1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
70 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
72 Report and patch from Dmitry Banschikov.
74 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
75 Removed SSL_clear() after SSL_new() which led to protocol negotiation
76 failures. We appear to now support TLS1.1+ with Exim.
78 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
79 lets Exim select keys and certificates based upon TLS SNI from client.
80 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
81 before an outbound SMTP session. New log_selector, +tls_sni.
83 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
84 NULL dereference. Report and patch from Alun Jones.
86 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
87 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
88 Not seeing resolver debug output on NetBSD, but suspect this is a
89 resolver implementation change.
91 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
92 Left warnings. Added "eximon gdb" invocation mode.
94 PP/21 Defaulting "accept_8bitmime" to true, not false.
96 PP/22 Added -bw for inetd wait mode support.
98 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
99 locate the relevant includes and libraries. Made this the default.
101 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
102 Bugzilla 1246, report and most of solution from Tomasz Kusy.
104 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
105 This may cause build issues on older platforms.
107 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
108 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
109 gnutls_require_mac & gnutls_require_protocols (no longer supported).
110 Added SNI support via GnuTLS too.
111 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
113 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
115 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
116 Applied second patch from Janne, implementing suggestion to default
117 multiple-strings-in-record handling to match SPF spec.
119 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
121 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
122 read-only, out of scope).
123 Patch from Wolfgang Breyha, report from Stuart Northfield.
125 PP/29 Fix three issues highlighted by clang analyser static analysis.
126 Only crash-plausible issue would require the Cambridge-specific
127 iplookup router and a misconfiguration.
128 Report from Marcin Mirosław.
130 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
136 PP/01 Solaris build fix for Oracle's LDAP libraries.
137 Bugzilla 1109, patch from Stephen Usher.
139 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
141 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
144 TF/02 Fix a couple more cases where we did not log the error message
145 when unlink() failed. See also change 4.74-TF/03.
147 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
148 lock up or crash if it happened to be inside a call to libc when it
149 got a SIGUSR1 from exiwhat.
151 The SIGUSR1 handler appends the current process status to the process
152 log which is later printed by exiwhat. It used to use the general
153 purpose logging code to do this, but several functions it calls are
154 not safe for signals.
156 The new output code in the SIGUSR1 handler is specific to the process
157 log, and simple enough that it's easy to inspect for signal safety.
158 Removing some special cases also simplifies the general logging code.
159 Removing the spurious timestamps from the process log simplifies
162 TF/04 Improved ratelimit ACL condition.
164 The /noupdate option has been deprecated in favour of /readonly which
165 has clearer semantics. The /leaky, /strict, and /readonly update modes
166 are mutually exclusive. The update mode is no longer included in the
167 database key; it just determines when the database is updated. (This
168 means that when you upgrde Exim will forget old rate measurements.)
170 Exim now checks that the per_* options are used with an update mode that
171 makes sense for the current ACL. For example, when Exim is processing a
172 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
173 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
174 must specify per_mail/readonly. If you omit the update mode it defaults to
175 /leaky where that makes sense (as before) or /readonly where required.
177 The /noupdate option is now undocumented but still supported for
178 backwards compatibility. It is equivalent to /readonly except that in
179 ACLs where /readonly is required you may specify /leaky/noupdate or
180 /strict/noupdate which are treated the same as /readonly.
182 A useful new feature is the /count= option. This is a generalization
183 of the per_byte option, so that you can measure the throughput of other
184 aggregate values. For example, the per_byte option is now equivalent
185 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
187 The per_rcpt option has been generalized using the /count= mechanism
188 (though it's more complicated than the per_byte equivalence). When it is
189 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
190 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
191 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
192 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
193 Note that using per_rcpt with a non-readonly update mode in more than
194 one ACL will cause the recipients to be double-counted. (The per_mail
195 and per_byte options don't have this problem.)
197 The handling of very low rates has changed slightly. If the computed rate
198 is less than the event's count (usually one) then this event is the first
199 after a long gap. In this case the rate is set to the same as this event's
200 count, so that the first message of a spam run is counted properly.
202 The major new feature is a mechanism for counting the rate of unique
203 events. The new per_addr option counts the number of different
204 recipients that someone has sent messages to in the last time period. It
205 behaves like per_rcpt if all the recipient addresses are different, but
206 duplicate recipient addresses do not increase the measured rate. Like
207 the /count= option this is a general mechanism, so the per_addr option
208 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
209 example, measure the rate that a client uses different sender addresses
210 with the options per_mail/unique=$sender_address. There are further
211 details in the main documentation.
213 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
215 TF/06 Removed a few PCRE remnants.
217 TF/07 Automatically extract Exim's version number from tags in the git
218 repository when doing development or release builds.
220 PP/02 Raise smtp_cmd_buffer_size to 16kB.
221 Bugzilla 879. Patch from Paul Fisher.
223 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
224 Heavily based on revision 40f9a89a from Simon Arlott's tree.
227 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
229 PP/05 Variable $av_failed, true if the AV scanner deferred.
230 Bugzilla 1078. Patch from John Horne.
232 PP/06 Stop make process more reliably on build failure.
233 Bugzilla 1087. Patch from Heiko Schlittermann.
235 PP/07 Make maildir_use_size_file an _expandable_ boolean.
236 Bugzilla 1089. Patch from Heiko Schlittermann.
238 PP/08 Handle ${run} returning more data than OS pipe buffer size.
239 Bugzilla 1131. Patch from Holger Weiß.
241 PP/09 Handle IPv6 addresses with SPF.
242 Bugzilla 860. Patch from Wolfgang Breyha.
244 PP/10 GnuTLS: support TLS 1.2 & 1.1.
246 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
249 PP/11 match_* no longer expand right-hand-side by default.
250 New compile-time build option, EXPAND_LISTMATCH_RHS.
251 New expansion conditions, "inlