 o adds the (disabled) notify extension core
2112 o adds some debug output for the result of if/elsif tests
2113 o points to the current vacation draft in the documentation
2114 and documents the missing references header update
2118 o fixes a bug in processing the envelope test (when testing
2119 multiple envelope elements, the last element determinted the
2122 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2123 Electronic Mail") by including:
2125 Auto-submitted: auto-generated
2127 in the messages that it generates (bounce messages and others, such as
2128 warnings). In the case of bounce messages for non-SMTP mesages, there was
2129 also a typo: it was using "Auto_submitted" (underscore instead of
2130 hyphen). Since every message generated by Exim is necessarily in response
2131 to another message, thes have all been changed to:
2133 Auto-Submitted: auto-replied
2135 in accordance with these statements in the RFC:
2137 The auto-replied keyword:
2139 - SHOULD be used on messages sent in direct response to another
2140 message by an automatic process,
2142 - MUST NOT be used on manually-generated messages,
2144 - MAY be used on Delivery Status Notifications (DSNs) and Message
2145 Disposition Notifications (MDNs),
2147 - MUST NOT be used on messages generated by automatic or periodic
2148 processes, except for messages which are automatic responses to
2151 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2152 to the default Received: header definition.
2154 PH/12 Added log selector acl_warn_skipped (default on).
2156 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2157 variables because (a) they are in the wrong storage pool and (b) even if
2158 they were copied, it wouldn't work properly because of the caching.
2160 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2161 checking when decoding. Apparently there are clients that generate
2162 overlong encoded strings. Why am I not surprised?
2164 PH/15 If the first argument of "${if match_address" was not empty, but did not
2165 contain an "@" character, Exim crashed. Now it writes a panic log message
2166 and treats the condition as false.
2168 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2170 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2171 extension "envelope-auth". The code is finished and in agreement with
2172 other implementations, but there is no documentation so far and in fact,
2173 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2174 not changing the active code.
2176 Print executed "if" and "elsif" statements when debugging is used. This
2177 helps a great deal to understand what a filter does.
2179 Document more things not specified clearly in RFC3028. I had all this
2180 sorted out, when out of a sudden new issues came to my mind. Oops."
2182 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2185 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2186 canonical form (as documented). However, after a host name lookup from
2187 the IP address, check_host() was doing a simple string comparison with
2188 addresses acquired from the DNS when checking that the found name did
2189 have the original IP as one of its addresses. Since any found IPv6
2190 addresses are likely to be in abbreviated form, the comparison could
2191 fail. Luckily, there already exists a function for doing the comparison
2192 by converting both addresses to binary, so now that is used instead of
2193 the text comparison.
2195 PH/20 There was another similar case to PH/19, when a complete host name was
2196 given in a host list; looking up its IP address could give an abbreviated
2197 form, whereas the current host's name might or might not be abbreviated.
2198 The same fix has been applied.
2204 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2205 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2208 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2209 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2211 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2213 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2214 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2215 Also, for "Auto-Submitted", treat anything other than "no" as
2216 non-personal, in accordance with RFC 3834. (Previously it treated
2217 anything starting "auto-" as non-personal.)
2219 TF/01 The control=submission/name=... option had a problem with syntax
2220 errors if the name included a slash character. The /name= option
2221 now slurps the rest of the string, so it can include any characters
2222 but it must come last in the list of options (after /sender_retain
2225 PH/05 Some modifications to the interface to the fake nameserver for the new
2233 TK/01 Added the "success_on_redirect" address verification option. See
2234 NewStuff for rationale and an example.
2236 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2238 PH/02 Patch to exigrep to allow it to work on syslog lines.
2240 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2241 fread() to skip over the body file's header line, because in Cygwin the
2242 header line is locked and is inaccessible.
2244 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2245 co-exist for some time) to make it clear that it is the Exim ID that is
2246 referenced, not the Message-ID: header line.
2248 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2249 string_format() function, because snprintf() does not exist on all
2252 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2253 ${stat: expansion item.
2255 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2256 protocol synchronization error", to keep the pedants happy.
2258 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2259 well as for IRIX systems, when gcc is being used. See the host.c source
2262 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2264 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2267 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2268 transport and to the smtp transport in LMTP mode.
2270 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2272 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2273 run a filter in a subprocess. This could lead to confusion in subsequent
2274 lookups in the parent process. There should also be a search_tidyup() at
2275 the end of the subprocess.
2277 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2278 only if the host matched helo_try_verify_hosts, which caused the
2279 verification to occur when the EHLO/HELO command was issued. The ACL just
2280 tested the remembered result. Now, if a previous verification attempt has
2281 not happened, "verify = helo" does it there and then.
2283 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2285 TK/03 Fix log output including CR from clamd.
2287 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2288 exist provoked a memory error which could cause a segfault.
2290 PH/15 Installed PCRE 6.2
2292 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2294 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2295 of the problem. Specifically, suggested +O2 rather than +O1 for the
2298 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2300 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2301 record, the port was not correctly logged when the outgoing_port log
2302 selector was set (it logged the transort's default port).
2304 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2305 fallback_hosts, and "hosts" in the smtp transport.
2307 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2308 host errors such as "Connection refused".
2310 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2311 authentication with radiusclient 0.4.9:
2313 - Error returned from rc_read_config was caught wrongly
2314 - Username/password not passed on to radius server due to wrong length.
2316 The presumption is that some radiusclient API changes for 4.51/PH/17
2317 were not taken care of correctly. The code is still untested by me (my
2318 Linux distribution still has 0.3.2 of radiusclient), but it was
2319 contributed by a Radius user.
2321 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2322 expanding the "port" option of the smtp transport.
2324 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2325 while reading a MIME header. Thanks to Tom Hughes for a patch.
2327 PH/24 Include config.h inside local_scan.h so that configuration settings are
2330 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2331 that in an ACL for RCPT (for example), you can examine exactly what was
2334 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2335 commands, but it was not correctly comparing the address with the actual
2336 client host address. Thus, it would show the EHLO address in Received:
2337 header lines when this was not necessary.
2339 PH/27 Added the % operator to ${eval:}.
2341 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2342 it should be ignoring failures (because with -C, for example, it has lost
2343 privilege). It wasn't ignoring creation failures other than "already
2346 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2349 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2350 because an input line was too long, either on its own, or by virtue of
2351 too many continuations, the temporary file was not being removed, and the
2352 return code was incorrect.
2354 PH/31 Missing "BOOL" in function definition in filtertest.c.
2356 PH/32 Applied Sieve patches from the maintainer.
2358 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2360 PH/33 Added "verify = not_blind".
2362 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2363 Local/Makefile (with some defaults set). These are used in built scripts
2364 such as exicyclog, but they have never been used in the exim_install
2365 script (though there are many overriding facilities there). I have
2366 arranged that the exim_install script now takes note of these two
2369 PH/35 Installed configuration files for Dragonfly.
2371 PH/36 When a locally submitted message by a trusted user did not contain a
2372 From: header, and the sender address was obtained from -f or from an SMTP
2373 MAIL command, and the trusted user did not use -F to supply a sender
2374 name, $originator_name was incorrectly used when constructing a From:
2375 header. Furthermore, $originator_name was used for submission mode
2376 messages from external hosts without From: headers in a similar way,
2377 which is clearly wrong.
2379 PH/37 Added control=suppress_local_fixups.
2381 PH/38 When log_selector = +received_sender was set, and the addition of the
2382 sender made the log line's construction buffer exactly full, or one byte
2383 less than full, an overflow happened when the terminating "\n" was
2386 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2387 when the result of a list match is failure because a DNS lookup failed.
2389 PH/40 RM_COMMAND is now used in the building process.
2391 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2392 the "build-* directories that it finds.
2394 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2395 address in a domain literal was a prefix of an interface address.
2397 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2398 when verifying a sender address, unless rewrite_headers is false.
2400 PH/44 Wrote a long comment about why errors_to addresses are verified as
2401 recipients, not senders.
2403 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2404 the ratelimit ACL was added.
2406 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2408 PH/46 Added extra information about PostgreSQL errors to the error string.
2410 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2411 avoiding the need to install special zones in a real server. This is
2412 backwards compatible; if it can't find the fake resolver, it drops back.
2413 Thus, both old and new test suites can be run.
2415 TF/02 Added util/ratelimit.pl
2417 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2418 clock is set back in time.
2420 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2421 Candler <B.Candler@pobox.com>.
2423 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2424 OK for addresses that are the result of redirections.
2426 PH/48 A number of further additions for the benefit of the new test suite,
2427 including a fake gethostbyname() that interfaces to the fake DNS resolver
2430 TF/06 The fix for widen_domains has also been applied to qualify_single and
2431 search_parents which are the other dnslookup options that can cause
2434 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2437 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2439 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2440 service" if it encountered an MX record with an empty target hostname.
2441 The message is now "an MX or SRV record indicated no SMTP service".
2443 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2444 if the DNS of the sending site is misconfigured. This is quite a
2445 common situation. This change restores the behaviour of treating a
2446 helo verification defer as a failure.
2448 PH/51 If self=fail was set on a router, the bounce message did not include the
2449 actual error message.
2455 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2457 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2458 command itself ended in error, the underlying message about the transport
2459 filter timeout was being overwritten with the pipe command error. Now the
2460 underlying error message should be appended to the second error message.
2462 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2464 PH/02 Reduce the amount of output that "make" produces by default. Full output
2465 can still be requested.
2467 PH/03 The warning log line about a condition test deferring for a "warn" verb
2468 was being output only once per connection, rather than after each
2469 occurrence (because it was using the same function as for successful
2470 "warn" verbs). This seems wrong, so I have changed it.
2472 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2473 it should not have, which might have caused a crash in the right
2474 circumstances, but probably never did.
2476 PH/04 Installed a modified version of Tony Finch's patch to make submission
2477 mode fix the return path as well as the Sender: header line, and to
2478 add a /name= option so that you can make the user's friendly name appear
2481 TF/03 Added the control = fakedefer ACL modifier.
2483 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2484 Mark Lowes for thorough testing.
2486 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2488 TK/03 Merged latest SRS patch from Miles Wilton.
2490 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2491 with the definition in sysexits.h (which is #included earlier).
2492 Fortunately, Exim does not actually use EX_OK. The code used to try to
2493 preserve the sysexits.h value, by assumimg that macro definitions were
2494 scanned for macro replacements. I have been disabused of this notion,
2495 so now the code just undefines EX_OK before #including unistd.h.
2497 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2498 in the smtp transport. When a block could not be written in a single
2499 write() function, the timeout was being re-applied to each part-write.
2500 This seems wrong - if the receiver was accepting one byte at a time it
2501 would take for ever. The timeout is now adjusted when this happens. It
2502 doesn't have to be particularly precise.
2504 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2505 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2507 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2508 without a domain if the check to <postmaster@domain> fails.
2510 SC/01 Eximstats: added -xls and the ability to specify output files
2511 (patch written by Frank Heydlauf).
2513 SC/02 Eximstats: use FileHandles for outputing results.
2515 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2517 SC/04 Eximstats: fixed display of large numbers with -nvr option
2519 SC/05 Eximstats: fixed merging of reports with empty tables.
2521 SC/06 Eximstats: added the -include_original_destination flag
2523 SC/07 Eximstats: removed tabs and trailing whitespace.
2525 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2527 TK/06 MBOX spool code: Add real "From " MBOX separator line
2528 so the .eml file is really in mbox format (even though
2529 most programs do not really care). Patch from Alex Miller.
2531 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2532 The latter is generated from $received_to and is only set if the
2533 message has one envelope recipient. SA can use these headers,
2534 obviously out-of-the-box. Patch from Alex Miller.
2536 PH/08 The ${def test on a variable was returning false if the variable's
2537 value was "0", contrary to what the specification has always said!
2538 The result should be true unless the variable is empty.
2540 PH/09 The syntax error of a character other than { following "${if
2541 def:variable_name" (after optional whitespace) was not being diagnosed.
2542 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2543 accidental colon was present, for example, could give incorrect results.
2545 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2546 result is used (not including appendfile, where other changes are about
2549 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2550 This involved changing a lot of size variables from int to off_t. It
2551 should work with maildirs and everything.
2553 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2554 spamd dying while we are connected to it.
2556 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2559 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2560 function for BDB 4.3.
2562 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2564 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2565 the sources for PCRE have been re-arranged and more files are now
2568 PH/15 The code I had for printing potentially long long variables in PH/11
2569 above was not the best (it lost precision). The length of off_t variables
2570 is now inspected at build time, and an appropriate printing format (%ld
2571 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2572 to be "long long int" or "long int". This is needed for the internal
2573 formatting function string_vformat().
2575 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2576 the configuration file to be ":syslog", then the script "guesses" where
2577 the logs files are, rather than using the compiled in default. In our
2578 case the guess is not the same as the compiled default, so the script
2579 suddenly stopped working when I started to use syslog. The patch checks
2580 to see if log_file_path is "". If so, it attempts to read it from exim
2581 with no configuration file to get the compiled in version, before it
2582 falls back to the previous guessing code."
2584 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2585 implementing BATV in an Exim configuration. See NewStuff for the gory
2588 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2589 Makefile that are specific to HP-UX.
2591 PH/18 If the "use_postmaster" option was set for a recipient callout together
2592 with the "random" option, the postmaster address was used as the MAIL
2593 FROM address for the random test, but not for the subsequent recipient
2594 test. It is now used for both.
2596 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2597 patch removes a few documentation additions to RFC 3028, because the
2598 latest draft now contains them. It adds the new en;ascii-case comparator
2599 and a new error check for 8bit text in MIME parts. Comparator and
2600 require names are now matched exactly. I enabled the subaddress
2601 extension, but it is not well tested yet (read: it works for me)."
2603 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2604 rework some of the code of TK/09 above to avoid the hardwired use of
2605 "%lld" and "long long". Replaced the call to snprintf() with a call to
2608 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2609 records point to non-existent hosts", "retry timeout exceeded", and
2610 "retry time not reached for any host after a long failure period".
2612 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2613 experimental DomainKeys support:
2615 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2616 (2) On an error such as an illegally used "control", the wrong name for
2617 the control was given.
2619 These problems did NOT occur unless DomainKeys support was compiled.
2621 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2623 PH/24 Added ${if match_ip condition.
2625 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2626 databases so that it will be absolutely obvious if a crash occurs in the
2627 DB library. This is a regular occurrence (often caused by mis-matched
2630 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2631 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2632 picked up on a user's system that detects such things. There doesn't seem
2633 to be a gcc warning option for this - only an attribute that has to be
2634 put on the function's prototype. It seems that in Fedora Core 4 they have
2635 set this on a number of new functions. No doubt there will be more in due
2638 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2639 specify a transport. However, if an address that was verified by such a
2640 router was the subject of a callout, Exim crashed because it tried to
2641 read the rcpt_include_affixes from the non-existent transport. Now it
2642 just assumes that the setting of that option is false. This bug was
2643 introduced by 4.51/PH/31.
2645 PH/28 Changed -d+all to exclude +memory, because that information is very
2646 rarely of interest, but it makes the output a lot bigger. People tend to
2647 do -d+all out of habit.
2649 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2650 code in os-type was giving problems when libc.so lives in lib64, like on
2653 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2654 aren't the modern standard, and it seems that some systems' include files
2655 don't always have them. Exim was already checking for some of the newer
2656 ones like T_AAAA, and defining it itself. I've added checks for all the
2657 record types that Exim uses.
2659 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2660 not automatically generating a new one, as it is supposed to. This
2661 prevented TLS from working. If the file did exist, but contained invalid
2662 data, a new version was generated, as expected. It was only the case of a
2663 non-existent file that was broken.
2665 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2666 with a change in libdomainkeys > 0.64.
2668 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2669 from DNS. If the selector record carries the flag, it now has
2670 precedence over the domain-wide flag.
2672 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2674 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2675 the use of a port name that isn't defined in /etc/services) occurred, the
2676 message was deferred as in a normal delivery, and thus remained on the
2677 spool, instead of being failed because of the mua_wrapper setting. This
2678 is now fixed, and I tidied up some of the mua_wrapper messages at the
2681 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2682 the messages in a hash of arrays rather than using individual hashes.
2683 This is a bit cleaner and results in dramatic memory savings, albeit
2684 at a slight CPU cost.
2686 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2687 as requested by Marc Sherman.
2689 SC/10 Eximstats: added histograms for user specified patterns as requested
2692 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2694 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2695 fopen() in the content-scanning modules that did not already have it.
2701 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2702 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2704 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2706 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2708 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2709 to test Sieve filters that use "vacation".
2711 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2712 that changes the way the GnuTLS parameters are stored in the cache file.
2713 The new format can be generated externally. For backward compatibility,
2714 if the data in the cache doesn't make sense, Exim assumes it has read an
2715 old-format file, and it generates new data and writes a new file. This
2716 means that you can't go back to an older release without removing the
2719 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2720 work if there are any delivery delays because "one_time" forces the
2721 parent to be marked "delivered", so its unseen clone is never tried
2722 again. For this reason, Exim now forbids the simultaneous setting of
2725 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2726 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2727 entry if you want to know the details. The fix, however, neglected to
2728 consider the case when local delivery batching is involved. The test for
2729 "previously delivered" was not happening when checking to see if an
2730 address could be batched with a previous (undelivered) one; under
2731 certain circumstances this could lead to multiple deliveries to the same
2734 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2735 in its include files, and this causes problems building Exim.
2737 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2738 header_syntax) but Exim was just ignoring anything given after a slash.
2739 In particular, this caused confusion with an attempt to use "verify =
2740 reverse_host_lookup/defer_ok". An error is now given when options are
2741 supplied for verify items that do not have them. (Maybe reverse_host_
2742 lookup should have a defer_ok option, but that's a different point.)
2744 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2745 defined by RFC 821) to 2048, because there were problems with some AUTH
2746 commands, and RFC 1869 says the size should be increased for extended
2747 SMTP commands that take arguments.
2749 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2752 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2753 "unknown" error; now it says that the functionality isn't in the binary.
2755 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2756 an address' error message when a string expansion fails (syntax or
2757 whatever). Otherwise the password may appear in the log. Following change
2758 PH/42 below, there is no longer a chance of it appearing in a bounce
2761 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2763 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2764 of its addresses was ever tried. (Bugzilla bug #2.)
2766 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2767 the result incorrectly in the debug output. (It correctly added a newline
2768 to what was transported.)
2770 TF/01 Added $received_time.
2772 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2773 commented out examples of how to interface to a virus scanner and to
2774 SpamAssassin. Also added commented examples of av_scanner and
2775 spamd_address settings.
2777 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2778 and controls are allowed in which ACLs. There were a couple of minor
2779 errors. Some of the entries in the conditions table (which is a table of
2780 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2781 negation of where the condition IS allowed.
2783 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2785 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2786 header file does not have a version number, so I've had to invent a new
2787 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2788 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2789 radiusclient), but it was contributed by a Radius user.
2791 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2792 files or directories, for OpenSSL.
2794 PH/19 When an Exim process that is running as root has to create an Exim log
2795 file, it does so in a subprocess that runs as exim:exim so as to get the
2796 ownership right at creation (otherwise, other Exim processes might see
2797 the file with the wrong ownership). There was no test for failure of this
2798 fork() call, which would lead to the process getting stuck as it waited
2799 for a non-existent subprocess. Forks do occasionally fail when resources
2800 run out. I reviewed all the other calls to fork(); they all seem to check
2803 PH/20 When checking for unexpected SMTP input at connect time (before writing
2804 the banner), Exim was not dealing correctly with a non-positive return
2805 from the read() function. If the client had disconnected by this time,
2806 the result was a log entry for a synchronization error with an empty
2807 string after "input=" when read() returned zero. If read() returned -1
2808 (an event I could not check), uninitialized data bytes were printed.
2809 There were reports of junk text (parts of files, etc) appearing after
2812 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2814 PH/22 Added support for macro redefinition, and (re)definition in between
2815 driver and ACL definitions.
2817 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2818 forgetting to use the resulting value; it was using the unexpanded value.
2820 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2821 hadn't been configured. The fix is from Juergen Kreileder, who
2822 understands it better than I do:
2824 "Here's what I see happening with three configured cyrus_sasl
2825 authenticators configured (plain, login, cram-md5):
2827 On startup auth_cyrus_sasl_init() gets called for each of these.
2828 This means three calls to sasl_listmech() without a specified mech_list.
2829 => SASL tests which mechs of all available mechs actually work
2830 => three warnings about OTP not working
2831 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2833 With the patch, sasl_listmech() also gets called three times. But now
2834 SASL's mech_list option is set to the server_mech specified in the the
2835 authenticator. Or in other words, the answer from sasl_listmech()
2836 gets limited to just the mech you're testing for (which is different
2838 => the return list contains just 'plain' or 'login', 'cram-md5' or
2839 nothing depending on the value of ob->server_mech.
2841 I've just tested the patch: Authentication still works fine,
2842 unavailable mechs specified in the exim configuration are still
2843 caught, and the auth.log warnings about OTP are gone."
2845 PH/25 When debugging is enabled, the contents of the command line are added
2846 to the debugging output, even when log_selector=+arguments is not
2849 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2850 answer is "GNU", and only if the return is "GNU/something" is the answer
2853 PH/27 $acl_verify_message is now set immediately after the failure of a
2854 verification in an ACL, and so is available in subsequent modifiers. In
2855 particular, the message can be preserved by coding like this:
2857 warn !verify = sender
2858 set acl_m0 = $acl_verify_message
2860 Previously, $acl_verify_message was set only while expanding "message"
2861 and "log_message" when a very denied access.
2863 PH/28 Modified OS/os.c-Linux with
2865 -#ifndef OS_LOAD_AVERAGE
2866 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2868 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2869 nomenclature these days.)
2871 PH/29 Installed patch from the Sieve maintainer that adds the options
2872 sieve_useraddress and sieve_subaddress to the redirect router.
2874 PH/30 In these circumstances:
2875 . Two addresses routed to the same list of hosts;
2876 . First host does not offer TLS;
2877 . First host accepts first address;
2878 . First host gives temporary error to second address;
2879 . Second host offers TLS and a TLS session is established;
2880 . Second host accepts second address.
2881 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2882 and peerdn, if requested) that were in fact used only for the second
2885 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2886 attention to any local part prefix or suffix that was matched by the
2887 router that accepted the address. It now behaves in the same way as it
2888 does for delivery: the affixes are removed from the local part unless
2889 rcpt_include_affixes is set on the transport.
2891 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2892 timeout during the DATA phase of an incoming message.
2894 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2895 applied a patch sanctioned by the Sieve maintainer.
2897 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2898 the uid or gid is negative. A case of a negative gid caused this to be
2899 noticed. The fix allows for either to be negative.
2901 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2902 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2905 PH/36 The change PH/12 above was broken. Fixed it.
2907 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2908 the grounds that routing the same address twice would always produce the
2909 same answer. This might have been true once, but it is certainly no
2910 longer true now. Routing a child address may depend on the previous
2911 routing that produced that child. Some complicated redirection strategies
2912 went wrong when messages had multiple recipients, and made Exim's
2913 behaviour dependent on the order in which the addresses were given.
2915 I have moved the duplicate checking until after the routing is complete.
2916 Exim scans the addresses that are assigned to local and remote
2917 transports, and removes any duplicates. This means that more work will be
2918 done, as duplicates will always all be routed, but duplicates are
2919 presumably rare, so I don't expect this is of any significance.
2921 For deliveries to pipes, files, and autoreplies, the duplicate checking
2922 still happens during the routing process, since they are not going to be
2925 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2926 It corrects a timeout issue with spamd. This is Ian's comment: "The
2927 background is that sometimes spamd either never reads data from a
2928 connection it has accepted, or it never writes response data. The exiscan
2929 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2930 blindly assumes that writes won't block so it may never time out."
2932 PH/39 Allow G after quota size as well as K and M.
2934 PH/40 The value set for $authenticated_id in an authenticator may not contain
2935 binary zeroes or newlines because the value is written to log lines and
2936 to spool files. There was no check on this. Now the value is run through
2937 the string_printing() function so that such characters are converted to
2938 printable escape sequences.
2940 PH/41 $message_linecount is a new variable that contains the total number of
2941 lines in the message. Compare $body_linecount, which is the count for the
2944 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2945 bounce and delay warning messages, except in certain special cases, which
2948 (a) An SMTP error message from a remote host;
2949 (b) A message specified in a :fail: redirection;
2950 (c) A message specified in a "fail" command in a system filter;
2951 (d) A message specified in a FAIL return from the queryprogram router;
2952 (e) A message specified by the cannot_route_message router option.
2954 In these cases only, Exim does include the error details in bounce and
2955 warning messages. There are also a few cases where bland messages such
2956 as "unrouteable address" or "local delivery error" are given.
2958 PH/43 $value is now also set for the "else" part of a ${run expansion.
2960 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2961 being worked on, but at least Exim now implements the latest version to
2964 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2965 process to complete was treated as a delivery failure, a timeout while
2966 writing the message to the pipe was logged, but erroneously treated as a
2967 successful delivery. Such timeouts include transport filter timeouts. For
2968 consistency with the overall process timeout, these timeouts are now
2969 treated as errors, giving rise to delivery failures by default. However,
2970 there is now a new Boolean option for the pipe transport called
2971 timeout_defer, which, if set TRUE, converts the failures into defers for
2972 both kinds of timeout. A transport filter timeout is now identified in
2975 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2976 systems where "make" and "gmake" are different, calling "gmake" at top
2977 level broke things. I've arranged for the value of $(MAKE) to be passed
2978 from the Makefile to this script so that it can call the same version of
2982 A note about Exim versions 4.44 and 4.50
2983 ----------------------------------------
2985 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2986 changes of various kinds. As a consequence, a big documentation update was
2987 needed. This delayed the release for rather longer than seemed good, especially
2988 in the light of a couple of (minor) security issues. Therefore, the changes
2989 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2990 release. So 4.44 and 4.50 are in effect two different branches that both start
2993 I have left the 4.50 change log unchanged; it contains all the changes since
2994 4.43. The change log for 4.44 is below; many of its items are identical to
2995 those for 4.50. This seems to be the most sensible way to preserve the
2996 historical information.
3002 1. Minor wording change to the doc/README.SIEVE file.
3004 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3005 computation of the current number of files was incorrect.
3007 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3008 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3011 4. Give more explanation in the error message when the command for a transport
3012 filter fails to execute.
3014 5. There are several places where Exim runs a non-Exim command in a
3015 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3016 was being done only for the command run by the queryprogram router. It is
3017 now done for all such subprocesses. The other cases are: ${run, transport
3018 filters, and the commands run by the lmtp and pipe transports.
3020 6. Added CONFIGURE_GROUP build-time option.
3022 7. Some older OS have a limit of 256 on the maximum number of file
3023 descriptors. Exim was using setrlimit() to set 1000 as a large value
3024 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3025 systems. I've change it so that if it can't get 1000, it tries for 256.
3027 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3028 was an oversight, and furthermore, ever since the addition of extra
3029 controls (e.g. 4.43/32), the checks on when to allow different forms of
3030 "control" were broken. There should now be diagnostics for all cases when a
3031 control that does not make sense is encountered.
3033 9. Added the /retain_sender option to "control=submission".
3035 10. $recipients is now available in the predata ACL (oversight).
3037 11. Tidy the search cache before the fork to do a delivery from a message
3038 received from the command line. Otherwise the child will trigger a lookup
3039 failure and thereby defer the delivery if it tries to use (for example) a
3040 cached ldap connection that the parent has called unbind on.
3042 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3043 of $address_data from the recipient verification was clobbered by the
3044 sender verification.
3046 13. The value of address_data from a sender verification is now available in
3047 $sender_address_data in subsequent conditions in the ACL statement.
3049 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3051 15. Added a new option "connect=<time>" to callout options, to set a different
3054 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3055 was its contents. (It was OK if the option was not defined at all.)
3057 17. A "Completed" log line is now written for messages that are removed from
3058 the spool by the -Mrm option.
3060 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3061 information about exactly what failed.
3063 19. Added -dd to debug only the daemon process.
3065 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3066 handles timeouts, both on the server side and network timeouts. Renamed the
3067 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3069 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3070 It is now set to "smtps".
3072 22. $host_address is now set to the target address during the checking of
3073 ignore_target_hosts.
3075 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3076 being passed; this would have caused $sender_host_name to have been used if
3077 matching the list had actually called for a host name (not very likely,
3078 since this list is usually IP addresses). A host name is now passed as
3081 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3082 code that specifies a non-restarting handler (typically sigaction() in
3083 modern systems) in an attempt to fix a rare and obscure crash bug.
3085 25. Narrowed the window for a race in the daemon that could cause it to ignore
3086 SIGCHLD signals. This is not a major problem, because they are used only to
3087 wake it up if nothing else does.
3089 26. A malformed maildirsize file could cause Exim to calculate negative values
3090 for the mailbox size or file count. Odd effects could occur as a result.
3091 The maildirsize information is now recalculated if the size or filecount
3094 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3095 support for a long time. Removed HAVE_SYS_VFS_H.
3097 28. Installed the latest version of exipick from John Jetmore.
3099 29. In an address list, if the pattern was not a regular expression, an empty
3100 subject address (from a bounce message) matched only if the pattern was an
3101 empty string. Non-empty patterns were not even tested. This was the wrong
3102 because it is perfectly reasonable to use an empty address as part of a
3103 database query. An empty address is now tested by patterns that are
3104 lookups. However, all the other forms of pattern expect the subject to
3105 contain a local part and a domain, and therefore, for them, an empty
3106 address still always fails if the pattern is not itself empty.
3108 30. Exim went into a mad DNS loop when attempting to do a callout where the
3109 host was specified on an smtp transport, and looking it up yielded more
3110 than one IP address.
3112 31. Re-factored the code for checking spool and log partition space into a
3113 function that finds that data and another that does the check. The former
3114 is then used to implement four new variables: $spool_space, $log_space,
3115 $spool_inodes, and $log_inodes.
3117 32. The RFC2047 encoding function was originally intended for short strings
3118 such as real names; it was not keeping to the 75-character limit for
3119 encoded words that the RFC imposes. It now respects the limit, and
3120 generates multiple encoded words if necessary. To be on the safe side, I
3121 have increased the buffer size for the ${rfc2047: expansion operator from
3124 33. It is now permitted to omit both strings after an "if" condition; if the
3125 condition is true, the result is "true". As before, when the second string
3126 is omitted, a false condition yields an empty string. This makes it less
3127 cumbersome to write custom ACL and router conditions.
3129 34. Failure to deliver a bounce message always caused it to be frozen, even if
3130 there was an errors_to setting on the router. The errors_to setting is now
3133 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3134 canonical form (fully expanded) before being placed in
3135 $sender_host_address.
3137 36. The table in the code that translates DNS record types into text (T_A to
3138 "A" for instance) was missing entries for NS and CNAME. It is just possible
3139 that this could have caused confusion if both these types were looked up
3140 for the same domain, because the text type is used as part of Exim's
3141 per-process caching. But the chance of anyone hitting this buglet seems
3144 37. The dnsdb lookup has been extended in a number of ways.
3146 (1) There is a new type, "zns", which walks up the domain tree until it
3147 finds some nameserver records. It should be used with care.
3149 (2) There is a new type, "mxh", which is like "mx" except that it returns
3150 just the host names, not the priorities.
3152 (3) It is now possible to give a list of domains (or IP addresses) to be
3153 looked up. The behaviour when one of the lookups defers can be
3154 controlled by a keyword.
3156 (4) It is now possible to specify the separator character for use when
3157 multiple records are returned.
3159 38. The dnslists ACL condition has been extended: it is now possible to supply
3160 a list of IP addresses and/or domains to be looked up in a particular DNS
3163 39. Added log_selector=+queue_time_overall.
3165 40. When running the queue in the test harness, wait just a tad after forking a
3166 delivery process, to get repeatability of debugging output.
3168 41. Include certificate and key file names in error message when GnuTLS fails
3169 to set them up, because the GnuTLS error message doesn't include the name
3170 of the failing file when there is a problem reading it.
3172 42. Allow both -bf and -bF in the same test run.
3174 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3176 44. The "Exiscan patch" is now merged into the mainline Exim source.
3178 45. Sometimes the final signoff response after QUIT could fail to get
3179 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3180 < 0 before doing a fflush(). This bug looks as though it goes back to the
3181 introduction of TLS in release 3.20, but "sometimes" must have been rare
3182 because the tests only now provoked it.
3184 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3185 (this can affect the format of dates).
3187 47. exim_tidydb, when checking for the continued existence of a message for
3188 which it has found a message-specific retry record, was not finding
3189 messages that were in split spool directories. Consequently, it was
3190 deleting retry records that should have stayed in existence.
3192 48. Steve fixed some bugs in eximstats.
3194 49. The SPA authentication driver was not abandoning authentication and moving
3195 on to the next authenticator when an expansion was forced to fail,
3196 contradicting the general specification for all authenticators. Instead it
3197 was generating a temporary error. It now behaves as specified.
3199 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3200 (the order specifies the preference for clients). The order is now AES256,
3201 AES128, 3DES, ARCFOUR128.
3203 51. Small patch to Sieve code - explicitly set From: when generating an
3206 52. Exim crashed if a remote delivery caused a very long error message to be
3207 recorded - for instance if somebody sent an entire SpamAssassin report back
3208 as a large number of 550 error lines. This bug was coincidentally fixed by
3209 increasing the size of one of Exim's internal buffers (big_buffer) that
3210 happened as part of the Exiscan merge. However, to be on the safe side, I
3211 have made the code more robust (and fixed the comments that describe what
3214 53. Now that there can be additional text after "Completed" in log lines (if
3215 the queue_time_overall log selector is set), a one-byte patch to exigrep
3216 was needed to allow it to recognize "Completed" as not the last thing in
3219 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3220 patch that reportedly fixes this has been added. I am not expert enough to
3221 create a test for it. This is what the patch creator wrote:
3223 "I found a little strange behaviour of ldap code when working with
3224 Windows 2003 AD Domain, where users was placed in more than one
3225 Organization Units. When I tried to give exim partial DN, the exit code
3226 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3227 But simultaneously result of request was absolutely normal ldap result,
3228 so I produce this patch..."
3230 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3231 REFERENCE, so I have modified the code to exclude the patch when that macro
3234 55. Some experimental protocols are using DNS PTR records for new purposes. The
3235 keys for these records are domain names, not reversed IP addresses. The
3236 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3237 leaves it alone. Component reversal etc. now happens only for IP addresses.
3240 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3242 57. Double the size of the debug message buffer (to 2048) so that more of very
3243 long debug lines gets shown.
3245 58. The exicyclog utility now does better if the number of log files to keep
3246 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3248 59. Two changes related to the smtp_active_hostname option:
3250 (1) $smtp_active_hostname is now available as a variable.
3251 (2) The default for smtp_banner uses $smtp_active_hostname instead
3252 of $primary_hostname.
3254 60. The host_aton() function is supposed to be passed a string that is known
3255 to be a valid IP address. However, in the case of IPv6 addresses, it was
3256 not checking this. This is a hostage to fortune. Exim now panics and dies
3257 if the condition is not met. A case was found where this could be provoked
3258 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3259 components; fortuitously, this particular loophole had already been fixed
3260 by change 4.50/55 above.
3262 If there are any other similar loopholes, the new check in host_aton()
3263 itself should stop them being exploited. The report I received stated that
3264 data on the command line could provoke the exploit when Exim was running as
3265 exim, but did not say which command line option was involved. All I could
3266 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3267 running as the user.
3270 61. There was a buffer overflow vulnerability in the SPA authentication code
3271 (which came originally from the Samba project). I have added a test to the
3272 spa_base64_to_bits() function which I hope fixes it.
3275 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3276 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3278 63. The daemon start-up calls getloadavg() while still root for those OS that
3279 need the first call to be done as root, but it missed one case: when
3280 deliver_queue_load_max is set with deliver_drop_privilege. This is
3281 necessary for the benefit of the queue runner, because there is no re-exec
3282 when deliver_drop_privilege is set.
3284 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3285 This has been fixed.
3287 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3288 was in use, was not putting the data itself into the right store pool;
3289 consequently, it could be overwritten for a subsequent message in the same
3290 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3293 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3295 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3296 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3297 the function were treating the return as a boolean value, which happened to
3298 work because 0=false and not-0=true, but is not correct code.
3300 68. The host_aton() function was not handling scoped IPv6 addresses (those
3301 with, for example, "%eth0" on the end) correctly.
3303 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3304 negated items (that is, ~something) in unsigned ints. Some compilers
3305 apparently mutter when there is no cast.
3307 70. If an address verification called from an ACL failed, and did not produce a
3308 user-specific message (i.e. there was only a "system" message), nothing was
3309 put in $acl_verify_message. In this situation, it now puts the system
3312 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3313 session; change 4.31/43 added the unwanted input to the log line - except
3314 that it did not do this in the start of session case. It now does.
3316 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3317 This is wrong and can cause the other end to generate a synchronization
3318 error if it is another Exim or anything else that does the synchronization
3319 check. A QUIT command is no longer sent after a timeout.
3321 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3322 during host lookups.
3324 74. The defer_ok option of callout verification was not working if it was used
3325 when verifying addresses in header lines, that is, for this case:
3327 verify = header_sender/callout=defer_ok
3329 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3330 those file descriptors could be used for SMTP connections. If anything
3331 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3332 could be sent to the SMTP client, causing chaos. The daemon now opens
3333 stdin, stdout, and stderr to /dev/null when it puts itself into the
3336 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3337 log by default. The user can override this with suitable Perl magic.
3339 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3340 the log message when discard triggers, was not working for the DATA ACL or
3341 for the non-SMTP ACL.
3343 78. Error message wording change in sieve.c.
3345 79. If smtp_accept_max_per_host was set, the number of connections could be
3346 restricted to fewer than expected, because the daemon was trying to set up
3347 a new connection before checking whether the processes handling previous
3348 connections had finished. The check for completed processes is now done
3349 earlier. On busy systems, this bug wouldn't be noticed because something
3350 else would have woken the daemon, and it would have reaped the completed
3353 80. If a message was submitted locally by a user whose login name contained one
3354 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3355 It caused a spool format error. I have fixed the spool reading code. A
3356 related problem was that the "from" clause in the Received: line became
3357 illegal because of the space(s). It is now covered by ${quote_local_part.
3359 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3362 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3364 83. Patch from Sieve maintainer to fix unterminated string problem in
3365 "vacation" handling.
3367 84. Some minor changes to the Linux configuration files to help with other
3368 OS variants using glibc.
3370 85. One more patch for Sieve to update vacation handling to latest spec.
3373 ----------------------------------------------------
3374 See the note above about the 4.44 and 4.50 releases.
3375 ----------------------------------------------------
3381 1. Change 4.43/35 introduced a bug that caused file counts to be
3382 incorrectly computed when quota_filecount was set in an appendfile
3385 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3386 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3389 3. Give more explanation in the error message when the command for a transport
3390 filter fails to execute.
3392 4. There are several places where Exim runs a non-Exim command in a
3393 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3394 was being done only for the command run by the queryprogram router. It is
3395 now done for all such subprocesses. The other cases are: ${run, transport
3396 filters, and the commands run by the lmtp and pipe transports.
3398 5. Some older OS have a limit of 256 on the maximum number of file
3399 descriptors. Exim was using setrlimit() to set 1000 as a large value
3400 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3401 systems. I've change it so that if it can't get 1000, it tries for 256.
3403 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3404 was an oversight, and furthermore, ever since the addition of extra
3405 controls (e.g. 4.43/32), the checks on when to allow different forms of
3406 "control" were broken. There should now be diagnostics for all cases when a
3407 control that does not make sense is encountered.
3409 7. $recipients is now available in the predata ACL (oversight).
3411 8. Tidy the search cache before the fork to do a delivery from a message
3412 received from the command line. Otherwise the child will trigger a lookup
3413 failure and thereby defer the delivery if it tries to use (for example) a
3414 cached ldap connection that the parent has called unbind on.
3416 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3417 of $address_data from the recipient verification was clobbered by the
3418 sender verification.
3420 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3421 was its contents. (It was OK if the option was not defined at all.)
3423 11. A "Completed" log line is now written for messages that are removed from
3424 the spool by the -Mrm option.
3426 12. $host_address is now set to the target address during the checking of
3427 ignore_target_hosts.
3429 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3430 being passed; this would have caused $sender_host_name to have been used if
3431 matching the list had actually called for a host name (not very likely,
3432 since this list is usually IP addresses). A host name is now passed as
3435 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3436 code that specifies a non-restarting handler (typically sigaction() in
3437 modern systems) in an attempt to fix a rare and obscure crash bug.
3439 15. Narrowed the window for a race in the daemon that could cause it to ignore
3440 SIGCHLD signals. This is not a major problem, because they are used only to
3441 wake it up if nothing else does.
3443 16. A malformed maildirsize file could cause Exim to calculate negative values
3444 for the mailbox size or file count. Odd effects could occur as a result.
3445 The maildirsize information is now recalculated if the size or filecount
3448 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3449 support for a long time. Removed HAVE_SYS_VFS_H.
3451 18. Updated exipick to current release from John Jetmore.
3453 19. Allow an empty sender to be matched against a lookup in an address list.
3454 Previously the only cases considered were a regular expression, or an
3457 20. Exim went into a mad DNS lookup loop when doing a callout where the
3458 host was specified on the transport, if the DNS lookup yielded more than
3461 21. The RFC2047 encoding function was originally intended for short strings
3462 such as real names; it was not keeping to the 75-character limit for
3463 encoded words that the RFC imposes. It now respects the limit, and
3464 generates multiple encoded words if necessary. To be on the safe side, I
3465 have increased the buffer size for the ${rfc2047: expansion operator from
3468 22. Failure to deliver a bounce message always caused it to be frozen, even if
3469 there was an errors_to setting on the router. The errors_to setting is now
3472 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3473 canonical form (fully expanded) before being placed in
3474 $sender_host_address.
3476 24. Updated eximstats to version 1.33
3478 25. Include certificate and key file names in error message when GnuTLS fails
3479 to set them up, because the GnuTLS error message doesn't include the name
3480 of the failing file when there is a problem reading it.
3482 26. Expand error message when OpenSSL has problems setting up cert/key files.
3485 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3486 (this can affect the format of dates).
3488 28. exim_tidydb, when checking for the continued existence of a message for
3489 which it has found a message-specific retry record, was not finding
3490 messages that were in split spool directories. Consequently, it was
3491 deleting retry records that should have stayed in existence.
3493 29. eximstats updated to version 1.35
3494 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3495 1.35 - bugfix such that pie charts by volume are generated correctly
3497 30. The SPA authentication driver was not abandoning authentication and moving
3498 on to the next authenticator when an expansion was forced to fail,
3499 contradicting the general specification for all authenticators. Instead it
3500 was generating a temporary error. It now behaves as specified.
3502 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3503 (the order specifies the preference for clients). The order is now AES256,
3504 AES128, 3DES, ARCFOUR128.
3506 31. Small patch to Sieve code - explicitly set From: when generating an
3509 32. Exim crashed if a remote delivery caused a very long error message to be
3510 recorded - for instance if somebody sent an entire SpamAssassin report back
3511 as a large number of 550 error lines. This bug was coincidentally fixed by
3512 increasing the size of one of Exim's internal buffers (big_buffer) that
3513 happened as part of the Exiscan merge. However, to be on the safe side, I
3514 have made the code more robust (and fixed the comments that describe what
3517 33. Some experimental protocols are using DNS PTR records for new purposes. The
3518 keys for these records are domain names, not reversed IP addresses. The
3519 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3520 leaves it alone. Component reversal etc. now happens only for IP addresses.
3523 34. The host_aton() function is supposed to be passed a string that is known
3524 to be a valid IP address. However, in the case of IPv6 addresses, it was
3525 not checking this. This is a hostage to fortune. Exim now panics and dies
3526 if the condition is not met. A case was found where this could be provoked
3527 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3528 components; fortuitously, this particular loophole had already been fixed
3529 by change 4.50/55 or 4.44/33 above.
3531 If there are any other similar loopholes, the new check in host_aton()
3532 itself should stop them being exploited. The report I received stated that
3533 data on the command line could provoke the exploit when Exim was running as
3534 exim, but did not say which command line option was involved. All I could
3535 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3536 running as the user.
3539 35. There was a buffer overflow vulnerability in the SPA authentication code
3540 (which came originally from the Samba project). I have added a test to the
3541 spa_base64_to_bits() function which I hope fixes it.
3544 36. The daemon start-up calls getloadavg() while still root for those OS that
3545 need the first call to be done as root, but it missed one case: when
3546 deliver_queue_load_max is set with deliver_drop_privilege. This is
3547 necessary for the benefit of the queue runner, because there is no re-exec
3548 when deliver_drop_privilege is set.
3550 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3551 was in use, was not putting the data itself into the right store pool;
3552 consequently, it could be overwritten for a subsequent message in the same
3553 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3556 38. Sometimes the final signoff response after QUIT could fail to get
3557 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3558 < 0 before doing a fflush(). This bug looks as though it goes back to the
3559 introduction of TLS in release 3.20, but "sometimes" must have been rare
3560 because the tests only now provoked it.
3566 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3567 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3568 Now it returns an integer. A number of calls were still expecting a T/F
3569 return. Fortuitously, in all cases, the tests worked in OK situations,
3570 which is the norm. However, things would have gone wrong on any write
3571 failures on the smtp file descriptor. This function is used when sending
3572 messages over SMTP and also when doing verify callouts.
3574 2. When Exim is called to do synchronous delivery of a locally submitted
3575 message (the -odf or -odi options), it no longer closes stderr before doing
3578 3. Implemented the mua_wrapper option.
3580 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3582 5. Implemented the functions header_remove(), header_testname(),
3583 header_add_at_position(), and receive_remove_recipient(), and exported them
3586 6. If an ACL "warn" statement specified the addition of headers, Exim already
3587 inserted X-ACL-Warn: at the start if there was no header name. However, it
3588 was not making this test for the second and subsequent header lines if
3589 there were newlines in the string. This meant that an invalid header could
3590 be inserted if Exim was badly configured.
3592 7. Allow an ACL "warn" statement to add header lines at the start or after all
3593 the Received: headers, as well as at the end.
3595 8. Added the rcpt_4xx retry error code.
3597 9. Added postmaster_mailfrom=xxx to callout verification option.
3599 10. Added mailfrom=xxxx to the callout verification option, for verify=
3602 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3603 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3605 12. Inserted some casts to stop certain compilers warning when using pointer
3606 differences as field lengths or precisions in printf-type calls (mostly
3607 affecting debugging statements).
3609 13. Added optional readline() support for -be (dynamically loaded).
3611 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3612 same clock tick as a message's arrival, so that its received time was the
3613 same as the "first fail" time on the retry record, and that message
3614 remained on the queue past the ultimate address timeout, every queue runner
3615 would try a delivery (because it was past the ultimate address timeout) but
3616 after another failure, the ultimate address timeout, which should have then
3617 bounced the address, did not kick in. This was a "< instead of <=" error;
3618 in most cases the first failure would have been in the next clock tick
3619 after the received time, and all would be well.
3621 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3622 being recognized when the domain list was tested by the match_domain
3623 condition in an expansion string.
3625 16. Added the ${str2b64: operator.
3627 17. Exim was always calling setrlimit() to set a large limit for the number of
3628 processes, without checking whether the existing limit was already
3629 adequate. (It did check for the limit on file descriptors.) Furthermore,
3630 errors from getrlimit() and setrlimit() were being ignored. Now they are
3631 logged to the main and panic logs, but Exim does carry on, to try to do its
3632 job under whatever limits there are.
3634 18. Imported PCRE 5.0.
3636 19. Trivial typo in log message " temporarily refused connection" (the leading
3639 20. If the log selector return_path_on_delivery was set and an address was
3640 redirected to /dev/null, the delivery process crashed because it assumed
3641 that a return path would always be set for a "successful" delivery. In this
3642 case, the whole delivery is bypassed as an optimization, and therefore no
3645 21. Internal re-arrangement: the function for sending a challenge and reading
3646 a response while authentication was assuming a zero-terminated challenge
3647 string. It's now changed to take a pointer and a length, to allow for
3648 binary data in such strings.
3650 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3652 23. Exim was not respecting finduser_retries when seeking the login of the
3653 uid under which it was called; it was always trying 10 times. (The default
3654 setting of finduser_retries is zero.) Also, it was sleeping after the final
3655 failure, which is pointless.
3657 24. Implemented tls_on_connect_ports.
3659 25. Implemented acl_smtp_predata.
3661 26. If the domain in control=submission is set empty, Exim assumes that the
3662 authenticated id is a complete email address when it generates From: or
3663 Sender: header lines.
3665 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3666 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3667 chown and chgrp in /bin and hostname in /usr/bin.
3669 28. Exim was keeping the "process log" file open after each use, just as it
3670 does for the main log. This opens the possibility of it remaining open for
3671 long periods when the USR1 signal hits a daemon. Occasional processlog
3672 errors were reported, that could have been caused by this. Anyway, it seems
3673 much more sensible not to leave this file open at all, so that is what now
3676 29. The long-running daemon process does not normally write to the log once it
3677 has entered its main loop, and it closes the log before doing so. This is
3678 so that log files can straightforwardly be renamed and moved. However,
3679 there are a couple of unusual error situations where the daemon does write
3680 log entries, and I had neglected to close the log afterwards.
3682 30. The text of an SMTP error response that was received during a remote
3683 delivery was being truncated at 512 bytes. This is too short for some of
3684 the long messages that one sometimes sees. I've increased the limit to
3687 31. It is now possible to make retry rules that apply only when a message has a
3688 specific sender, in particular, an empty sender.
3690 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3691 it possible to be selective about when SMTP synchronization is enforced.
3693 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3695 32. Implemented hosts_connection_nolog.
3697 33. Added an ACL for QUIT.
3699 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3702 35. Added mailbox_size and mailbox_filecount to appendfile.
3704 36. Added control = no_multiline_responses to ACLs.
3706 37. There was a bug in the logic of the code that waits for the clock to tick
3707 in the case where the clock went backwards by a substantial amount such
3708 that the microsecond fraction of "now" was more than the microsecond
3709 fraction of "then" (but the whole seconds number was less).
3711 38. Added support for the libradius Radius client library this is found on
3712 FreeBSD (previously only the radiusclient library was supported).
3718 1. When certain lookups returned multiple values in the form name=value, the
3719 quoting of the values was not always being done properly. Specifically:
3720 (a) If the value started with a double quote, but contained no whitespace,
3722 (b) If the value contained whitespace other than a space character (i.e.
3723 tabs or newlines or carriage returns) it was not quoted.
3724 This fix has been applied to the mysql and pgsql lookups by writing a
3725 separate quoting function and calling it from the lookup code. The fix
3726 should probably also be applied to nisplus, ibase and oracle lookups, but
3727 since I cannot test any of those, I have not disturbed their existing code.
3729 2. A hit in the callout cache for a specific address caused a log line with no
3730 reason for rejecting RCPT. Now it says "Previous (cached) callout
3731 verification failure".
3733 3. There was an off-by-one bug in the queryprogram router. An over-long
3734 return line was truncated at 256 instead of 255 characters, thereby
3735 overflowing its buffer with the terminating zero. As well as fixing this, I
3736 have increased the buffer size to 1024 (and made a note to document this).
3738 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3739 when Exim is waiting for an SMTP response from a remote server, Exim
3740 restarts its select() call on the socket, thereby resetting its timeout.
3741 This is not a problem when such interrupts are rare. Somebody set up a cron
3742 job to run exiwhat every 2 minutes, which is less than the normal select()
3743 timeout (5 or 10 minutes). This meant that the select() timeout never
3744 kicked in because it was always reset. I have fixed this by comparing the
3745 time when an interrupt arrives with the time at the start of the first call
3746 to select(). If more time than the timeout has elapsed, the interrupt is
3747 treated as a timeout.
3749 5. Some internal re-factoring in preparation for the addition of Sieve
3750 extensions (by MH). In particular, the "personal" test is moved to a
3751 separate function, and given an option for scanning Cc: and Bcc: (which is
3752 not set for Exim filters).
3754 6. When Exim created an email address using the login of the caller as the
3755 local part (e.g. when creating a From: or Sender: header line), it was not
3756 quoting the local part when it contained special characters such as @.
3758 7. Installed new OpenBSD configuration files.
3760 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3761 try to make them clearer.
3763 9. Callout options, other than the timeout value, were being ignored when
3764 verifying sender addresses in header lines. For example, when using
3766 verify = header_sender/callout=no_cache
3768 the cache was (incorrectly) being used.
3770 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3771 only the Cygwin environment.
3773 11. When return_path_on_delivery was set as a log selector, if different remote
3774 addresses in the same message used different return paths and parallel
3775 remote delivery occurred, the wrong values would sometimes be logged.
3776 (Whenever a remote delivery process finished, the return path value from
3777 the most recently started remote delivery process was logged.)
3779 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3780 lines when AUTH and/or TLS are in use. This is the "received protocol"
3781 field. Exim used to use "asmtp" for authenticated SMTP, without any
3782 indication (in the protocol name) for TLS use. Now it follows the RFC and
3783 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3784 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3785 names appear in log lines as well as in Received: header lines.
3787 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3788 extensions, and comparison tests, and to fix some bugs.
3790 14. Changes to the "personal" filter test:
3792 (1) The test was buggy in that it was just doing the equivalent of
3793 "contains" tests on header lines. For example, if a user's address was
3794 anne@some.where, the "personal" test would incorrectly be true for
3796 To: susanne@some.where
3798 This test is now done by extracting each address from the header in turn,
3799 and checking the entire address. Other tests that are part of "personal"
3800 are now done using regular expressions (for example, to check local parts
3801 of addresses in From: header lines).
3803 (2) The list of non-personal local parts in From: addresses has been
3804 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3805 taken from the Sieve specification recommendations.
3807 (3) If the message contains any header line starting with "List-" it is
3808 treated as non-personal.
3810 (4) The test for "circular" in the Subject: header line has been removed
3811 because it now seems ill-conceived.
3813 15. Minor typos in src/EDITME comments corrected.
3815 16. Installed latest exipick from John Jetmore.
3817 17. If headers_add on a router specified a text string that was too long for
3818 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3819 of string_sprintf() is now avoided.
3821 18. $message_body_size was not set (it was always zero) when running the DATA
3822 ACL and the local_scan() function.
3824 19. For the "mail" command in an Exim filter, no default was being set for
3825 the once_repeat time, causing a random time value to be used if "once" was
3826 specified. (If the value happened to be <= 0, no repeat happened.) The
3827 default is now 0s, meaning "never repeat". The "vacation" command was OK
3828 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3829 (I found it when inspecting the code).
3831 20. There is now an overall timeout for performing a callout verification. It
3832 defaults to 4 times the callout timeout, which applies to individual SMTP
3833 commands during the callout. The overall timeout applies when there is more
3834 than one host that can be tried. The timeout is checked before trying the
3835 next host. This prevents very long delays if there are a large number of
3836 hosts and all are timing out (e.g. when the network connections are timing
3837 out). The value of the overall timeout can be changed by specifying an
3838 additional sub-option for "callout", called "maxwait". For example:
3840 verify = sender/callout=5s,maxwait=20s
3842 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3843 to the end before writing, but this should make it even safer).
3845 22. Exim was forgetting that it had advertised PIPELINING for the second and
3846 subsequent messages on an SMTP connection. It was also not resetting its
3847 memory on STARTTLS and an internal HELO.
3849 23. When Exim logs an SMTP synchronization error within a session, it now
3850 records whether PIPELINING has been advertised or not.
3852 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3853 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3854 rather than long int.
3856 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3858 26. Added the never_mail option to autoreply.
3864 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3865 crash if the getsockname() call failed; this can happen if a connection is
3866 closed very soon after it is established. The problem was simply in the
3867 order in which certain operations were done, causing Exim to try to write
3868 to the SMTP stream before it had set up the file descriptor. The bug has
3869 been fixed by making things happen in the correct order.
3875 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3876 before the connection was closed, thus losing the rejection response.
3878 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3879 some early Solaris releases, but causes trouble in current releases where
3880 socklen_t is defined.
3882 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3885 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3888 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3889 that releasing the top of it at the end releases what was used for sub-
3890 expansions (unless the block got too big). However, discard this block if
3891 the first thing is a variable or header, so that we can use its block when
3892 it is dynamic (useful for very large $message_headers, for example).
3894 6. Lookups now cache *every* query, not just the most recent. A new, separate
3895 store pool is used for this. It can be recovered when all lookup caches are
3896 flushed. Lookups now release memory at the end of their result strings.
3897 This has involved some general refactoring of the lookup sources.
3899 7. Some code has been added to the store_xxx() functions to reduce the amount
3900 of flapping under certain conditions.
3902 8. log_incoming_interface used to affect only the <= reception log lines. Now
3903 it causes the local interface and port to be added to several more SMTP log
3904 lines, for example "SMTP connection from", and rejection lines.
3906 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3908 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3910 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3911 could be overwritten at the end of the current message (or the start of a
3912 new message if it was set in a HELO ACL). The value is now preserved for
3913 the duration of the SMTP connection.
3915 12. If a transport had a headers_rewrite setting, and a matching header line
3916 contained an unqualified address, that address was qualified, even if it
3917 did not match any rewriting rules. The underlying bug was that the values
3918 of the flags that permit the existence of unqualified sender and recipient
3919 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3920 non-local messages, and by -bnq for local messages) were not being
3921 preserved with the message after it was received.
3923 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3924 "next input=" as part of the text comprising the host identity instead of
3925 the correct text. The code was using the same buffer for two different
3926 strings. However, depending on which order the printing function evaluated
3927 its arguments, the bug did not always show up. Under Linux, for example, my
3928 test suite worked just fine.
3930 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3931 doesn't work with some older versions of Perl. It has been changed to "my",
3932 which in any case is probably the better facility to use.
3934 15. A really picky compiler found some instances of statements for creating
3935 error messages that either had too many or two few arguments for the format
3938 16. The size of the buffer for calls to the DNS resolver has been increased
3939 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3940 for addresses that have a lot of PTR records. This alleviates a problem; it
3941 does not fully solve it.
3943 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3944 buffer now truncates the list and logs the incident, which is the same
3945 action as happens when Exim is looking up a host name and its aliases.
3946 Previously in this situation something unpredictable would happen;
3947 sometimes it was "internal error: store_reset failed".
3949 18. If a server dropped the connection unexpectedly when an Exim client was
3950 using GnuTLS and trying to read a response, the client delivery process
3951 crashed while trying to generate an error log message.
3953 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3954 string, for example:
3956 warn message = H1: something\nH2: something
3958 the text was added as a single header line from Exim's point of view
3959 though it ended up OK in the delivered message. However, searching for the
3960 second and subsequent header lines using $h_h2: did not work. This has been
3961 fixed. Similarly, if a system filter added multiple headers in this way,
3962 the routers could not see them.
3964 20. Expanded the error message when iplsearch is called with an invalid key to
3965 suggest using net-iplsearch in a host list.
3967 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3968 ACLs are no longer actually imposed (and a message to that effect is
3971 22. If a "gecos" field in a passwd entry contained escaped characters, in
3972 particular, if it contained a \" sequence, Exim got it wrong when building
3973 a From: or a Sender: header from that name. A second bug also caused
3974 incorrect handling when an unquoted " was present following a character
3975 that needed quoting.
3977 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3978 was not being matched caselessly.
3980 24. Arranged for all hyphens in the exim.8 source to be escaped with
3983 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3984 properly. Recipient callout cache records were still being keyed to include
3985 the sender, even when use_sender was set false. This led to far more
3986 callouts that were necessary. The sender is no longer included in the key
3987 when use_sender is false.
3989 26. Added "control = submission" modifier to ACLs.
3991 27. Added the ${base62d: operator to decode base 62 numbers.
3993 28. dnsdb lookups can now access SRV records.
3995 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3996 the configuration file.
3998 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3999 (-v) mode. This makes the output for a verbose queue run more intelligible.
4001 31. Added a use_postmaster feature to recipient callouts.
4003 32. Added the $body_zerocount variable, containing the number of binary zero
4004 bytes in the message body.
4006 33. The time of last modification of the "new" subdirectory is now used as the
4007 "mailbox time last read" when there is a quota error for a maildir
4010 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4012 35. Added +ignore_unknown as a special item in host lists.
4014 36. Code for decoding IPv6 addresses in host lists is now included, even if
4015 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4016 address was recognized as an IP address, but was then not correctly decoded
4017 into binary, causing unexpected and incorrect effects when compared with
4024 1. Very minor rewording of debugging text in manualroute to say "list of
4025 hosts" instead of "hostlist".
4027 2. If verify=header_syntax was set, and a header line with an unqualified
4028 address (no domain) and a large number of spaces between the end of the
4029 name and the colon was received, the reception process suffered a buffer
4030 overflow, and (when I tested it) crashed. This was caused by some obsolete
4031 code that should have been removed. The fix is to remove it!
4033 3. When running in the test harness, delay a bit after writing a bounce
4034 message to get a bit more predictability in the log output.
4036 4. Added a call to search_tidyup() just before forking a reception process. In
4037 theory, someone could use a lookup in the expansion of smtp_accept_max_
4038 per_host which, without the tidyup, could leave open a database connection.
4040 5. Added the variables $recipient_data and $sender_data which get set from a
4041 lookup success in an ACL "recipients" or "senders" condition, or a router
4042 "senders" option, similar to $domain_data and $local_part_data.
4044 6. Moved the writing of debug_print from before to after the "senders" test
4047 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4048 problems for message scanning, either using a data ACL, or using
4049 local_scan() because the Received: header was not generated till after they
4050 were called (in order to set the time as the time of reception completion).
4051 I have revised the way this works. The header is now generated after the
4052 body is received, but before the ACL or local_scan() are called. After they
4053 are run, the timestamp in the header is updated.
4059 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4060 before starting a queue runner without re-exec. This happened only when
4061 deliver_drop_privilege was set or when the Exim user was set to root. The
4062 effect of the bug was that timeouts during subsequent deliveries caused
4063 crashes instead of being properly handled. The handler is now left at its
4064 default (and expected) setting.
4066 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4067 message, again when deliver_drop_privilege is set or Exim is run as root.
4068 The bug described in (1) was not present in this case, but the tidying up
4069 of the other signals was missing. I have made the two cases consistent.
4071 3. The ignore_target_hosts setting on a manualroute router was being ignored
4072 for hosts that were looked up using the /MX notation.
4074 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4077 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4078 operated on the sender address. After changing the $sender_address to <>
4079 for the sender address verify, Exim was re-instated it as the original
4080 (before rewriting) address, but remembering that it had rewritten it, so it
4081 wasn't rewriting it again. This bug also had the effect of breaking the
4082 sender address verification caching when the sender address was rewritten.
4084 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4085 This has been changed so that if the ip literal address matches
4086 ignore_target_hosts, the router declines.
4088 7. Added expansion conditions match_domain, match_address, and match_local_
4089 part (NOT match_host).
4091 8. The placeholder for the Received: header didn't have a length field set.
4093 9. Added code to Exim itself and to exim_lock to test for a specific race
4094 condition that could lead to file corruption when using MBX delivery. The
4095 issue is with the lockfile that is created in /tmp. If this file is removed
4096 after a process has opened it but before that process has acquired a lock,
4097 there is the potential for a second process to recreate the file and also
4098 acquire a lock. This could lead to two Exim processes writing to the file
4099 at the same time. The added code performs the same test as UW imapd; it
4100 checks after acquiring the lock that its file descriptor still refers to
4101 the same named file.
4103 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4104 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4107 11. Added the smtp_active_hostname option. If used, this will typically be made
4108 to depend on the incoming interface address. Because $interface_address is
4109 not set up until the daemon has forked a reception process, error responses
4110 that can happen earlier (such as "too many connections") no longer contain
4113 12. If an expansion in a condition on a "warn" statement fails because a lookup
4114 defers, the "warn" statement is abandoned, and the next ACL statement is
4115 processed. Previously this caused the whole ACL to be aborted.
4117 13. Added the iplsearch lookup type.
4119 14. Added ident_timeout as a log selector.
4121 15. Added tls_certificate_verified as a log selector.
4123 16. Added a global option tls_require_ciphers (compare the smtp transport
4124 option of the same name). This controls incoming TLS connections.
4126 17. I finally figured out how to make tls_require_ciphers do a similar thing
4127 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4128 before starting the TLS session.
4130 18. Tabs are now shown as \t in -bP output.
4132 19. If the log selector return_path_on_delivery was set, Exim crashed when
4133 bouncing a message because it had too many Received: header lines.
4135 20. If two routers both had headers_remove settings, and the first one included
4136 a superfluous trailing colon, the final name in the first list and the
4137 first name in the second list were incorrectly joined into one item (with a
4138 colon in the middle).
4144 1. Added -C and -D options to the exinext utility, mainly to make it easier
4145 to include in the automated testing, but these could be helpful when
4146 multiple configurations are in use.
4148 2. The exinext utility was not formatting the output nicely when there was
4149 an alternate port involved in the retry record key, nor when there was a
4150 message id as well (for retries that were specific to a specific message
4151 and a specific host). It was also confused by IPv6 addresses, because of
4152 the additional colons they contain. I have fixed the IPv4 problem, and
4153 patched it up to do a reasonable job for IPv6.
4155 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4156 delivery, the log line now contains "pipelined" if PIPELINING was used.
4158 4. An SMTP transport process used to panic and die if the bind() call to set
4159 an explicit outgoing interface failed. This has been changed; it is now
4160 treated in the same way as a connect() failure.
4162 5. A reference to $sender_host_name in the part of a conditional expansion
4163 that was being skipped was still causing a DNS lookup. This no longer
4166 6. The def: expansion condition was not recognizing references to header lines
4167 that used bh_ and bheader_.
4169 7. Added the _cache feature to named lists.
4171 8. The code for checking quota_filecount in the appendfile transport was
4172 allowing one more file than it should have been.
4174 9. For compatibility with Sendmail, the command line option
4182 and sets the incoming protocol and host name (for trusted callers). The
4183 host name and its colon can be omitted when only the protocol is to be set.
4184 Note the Exim already has two private options, -pd and -ps, that refer to
4185 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4186 "s", but I don't think that's a major issue.
4188 10. A number of refactoring changes to the code, none of which should affect
4191 (a) The number of logging options was getting close to filling up the
4192 32-bit word that was used as a bit map. I have split them into two classes:
4193 those that are passed in the argument to log_write(), and those that are
4194 only ever tested independently outside of that function. These are now in
4195 separate 32-bit words, so there is plenty of room for expansion again.
4196 There is no change in the user interface or the logging behaviour.
4198 (b) When building, for example, log lines, the code previously used a
4199 macro that called string_cat() twice, in order to add two strings. This is
4200 not really sufficiently general. Furthermore, there was one instance where
4201 it was actually wrong because one of the argument was used twice, and in
4202 one call a function was used. (As it happened, calling the function twice
4203 did not affect the overall behaviour.) The macro has been replaced by a
4204 function that can join an arbitrary number of extra strings onto a growing
4207 (c) The code for expansion conditions now uses a table and a binary chop
4208 instead of a serial search (which was left over from when there were very
4209 few conditions). Also, it now recognizes conditions like "pam" even when
4210 the relevant support is not compiled in: a suitably worded error message is
4211 given if an attempt is made to use such a condition.
4213 11. Added ${time_interval:xxxxx}.
4215 12. A bug was causing one of the ddress fields not to be passed back correctly
4216 from remote delivery subprocesses. The field in question was not being
4217 subsequently used, so this caused to problems in practice.
4219 13. Added new log selectors queue_time and deliver_time.
4221 14. Might have fixed a bug in maildirsizefile handling that threw up
4222 "unexpected character" debug warnings, and recalculated the data
4223 unnecessarily. In any case, I expanded the warning message to give more
4226 15. Added the message "Restricted characters in address" to the statements in
4227 the default ACL that block characters like @ and % in local parts.
4229 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4230 Three changes have been made:
4232 (a) There was a serious bug; a negative response to MAIL caused the whole
4233 recipient domain to be cached as invalid, thereby blocking all messages
4234 to all local parts at the same domain, from all senders. This bug has
4235 been fixed. The domain is no longer cached after a negative response to
4236 MAIL if the sender used is not empty.
4238 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4241 (c) A new callout option, "use_sender" has been added for people who want
4242 the modified behaviour.
4248 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4251 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4252 indeed breaks things for older releases.
4254 3. Added additional logging to the case where there is a problem reading data
4255 from a filter that is running in a subprocess using a pipe, in order to
4256 try to track down a specific problem.
4258 4. Testing facility fudge: when running in the test harness and attempting
4259 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4260 getting "No route to host". Convert this to a timeout.
4262 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4265 6. Some OS don't have socklen_t but use size_t instead. This affects the
4266 fifth argument of getsockopt() amongst other things. This is now
4267 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4268 can be set for individual OS. I have set it for SunOS5, OSF1, and
4269 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4270 some earlier ones do not.
4272 7. Change 4.30/15 was not doing the test caselessly.
4274 8. The standard form for an IPv6 address literal was being rejected by address
4275 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4276 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4277 this, as well as the form without the "IPv6" on the front (but only when
4278 address literals are enabled, of course).
4280 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4282 10. Exim crashed if a message with an empty sender address specified by -f
4283 encountered a router with an errors_to setting. This could be provoked only
4284 by a command such as
4288 where an empty string was supplied; "<>" did not hit this bug.
4290 11. Installed PCRE release 4.5.
4292 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4293 remained set. It is now erased.
4295 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4296 times from message ids (which are base 36 rather than the normal 62).
4298 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4299 were being counted as actual protocol errors, and logged if the log
4300 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4301 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4302 and DATA following a set of rejected RCPTs do not count as protocol errors.
4303 In other words, Exim assumes they were pipelined, though this may not
4304 actually be the case. Of course, in all cases the client gets an
4305 appropriate error code.
4307 15. If a lookup fails in an ACL condition, a message about the failure may
4308 be available; it is used if testing the ACL cannot continue, because most
4309 such messages specify what the cause of the deferral is. However, some
4310 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4311 that caused an old message to be retained and used if a later statement
4312 caused a defer, replacing the real cause of the deferral.
4314 16. If an IP address had so many PTR records that the DNS lookup buffer
4315 was not large enough to hold them, Exim could crash while trying to process
4316 the truncated data. It now detects and logs this case.
4318 17. Further to 4.21/58, another change has been made: if (and only if) the
4319 first line of a message (the first header line) ends with CRLF, a bare LF
4320 in a subsequent header line has a space inserted after it, so as not to
4321 terminate the header.
4323 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4324 unnecessarily, used atoi() instead of strtol(), and didn't check the
4325 termination when getting file sizes from file names by regex.
4327 19. Completely re-implemented the support for maildirsize files, in the light
4328 of a number of problems with the previous contributed implementation
4329 (4.30/29). In particular:
4331 . If the quota is zero, the maildirsize file is maintained, but no quota is
4334 . If the maildir directory does not exist, it is created before any attempt
4335 to write a maildirsize file.
4337 . The quota value in the file is just a cache; if the quota is changed in
4338 the transport, the new value overrides.
4340 . A regular expression is available for excluding directories from the
4343 20. The autoreply transport checks the characters in options that define the
4344 message's headers; it allows continued headers, but it was checking with
4345 isspace() after an embedded newline instead of explicitly looking for a
4348 21. If all the "regular" hosts to which an address was routed had passed their
4349 expiry times, and had not reached their retry times, the address was
4350 bounced, even if fallback hosts were defined. Now Exim should go on to try
4353 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4354 equivalent code in the SMTP transport. Some hosts send humungous responses
4355 to HELO/EHLO, more than 1024 it seems.
4357 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4358 gives compiler warnings in some environments. I've now done it "properly",
4361 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4362 (because of problems with the built-in one) was declared to return uschar *
4363 instead of char *, causing compiler failure.
4365 25. Fixed a file descriptor leak when processing alias/forward files.
4367 26. Fixed a minor format string issue in dbfn.c.
4369 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4371 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4372 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4375 29. When a sender address is verified, it is cached, to save repeating the test
4376 when there is more than one recipient in a message. However, when the
4377 verification involves a callout, it is possible for different callout
4378 options to be set for different recipients. It is too complicated to keep
4379 track of this in the cache, so now Exim always runs a verification when a
4380 callout is required, relying on the callout cache for the optimization.
4381 The overhead is duplication of the address routing, but this should not be
4384 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4385 to be verified with callout=postmaster, and the main callout worked but the
4386 postmaster check failed, the verification correctly failed. However, if a
4387 subsequent RCPT command asked for sender verification *without* the
4388 postmaster check, incorrect caching caused this verification also to fail,
4391 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4392 it was not caching the DNS options (qualify_single, search_parents) that
4393 were used when the lookup failed. A subsequent lookup with different
4394 options therefore always gave the same answer, though there were cases
4395 where it should not have. (Example: a "domains = !$mx_any" option on a
4396 dnslookup router: the "domains" option is always processed without any
4397 widening, but the router might have qualify_single set.) Now Exim uses the
4398 cached value only when the same options are set.
4400 32. Added John Jetmore's "exipick" utility to the distribution.
4402 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4403 than a timeout (e.g. a certificate is required, and is not provided), an
4404 Exim server now closes the connection immediately. Previously it waited for
4405 the client to close - but if the client is SSL, it seems that they each
4406 wait for each other, leading to a delay before one of them times out.
4408 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4409 maintained 0.8.x compatibility because I don't think many are using it, and
4410 it is clearly obsolete.
4412 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4415 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4416 client certificate was expired. A simple patch fixes this, though I don't
4417 understand the full logic of why the verify callback is called multiple
4420 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4421 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4422 which causes problems with some clients (such as the Certicom SSL Plus
4423 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4424 disables the coutermeasure allowing Eudora to connect."
4426 38. Exim was not checking that a write() to a log file succeeded. This could
4427 lead to Bad Things if a log got too big, in particular if it hit a file
4428 size limit. Exim now panics and dies if it cannot write to a log file, just
4429 as it does if it cannot open a log file.
4431 39. Modified OS/Makefile-Linux so that it now contains
4433 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4435 The two -D definitions ensure that Exim is compiled with large file
4436 support, which makes it possible to handle log files that are bigger than
4439 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4440 instance) a domain was checked against a named list that involved a lookup,
4441 causing $domain_data to be set, then another domain was checked against the
4442 same list, then the first domain was re-checked, the value of $domain_data
4443 after the final check could be wrong. In particular, if the second check
4444 failed, it could be set empty. This bug probably also applied to
4447 41. The strip_trailing_dot option was not being applied to the address given
4448 with the -f command-line option.
4450 42. The code for reading a message's header from the spool was incrementing
4451 $received_count, but never initializing it. This meant that the value was
4452 incorrect (doubled) while delivering a message in the same process in which
4453 it was received. In the most common configuration of Exim, this never
4454 happens - a fresh exec is done - but it can happen when
4455 deliver_drop_privilege is set.
4457 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4458 it now includes up to 150 characters of the unexpected data in the log
4461 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4462 and building data strings. The size of both of these buffers was 10 000
4463 bytes - far larger than anybody would *ever* want, thought I. Needless to
4464 say, somebody hit the limit. I have increased the maximum line length to
4465 20 000 and the maximum data length of concatenated lines to 100 000. I have
4466 also fixed two bugs, because there was no checking on these buffers. Tsk,
4467 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4468 buffer is too small.
4470 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4471 lsearch lookups. Now it does.
4473 46. When parsing a route_list item in a manualroute router, a fixed-length
4474 buffer was used for the list of hosts. I made this 1024 bytes long,
4475 thinking that nobody would ever have a list of hosts that long. Wrong.
4476 Somebody had a whole pile of complicated expansion conditions, and the
4477 string was silently truncated, leading to an expansion error. It turns out
4478 that it is easier to change to an unlimited length (owing to other changes
4479 that have happened since this code was originally written) than to build
4480 structure for giving a limitation error. The length of the item that
4481 expands into the list of hosts is now unlimited.
4483 47. The lsearch lookup could not handle data where the length of text line was
4484 more than 4095 characters. Such lines were truncated, leading to shortened
4485 data being returned. It should now handle lines of any length.
4487 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4488 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4491 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4492 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4493 no longer happens in comment lines. A list of replacements is now placed
4494 at the head of all of the source files, except those whose only change is
4495 to replace PERL_COMMAND in the very first #! line.
4497 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4498 messages on the queue, with a bottom-up merge sort, using code contributed
4499 by Michael Haardt. This should make operations like -bp somewhat faster on
4500 large queues. It won't affect queue runners, except when queue_run_in_order
4503 51. Installed eximstats 1.31 in the distribution.
4505 52. Added support for SRV lookups to the dnslookup router.
4507 53. If an ACL referred to $message_body or $message_body_end, the value was not
4508 reset for any messages that followed in the same SMTP session.
4510 54. The store-handling optimization for building very long strings was not
4511 differentiating between the different store pools. I don't think this
4512 actually made any difference in practice, but I've tidied it.
4514 55. While running the routers to verify a sender address, $sender_address
4515 was still set to the sender address. This is wrong, because when routing to
4516 send a bounce to the sender, it would be empty. Therefore, I have changed
4517 it so that, while verifying a sender address, $sender_address is set to <>.
4518 (There is no change to what happens when verifying a recipient address.)
4520 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4521 target A or AAAA records (if not already returned) without resetting the
4522 qualify_single or search_parents options of the DNS resolver. These are
4523 inappropriate in this case because the targets of MX and SRV records must
4524 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4525 target that, when qualified, matched something in the local domain. These
4526 two options are now turned off when doing these lookups.
4528 57. It seems that at least some releases of Reiserfs (which does not have the
4529 concept of a fixed number of inodes) returns zero and not -1 for the
4530 number of available inodes. This interacted badly with check_spool_inodes,
4531 which assumed that -1 was the "no such thing" setting. What I have done is
4532 to check that the total number of inodes is greater than zero before doing
4533 the test of how many are available.
4535 58. When a "warn" ACL statement has a log_message modifier, the message is
4536 remembered, and not repeated. This is to avoid a lot of repetition when a
4537 message has many recipients that cause the same warning to be written.
4538 Howewer, Exim was preserving the list of already written lines for an
4539 entire SMTP session, which doesn't seem right. The memory is now reset if a
4540 new message is started.
4542 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4543 debugging output unless log_rewrite was also set.
4545 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4546 of (int)(handle) when we know that handle contains (void *)(-1).
4548 61. The Exim daemon panic-logs an error return when it closes the incoming
4549 connection. However "connection reset by peer" seems to be common, and
4550 isn't really an error worthy of noting specially, so that particular error
4553 62. When Exim is trying to find all the local interfaces, it used to panic and
4554 die if the ioctl to get the interface flags failed. However, it seems that
4555 on at least one OS (Solaris 9) it is possible to have an interface that is
4556 included in the list of interfaces, but for which you get a failure error
4557 for this call. This happens when the interface is not "plumbed" into a
4558 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4559 failure of the "get flags" call assumes that the interface is down.
4561 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4562 makes life easier for people who are doing arithmetic on fields extracted
4563 from dates, where you often get leading zeros that should not be
4564 interpreted as octal.
4566 64. Added qualify_domain to the redirect router, to override the global
4569 65. If a pathologically long header line contained very many addresses (the
4570 report of this problem mentioned 10 000) and each of them was rewritten,
4571 Exim could use up a very large amount of memory. (It kept on making new
4572 copies of the header line as it rewrote, and never released the old ones.)
4573 At the expense of a bit more processing, the header rewriting function has
4574 been changed so that it no longer eats memory in this way.
4576 66. The generation of the Received: header has been moved from the time that a
4577 message starts to be received, to the time that it finishes. The timestamp
4578 in the Received: header should now be very close to that of the <= log
4579 line. There are two side-effects of this change:
4581 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4582 logged header lines no longer include the local Received: line, because
4583 it has not yet been created. The same applies to a copy of the message
4584 that is returned to a non-SMTP sender when a message is rejected.
4586 (b) When a filter file is tested using -bf, no additional Received: header
4587 is added to the test message. After some thought, I decided that this
4590 This change does not affect the value of $received_for. It is still set
4591 after address rewriting, but before local_scan() is called.
4593 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4595 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4596 gave an unhelpful panic error message, and a defer error. I have managed to
4597 change this behaviour so that it now rejects any supplied certificate,
4598 which seems right, as the list of acceptable certificates is empty.
4600 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4601 gave an unhelpful defer error. I have not managed to make this reject any
4602 supplied certificates, but the error message it gives is "no certificate
4603 supplied", which is not helpful.
4605 70. exigrep's output now also includes lines that are not associated with any
4606 message, but which match the given pattern. Implemented by a patch from
4607 Martin Sluka, which also tidied up the Perl a bit.
4609 71. Recipient callout verification, like sender verification, was using <> in
4610 the MAIL FROM command. This isn't really the right thing, since the actual
4611 sender may affect whether the remote host accepts the recipient or not. I
4612 have changed it to use the actual sender in the callout; this means that
4613 the cache record is now keyed on a recipient/sender pair, not just the
4614 recipient address. There doesn't seem to be a real danger of callout loops,
4615 since a callout by the remote host to check the sender would use <>.
4616 [SEE ABOVE: changed after hitting problems.]
4618 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4619 temporary errors. However, in the case of such a code being given after
4620 the end of a data transmission (i.e. after ".") Exim was failing to write
4621 a retry record for the message. (Yes, there was some broken host that was
4622 actually sending 8xx at this point.)
4624 73. An unknown lookup type in a host list could cause Exim to panic-die when
4625 the list was checked. (An example that provoked this was putting <; in the
4626 middle of a list instead of at the start.) If this happened during a DATA
4627 ACL check, a -D file could be left lying around. This kind of configuration
4628 error no longer causes Exim to die; instead it causes a defer errror. The
4629 incident is still logged to the main and panic logs.
4631 74. Buglet left over from Exim 3 conversion. The message "too many messages
4632 in one connection" was written to the rejectlog but not the mainlog, except
4633 when address rewriting (yes!) was being logged.
4635 75. Added write_rejectlog option.
4637 76. When a system filter was run not as root (that is, when system_filter_user
4638 was set), the values of the $n variables were not being returned to the
4639 main process; thus, they were not subsequently available in the $sn
4642 77. Added +return_path_on_delivery log selector.
4644 78. A connection timeout was being treated differently from recipients deferred
4645 when testing hosts_max_try with a message that was older than the host's
4646 retry timeout. (The host should not be counted, thus allowing all hosts to
4647 be tried at least once before bouncing.) This may have been the cause of an
4648 occasionally reported bug whereby a message would remain on the queue
4649 longer than the retry timeout, but would be bounced if a delivery was
4650 forced. I say "may" because I never totally pinned down the problem;
4651 setting up timeout/retry tests is difficult. See also the next item.
4653 79. The ultimate address timeout was not being applied to errors that involved
4654 a combination of host plus message (for example, a timeout on a MAIL
4655 command). When an address resolved to a number of possible hosts, and they
4656 were not all tried for each delivery (e.g. because of hosts_max_try), a
4657 message could remain on the queue longer than the retry timeout.
4659 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4662 81. Fixed an obscure SMTP outgoing bug which required at least the following
4663 conditions: (a) there was another message waiting for the same server;
4664 (b) the server returned 5xx to all RCPT commands in the first message so
4665 that the message was not completed; (c) the server dropped the connection
4666 or gave a negative response to the RSET that Exim sends to abort the
4667 transaction. The observed case was a dropped connection after DATA that had
4668 been sent in pipelining mode. That is, the server had advertised PIPELINING
4669 but was not implementing it correctly. The effect of the bug was incorrect
4670 behaviour, such as trying another host, and this could lead to a crash.
4676 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4677 and daemon.c were passed as pointers to ints; they should have been
4678 pointers to socklen_t variables (which are typically unsigned ints).
4680 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4683 3. Fixed a really odd bug that affected only the testing scheme; patching a
4684 certain fixed string in the binary changed the value of another string that
4685 happened to be identical to the end of the original first string.
4687 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4688 name", it returns that address as the IP address. On some operating
4689 systems (e.g. Solaris), it also passes back the IP address string as the
4690 "host name". However, on others (e.g. Linux), it passes back an empty
4691 string. Exim wasn't checking for this, and was changing the host name to an
4692 empty string, assuming it had been canonicized.
4694 5. Although rare, it is permitted to have more than one PTR record for a given
4695 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4696 all the names associated with an address, because they do in Solaris.
4697 However, it seems that they do not in Linux for data that comes from the
4698 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4699 I found this out when I moved to a new Linux workstation and tried to run
4700 the Exim test suite.
4702 To get round this problem I have changed the code so that it now does its
4703 own call to the DNS to look up PTR records when searching for a host name.
4704 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4705 addresses that are only in /etc/hosts are still found.
4707 This behaviour is, however, controlled by an option called host_lookup_
4708 order, which defaults to "bydns:byaddr". If people want to use the other
4709 order, or indeed, just use one or the other means of lookup, they can
4710 specify it in this variable.
4712 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4713 some operating systems, this comes back from gethostbyaddr() as an empty
4714 string, and this is what Exim used to test for. However, it seems that in
4715 other systems, "." is yielded. Exim now tests for this case too.
4717 7. The values of check_spool_space and check_log_space are now held internally
4718 as a number of kilobytes instead of an absolute number of bytes. If a
4719 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4720 kilobyte. This means that much larger values can be stored.
4722 8. Exim monitor: an attempt to get the action menu when not actually pointing
4723 at a message produces an empty menu entitled "No message selected". This
4724 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4725 no entries in it ("Shell widget menu has zero width and/or height"). So I
4726 have added a single, blank menu entry in this case.
4728 9. Added ${quote_local_part.
4730 10. MIME decoding is now applied to the contents of Subject: header lines when
4733 11. Now that a reference to $sender_host_address automatically causes a reverse
4734 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4735 host lookup before query-style lookups in lists that might use this
4736 variable. This has therefore been abolished, and the "net-" prefix is no
4737 longer necessary for query-style lookups.
4739 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4740 have been a typo for LFLAGS, so it has been changed.
4742 13. The install script calls Exim with "-C /dev/null" in order to find the
4743 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4744 to be output. Howeve, since Exim outputs its version number before the
4745 error, it didn't break the script. It just looked ugly. I fixed this by
4746 always allowing "-C /dev/null" if the caller is root.
4748 14. Ignore overlarge ACL variable number when reading spool file - insurance
4749 against a later release with more variables having written the file.
4751 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4752 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4753 this, as well as the form without the "IPv6" on the front.
4755 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4756 OS/Makefile-Darwin file.
4758 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4759 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4761 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4762 variable. However, some OS use ints for inodes. I've added cast to long int
4763 to get rid of the compiler warning.
4765 19. I had forgotten to lock out "/../" in configuration file names when
4766 ALT_CONFIG_PREFIX was set.
4768 20. Routers used for verification do not need to specify transports. However,
4769 if such a router generated a host list, and callout was configured, Exim
4770 crashed, because it could not find a port number from the (non-existent)
4771 transport. It now assumes port 25 in this circumstance.
4773 21. Added the -t option to exigrep.
4775 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4776 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4777 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4778 them is set, LOOKUP_LSEARCH is forced.
4780 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4782 24. Added sender and host information to the "rejected by local_scan()" log
4783 line; previously there was no indication of these.
4785 25. Added .include_if_exists.
4787 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4788 Linux. It turns out that not all file systems support this. Apparently some
4789 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4790 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4791 not supported on the file descriptor, is now ignored when Exim is trying to
4792 sync a directory. This applies only to Linux.
4794 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4796 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4797 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4800 29. Added support for maildirsize files from supplied patch (modified a bit).
4802 30. The use of :fail: followed by an empty string could lead Exim to respond to
4803 sender verification failures with (e.g.):
4805 550 Verification failed for <xxx>
4806 550 Sender verify failed
4808 where the first response line was missing the '-' that indicates it is not
4809 the final line of the response.
4811 31. The loop for finding the name of the user that called Exim had a hardwired
4812 limit of 10; it now uses the value of finduser_retries, which is used for
4813 all other user lookups.
4815 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4818 33. Exim was neglecting to zero errno before one call of strtol() when
4819 expanding a string and expecting an integer value. On some systems this
4820 resulted in spurious "integer overflow" errors. Also, it was casting the
4821 result into an int without checking.
4823 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4824 did not work. The code looks as if it has *never* worked, though it appears
4825 to have been documented since at least releast 1.62. I have made it work.
4827 35. The "timeout_DNS" error in retry rules, also documented since at least
4828 1.62, also never worked. As it isn't clear exactly what this means, and
4829 clearly it isn't a major issue, I have abolished the feature by treating it
4830 as "timeout", and writing a warning to the main and panic logs.
4832 36. The display of retry rules for -brt wasn't always showing the error code
4835 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4836 timeout_connect_A, timeout_connect_MX.
4838 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4839 to the empty sender.
4841 39. The daemon was not analysing the content of -oX till after it had closed
4842 stderr and disconnected from the controlling terminal. This meant that any
4843 syntax errors were only noted on the panic log, and the return code from
4844 the command was 0. By re-arranging the code a little, I've made the
4845 decoding happen first, so such errors now appear on stderr, and the return
4846 code is 1. However, the actual setting up of the sockets still happens in
4847 the disconnected process, so errors there are still only recorded on the
4850 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4851 connections (as happens on some IP stacks) was logged at start up time as
4852 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4853 it from "IPv6 and IPv4", which means that two separate sockets are being
4856 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4857 says whether AF_INET or AF_INET6 was passed as an argument.
4859 42. Exiwhat output was messed up when time zones were included in log
4862 43. Exiwhat now gives more information about the daemon's listening ports,
4863 and whether -tls-on-connect was used.
4865 44. The "port" option of the smtp transport is now expanded.
4867 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4868 silently ignored. Now an error message is written to the main and panic
4871 46. There's a new ACL modifier called "logwrite" which writes to a log file
4872 as soon as it is encountered.
4874 47. Added $local_user_uid and $local_user_gid at routing time.
4876 48. Exim crashed when trying to verify a sender address that was being
4879 49. Exim was recognizing only a space character after ".include". It now also
4880 recognizes a tab character.
4882 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4883 extracting the relevant information from the specification. The man page no
4884 longer contains scrambled data for the -d option, and I've added a section
4885 at the front about calling Exim under different names.
4887 51. Added "extra_headers" argument to the "mail" command in filter files.
4889 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4892 53. Installed eximstats 1.29.
4894 54. Added transport_filter_timeout as a generic transport option.
4896 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4897 Cc: header lines. This was required by RFC 822, but it not required by RFC
4900 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4901 incoming messages that did not have them. Now it does so only if the
4902 message originates locally, that is, if there is no associated remote host
4903 address. When Resent- header lines are present, this applies to the Resent-
4904 lines rather than the non-Resent- lines.
4906 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4907 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4909 58. Messages for configuration errors now include the name of the main
4910 configuration file - useful now that there may be more than one file in a
4911 list (.included file names were always shown).
4913 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4914 for those rare installations that do not start the daemon as root or run it
4915 setuid root. I've cut out the call to initgroups() if the daemon is not
4918 60. The Exim user and group can now be bound into the binary as text strings
4919 that are looked up at the start of Exim's processing.
4921 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4923 62. Added $mailstore_basename variable.
4925 63. Installed patch to sieve.c from Michael Haardt.
4927 64. When Exim failed to open the panic log after failing to open the main log,
4928 the original message it was trying to log was written to stderr and debug
4929 output, but if they were not available (the usual case in production), it
4930 was lost. Now it is written to syslog before the two lines that record the
4931 failures to open the logs.
4933 65. Users' Exim filters run in subprocesses under the user's uid. It is
4934 possible for a "deliver" command or an alias in a "personal" command to
4935 provoke an address rewrite. If logging of address rewriting is configured,
4936 this fails because the process is not running as root or exim. There may be
4937 a better way of dealing with this, but for the moment (because 4.30 needs
4938 to be released), I have disabled address rewrite logging when running a
4939 filter in a non-root, non-exim process.
4945 1. The buildconfig auxiliary program wasn't quoting the value set for
4946 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4947 not defined. This bug was masked in 4.22 by the effect that was fixed in
4950 2. Some messages that were rejected after a message id was allocated were
4951 shown as "incomplete" by exigrep. It no longer does this for messages that
4952 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4954 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4955 have allow_domain_literals set, the ID did not get logged in the <= line.
4956 Domain literals are now always recognized in Message-ID: header lines.
4958 4. The first argument for a ${extract expansion item is the key name or field
4959 number. Leading and trailing spaces in this item were not being ignored,
4960 causing some misleading effects.
4962 5. When deliver_drop_privilege was set, single queue runner processes started
4963 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4964 same command in the process it spins off) were not dropping privilege.
4966 6. When the daemon running as "exim" started a queue runner, it always
4967 re-executed Exim in the spun-off process. This is a waste of effort when
4968 deliver_drop_privilege is set. The new process now just calls the
4969 queue-runner function directly.
4975 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4978 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4979 ignored. Though the use of -oP was forcing the writing of a pid file, it
4980 was always written to the default place.
4982 3. If the message "no IP address found for host xxxx" is generated during
4983 incoming verification, it is now followed by identification of the incoming
4984 connection (so you can more easily find what provoked it).
4986 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4988 5. Added some features to "harden" Exim a bit more against certain attacks:
4990 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4991 be put in Local/Makefile. This is like the never_users runtime option,
4992 but it cannot be overridden. The default setting is "root".
4994 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4995 prefix string with which any file named in a -C command line option
4998 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4999 is retained for -C and -D only if the caller of Exim is root. Without
5000 it, the exim user may also use -C and -D and retain privilege.
5002 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5003 command line option is disabled.
5005 6. Macro names set by the -D option must start with an upper case letter, just
5006 like macro names defined in the configuration file.
5008 7. Added "dereference=" facility to LDAP.
5010 8. Two instances of the typo "uknown" in the source files are fixed.
5012 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5013 the Configure-Makefile script screwed up while processing it.
5015 10. Incorporated PCRE 4.4.
5017 11. The SMTP synchronization check was not operating right at the start of an
5018 SMTP session. For example, it could not catch a HELO sent before the client
5019 waited for the greeting. There is now a check for outstanding input at the
5020 point when the greeting is written. Because of the duplex, asynchronous
5021 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5022 way, but not yet received, when the check is performed.
5024 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5025 on TCP/IP sockets, because this apparently causes some broken clients to
5028 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5029 unchanged) from the Cygwin maintainer.
5031 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5032 was supported instead of testing for mbx. Effectively a typo.
5034 15. The spa authenticator server code was not checking that the input it
5035 received was valid base64.
5037 16. The debug output line for the "set" modifier in ACLs was not showing the
5038 name of the variable that was being set.
5040 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5042 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5043 lookup on its own. Something else was needed to trigger the lookup. For
5044 example, a match in host_lookup or the need for a host name in a host list.
5045 Now, if $sender_host_name is referenced and the host name has not yet been
5046 looked up, a lookup is performed. If the lookup fails, the variable remains
5047 empty, and $host_lookup_failed is set to "1".
5049 19. Added "eqi" as a case-independent comparison operator.
5051 20. The saslauthd authentication condition could segfault if neither service
5052 nor realm was specified.
5054 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5055 error message that was logged was misleading, and incoming SMTP
5056 connections were dropped. The message is now more accurate, and temporary
5057 errors are given to SMTP connections.
5059 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5060 (except RSET) with a 421 error, until QUIT is received. However, it was
5061 failing to send a response to QUIT.
5063 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5064 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5065 ACL is now run after the helo_try_verify_hosts code.
5067 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5068 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5069 case-independent, so other case variants are also recognized). Apparently
5070 some systems use these upper case variants.
5072 25. If more than two messages were waiting for the same host, and a transport
5073 filter was specified for the transport, Exim sent two messages over the
5074 same TCP/IP connection, and then failed with "socket operation on non-
5075 socket" when it tried to send the third.
5077 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5079 27. The extern definition of crypt16() in expand.c was not being excluded when
5080 the OS had its own crypt16() function.
5082 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5083 as a preferred synonym for return_size_limit, both as an option and as an
5086 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5088 30. Changed the default configuration ACL to relax the local part checking rule
5089 for addresses that are not in any local domains. For these addresses,
5090 slashes and pipe symbols are allowed within local parts, but the sequence
5091 /../ is explicitly forbidden.
5093 31. SPA server authentication was not clearing the challenge buffer before
5096 32. log_message in a "warn" ACL statement was writing to the reject log as
5097 well as to the main log, which contradicts the documentation and doesn't
5098 seem right (because no rejection is happening). So I have stopped it.
5100 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5101 However, I am unable to do any testing of this.
5103 34. Fixed an infelicity in the appendfile transport. When checking directories
5104 for a mailbox, to see if any needed to be created, it was accidentally
5105 using path names with one or more superfluous leading slashes; tracing
5106 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5108 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5109 contents are added to the log line that is written for every discarded
5110 recipient. (Previously a log_message setting was ignored.)
5112 36. The ${quote: operator now quotes the string if it is empty.
5114 37. The install script runs exim in order to find its version number. If for
5115 some reason other than non-existence or emptiness, which it checks, it
5116 could not run './exim', it was installing it with an empty version number,
5117 i.e. as "exim-". This error state is now caught, and the installation is
5120 38. An argument was missing from the function that creates an error message
5121 when Exim fails to connect to the socket for saslauthd authentication.
5122 This could cause Exim to crash, or give a corrupted message.
5124 39. Added isip, isip4, and isip6 to ${if conditions.
5126 40. The ACL variables $acl_xx are now saved with the message, and can be
5127 accessed later in routers, transports, and filters.
5129 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5130 strings in the file are not string-expanded.
5132 42. If a MAIL command specified a SIZE value that was too large to fit into an
5133 int variable, the check against message_size_limit failed. Such values are
5134 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5135 day this will have to be increased, but I don't think I want to be around
5136 when emails are that large.
5143 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5144 iconv() is not standard in FreeBSD.
5146 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5147 IPv6 enabled. The observed symptom was a segmentation fault on return from
5148 the function os_common_find_running_interfaces() in src/os.c.
5150 3. In the check_special_case() function in daemon.c I had used "errno" as an
5151 argument name, which causes warnings on some systems. This was basically a
5152 typo, since it was named "eno" in the comments!
5154 4. The code that waits for the clock to tick (at a resolution of some fraction
5155 of a second) so as to ensure message-id uniqueness was always waiting for
5156 at least one whole tick, when it could have waited for less. [This is
5157 almost certainly not relevant at current processor speeds, where it is
5158 unlikely to ever wait at all. But we try to future-proof.]
5160 5. The function that sleeps for a time interval that includes fractions of a
5161 second contained a race. It did not block SIGALRM between setting the
5162 timer, and suspending (a couple of lines later). If the interval was short
5163 and the sigsuspend() was delayed until after it had expired, the suspension
5164 never ended. On busy systems this could lead to processes getting stuck for
5167 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5168 process, before it forks any delivery processes. The open lookup caching
5169 mechanism meant that the open file or database connection was passed into
5170 the delivery process. The problem was that delivery processes always tidy
5171 up cached lookup data. This could cause a problem for the next delivery
5172 process started by the queue runner, because the external queue runner
5173 process does not know about the closure. So the next delivery process
5174 still has data in the lookup cache. In the case of a file lookup, there was
5175 no problem because closing a file descriptor in a subprocess doesn't affect
5176 the parent. However, if the lookup was caching a connection to a database,
5177 the connection was closed, and the second delivery process was likely to
5178 see errors such as "PGSQL: query failed: server closed the connection
5179 unexpectedly". The problem has been fixed by closing all cached lookups
5180 in a queue runner before running a delivery process.
5182 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5183 seem to have the "const" qualifier which it has on other OS. I've
5186 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5187 *of the same type* (client or server) with the same public name that an
5188 error should be diagnosed.
5190 9. When Exim looked up a host name for an IP address, but failed to find the
5191 original IP address when looking up the host name (a safety check), it
5192 output the message "<ip address> does not match any IP for NULL", which was
5193 confusing, to say the least. The bug was that the host name should have
5194 appeared instead of "NULL".
5196 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5197 user that is built into the binary, and the -C or -D options is used, root
5198 privilege is dropped before the configuration file is read. In addition,
5199 logging is switched to stderr instead of the normal log files. If the
5200 configuration then re-defines the Exim user, the unprivileged environment
5201 is probably not what is expected, so Exim logs a panic warning message (but
5204 However, if deliver_drop_privilege is set, the unprivileged state may well
5205 be exactly what is intended, so the warning has been cut out in that case,
5206 and Exim is allowed to try to write to its normal log files.
5212 1. smtp_return_error_details was not giving details for temporary sender
5213 or receiver verification errors.
5215 2. Diagnose a configuration error if two authenticators have the same public
5218 3. Exim used not to create the message log file for a message until the first
5219 delivery attempt. This could be confusing when incoming messages were held
5220 for policy or load reasons. The message log file is now created at the time
5221 the message is received, and an initial "Received" line is written to it.
5223 4. The automatically generated man page for command line options had a minor
5224 bug that caused no ill effects; however, a more serious problem was that
5225 the procedure for building the man page automatically didn't always
5226 operate. Consequently, release 4.20 contains an out-of-date version. This
5227 shouldn't happen again.
5229 5. When building Exim with embedded Perl support, the script that builds the
5230 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5231 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5233 6. The freeze_tell option was not being used for messages that were frozen on
5234 arrival, either by an ACL or by local_scan().
5236 7. Added the smtp_incomplete_transaction log selector.
5238 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5239 was accepting AUTH without a new EHLO.
5241 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5242 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5245 10. Logging of TCP/IP connections (when configured) now happens in the main
5246 daemon process instead of the child process, so that the TCP/IP connection
5247 count is more accurate (but it can never be perfect).
5249 11. The use of "drop" in a nested ACL was not being handled correctly in the
5250 outer ACL. Now, if condition failure induced by the nested "drop" causes
5251 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5252 or "require"), the connection is dropped.
5254 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5255 that yield "discard" can now be used with an "accept" or a "discard" verb,
5256 but an error is generated for any others (because I can't see a useful way
5257 to define what should happen).
5259 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5260 are now processed in the same way as lines in the Exim configuration file.
5261 In particular, continuation lines are supported.
5263 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5265 15. Added -ti meaning -t -i.
5267 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5268 domains, and warn by logging if others are found.
5270 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5271 returned from GIFCONF when Exim is trying to find the list of interfaces on
5272 a host. The code in os.c has been modified to copy each ifreq to an aligned
5273 structure in all cases.
5275 Also, in some cases, the returned ifreq's were being copied to a 'struct
5276 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5277 means the last couple of bytes of an IPv6 address could be chopped if the
5278 ifreq contained only a normal sockaddr (14 bytes storage).
5280 18. Named domain lists were not supported in the hosts_treat_as_local option.
5281 An entry such as +xxxx was not recognized, and was treated as a literal
5284 19. Ensure that header lines added by a DATA ACL are included in the reject log
5285 if the ACL subsequently rejects the message.
5287 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5288 MD5 (which is deprecated).
5290 21. When testing a filter file using -bf, Exim was writing a message when it
5291 took the sender from a "From " line in the message, but it was not doing so
5292 when it took $return_path from a Return-Path: header line. It now does.
5294 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5295 with a valid header line field name (a series of printing characters
5296 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5298 23. Changed "disc" in the source to "disk" to conform to the documentation and
5299 the book and for uniformity.
5301 24. Ignore Sendmail's -Ooption=value command line item.
5303 25. When execve() failed while trying to run a command in a pipe transport,
5304 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5305 could be confused with a return value of 69 from the command itself. This
5306 has been changed to 127, the value the shell returns if it is asked to run
5307 a non-existent command. The wording for the related log line suggests a
5308 non-existent command as the problem.
5310 26. If received_header_text expands to an empty string, do not add a Received:
5311 header line to the message. (Well, it adds a token one on the spool, but
5312 marks it "old" so that it doesn't get used or transmitted.)
5314 27. Installed eximstats 1.28 (addition of -nt option).
5316 28. There was no check for failure on the call to getsockname() in the daemon
5317 code. This can fail if there is a shortage of resources on the system, with
5318 ENOMEM, for example. A temporary error is now given on failure.
5320 29. Contrary to the C standard, it seems that in some environments, the
5321 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5322 program. Exim now does this explicitly; it affects the formatting of
5323 timestamps using strftime().
5325 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5326 complaints. I've now initialized all the variables, to avoid this.
5328 32. Header lines added by a system filter were not being "seen" during
5329 transport-time rewrites.
5331 33. The info_callback() function passed to OpenSSL is set up with type void
5332 (*)(SSL *, int, int), as described somewhere. However, when calling the
5333 function (actually a macro) that sets it up, the type void(*)() is
5334 expected. I've put in a cast to prevent warnings from picky compilers.
5336 34. If a DNS black list lookup found a CNAME record, but there were no A
5337 records associated with the domain it pointed at, Exim crashed.
5339 35. If a DNS black list lookup returned more than one A record, Exim ignored
5340 all but the first. It now scans all returned addresses if a particular IP
5341 value is being sought. In this situation, the contents of the
5342 $dnslist_value variable are a list of all the addresses, separated by a
5345 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5346 to accept a host name and all its aliases if the forward lookup for any of
5347 them yielded the IP address of the incoming connection. Now it accepts only
5348 those names whose forward lookup yields the correct IP address. Any other
5349 names are discarded. This closes a loophole whereby a rogue DNS
5350 administrator could create reverse DNS records to break through a
5351 wildcarded host restriction in an ACL.
5353 37. If a user filter or a system filter that ran in a subprocess used any of
5354 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5355 the wrong values were passed to the pipe command ($thisaddress had the
5356 value of $0, $0 had the value of $1, etc). This bug was introduced by
5357 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5359 38. Improved the line breaking for long SMTP error messages from ACLs.
5360 Previously, if there was no break point between 40 and 75 characters, Exim
5361 left the rest of the message alone. Two changes have been made: (a) I've
5362 reduced the minimum length to 35 characters; (b) if it can't find a break
5363 point between 35 and 75 characters, it looks ahead and uses the first one
5364 that it finds. This may give the occasional overlong line, but at least the
5365 remaining text gets split now.
5367 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5368 file descriptors might be low, and that setting 1000 would always raise it.
5369 It turns out that in some environments, the limit is already over 1000 and
5370 that lowering it causes trouble. So now Exim takes care not to decrease it.
5372 40. When delivering a message, the value of $return_path is set to $sender_
5373 address at the start of routing (routers may change the value). By an
5374 oversight, this default was not being set up when an address was tested by
5375 -bt or -bv, which affected the outcome if any router or filter referred to
5378 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5379 log only when "message" or "log_message" are set. However, if one of the
5380 conditions was an address verification, or a call to a nested ACL, the
5381 messages generated by the underlying test were being passed through. This
5382 no longer happens. The underlying message is available in $acl_verify_
5383 message for both "message" and "log_message" expansions, so it can be
5384 passed through if needed.
5386 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5387 new expansion $bh_ to give the encoded byte string without charset
5388 translation. Translation happens only if iconv() is available; HAVE_ICONV
5389 indicates this at build time. HEADERS_CHARSET gives the charset to
5390 translate to; headers_charset can change it in the configuration, and
5391 "headers charset" can change it in an individual filter file.
5393 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5394 that creates RFC 2047 encoded "words" labels them as that charset instead
5395 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5396 expansion operator; (ii) when Exim creates a From: line for a local
5397 message; (iii) when a header line is rewritten to include a "phrase" part.
5399 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5400 buggy, causing it to skip the first lines of messages whose message ID
5401 ended in 'D'. This would not have bitten before Exim release 4.14, because
5402 message IDs were unlikely to end in 'D' before then. The effect was to have
5403 incorrect size information for certain domains.
5405 45. #include "config.h" was missing at the start of the crypt16.c module. This
5406 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5409 46. If there was a timeout during a "random" callout check, Exim treated it as
5410 a failure of the random address, and carried on sending RSET and the real
5411 address. If the delay was just some slowness somewhere, the response to the
5412 original RCPT would be taken as a response to RSET and so on, causing
5413 mayhem of various kinds.
5415 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5416 when I implemented it. It didn't allow for the fact that some option values
5417 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5418 the right test for positive values.
5420 48. Domain names in DNS records are case-independent. Exim always looks them up
5421 in lower case. Some resolvers return domain names in exactly the case they
5422 appear in the zone file, that is, they may contain uppercase letters. Not
5423 all resolvers do this - some return always lower case. Exim was treating a
5424 change of case by a resolver as a change of domain, similar to a widening
5425 of a domain abbreviation. This triggered its re-routing code and so it was
5426 trying to route what was effectively the same domain again. This normally
5427 caused routing to fail (because the router wouldn't handle the domain
5428 twice). Now Exim checks for this case specially, and just changes the
5429 casing of the domain that it ultimately uses when it transmits the message
5432 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5435 50. If a filter generated a file delivery with a non-absolute name (possible if
5436 no home directory exists for the router), the forbid_file option was not
5439 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5440 the existing equality matching.
5442 52. Exim was using ints instead of ino_t variables in some places where it was
5443 dealing with inode numbers.
5445 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5446 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5447 called TMPDIR, and if it finds it is different, it changes its value.
5449 54. The smtp_printf() function is now made available to local_scan() so
5450 additional output lines can be written before returning. There is also an
5451 smtp_fflush() function to enable the detection of a dropped connection.
5452 The variables smtp_input and smtp_batched_input are exported to
5455 55. Changed the default runtime configuration: the message "Unknown user"
5456 has been removed from the ACL, and instead placed on the localuser router,
5457 using the cannot_route_message feature. This means that any verification
5458 failures that generate their own messages won't get overridden. Similarly,
5459 the "Unrouteable address" message that was in the ACL for unverifiable
5460 relay addresses has also been removed.
5462 56. Added hosts_avoid_esmtp to the smtp transport.
5464 57. The exicyclog script was not checking for the esoteric option
5465 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5466 will work only if exicyclog is run under the appropriate euid.
5468 58. Following a discussion on the list, the rules by which Exim recognises line
5469 endings on incoming messages have been changed. The -dropcr and drop_cr
5470 options are now no-ops, retained only for backwards compatibility. The
5471 following line terminators are recognized: LF CRLF CR. However, special
5472 processing applies to CR:
5474 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5475 nor a local message in the state where . is a terminator.
5477 (ii) If a bare CR is encountered in a header line, an extra space is added
5478 after the line terminator so as not to end the header. The reasoning
5479 behind this is that bare CRs in header lines are most likely either
5480 to be mistakes, or people trying to play silly games.
5482 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5483 was being incorrectly given as 18 bytes larger than it should have been.
5484 This is a VOB (very old bug).
5486 60. This may never have affected anything current, but just in case it has:
5487 When the local host is found other than at the start of a list of hosts,
5488 the local host, those with the same MX, and any that follow, are discarded.
5489 When the list in question was part of a longer list of hosts, the following
5490 hosts (not currently being processed) were also being discarded. This no
5491 longer happens. I'm not sure if this situation could ever has previously
5494 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5497 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5498 header. This is something that is recommended in a new Internet Draft, and
5499 is something that is documented as being done by Sendmail. There are two
5500 possible values. For messages generated by the autoreply transport, Exim
5503 Auto-Submitted: auto-replied
5505 whereas for all other generated messages (e.g. bounces) it adds
5507 Auto-Submitted: auto-generated
5509 63. The "personal" condition in filters now includes a test for the
5510 Auto-Submitted: header. If it contains the string "auto-" the message it
5511 not considered personal.
5513 64. Added rcpt_include_affixes as a generic transport option.
5515 65. Added queue_only_override (default true).
5517 66. Added the syslog_duplication option.
5519 67. If what should have been the first header line of a message consisted of
5520 a space followed by a colon, Exim was mis-interpreting it as a header line.
5521 It isn't of course - it is syntactically invalid and should therefore be
5522 treated as the start of the message body. The misbehaviour could have
5523 caused a number of strange effects, including loss of data in subsequent
5524 header lines, and spool format errors.
5526 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5527 client host had authenticated. This control can now be exercised by an ACL
5528 for more flexibility.
5530 69. By default, callouts do not happen when testing with -bh. There is now a
5531 variant, -bhc, which does actually run the callout code, including
5532 consulting and updating the callout cache.
5534 70. Added support for saslauthd authentication, courtesy of Alexander
5537 71. If statvfs() failed on the spool or log directories while checking their
5538 size for availability, Exim confusingly gave the error "space shortage".
5539 Furthermore, in debugging mode it crashed with a floating point exception.
5540 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5541 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5542 serious problem, Exim now writes to the main and panic logs when this
5543 happens, with details of the failure. It then refuses to accept the
5544 incoming message, giving the message "spool directory problem" or "log
5545 directory problem" with a 421 code for SMTP messages.
5547 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5548 0, 1, and 2 exist, because some OS complain for execs without them (see
5549 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5550 descriptors. However, the code omitted to check that the open succeeded,
5551 causing mysterious errors if for some reason the permissions on /dev/null
5552 got screwed. Now Exim writes a message to the main and panic logs, and
5553 bombs out if it can't open /dev/null.
5555 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5556 interact so that it is all more flexible. It is supposed to remain
5557 backwards compatible. Also added extra_local_interfaces.
5559 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5560 to bomb out with an assertion failure - to the client this appears as a
5561 connection drop. This problem occurs in the part of the code that was taken
5562 from the Samba project. Fortunately, the assertion is in a very simple
5563 function, so I have fixed this by reproducing the function inline in the
5564 one place where it is called, and arranging for authentication to fail
5565 instead of killing the process with assert().
5567 75. The SPA client code was not working when the server requested OEM rather
5568 than Unicode encoding.
5570 76. Added code to make require_files with a specific uid setting more usable in
5571 the case where statting the file as root fails - usually a non-root-mounted
5572 NFS file system. When this happens and the failure is EACCES, Exim now
5573 forks a subprocess and does the per-uid checking as the relevant uid.
5575 77. Added process_log_path.
5577 78. If log_file_path was not explicitly set, a setting of check_log_space or
5578 check_log_inodes was ignored.
5580 79. If a space check for the spool or log partitions fails, the incident is now
5581 logged. Of course, in the latter case the data may get lost...
5583 80. Added the %p formatting code to string_format() so that it can be used to
5584 print addresses in debug_print(). Adjusted all the address printing in the
5585 debugging in store.c to use %p rather than %d.
5587 81. There was a concern that a line of code in smtp_in.c could overflow a
5588 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5589 initially expressed, the concern was not well-founded, because trailing
5590 spaces are removed early. However, if the trailing spaces were followed by
5591 a NULL, they did not get removed, so the overflow was possible. Two fixes
5594 (a) I re-wrote the offending code in a cleaner fashion.
5595 (b) If an incoming SMTP command contains a NULL character, it is rejected
5598 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5599 runs initgroups(), so that if the Exim user is in any additional groups,
5600 they will be used during message reception.
5606 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob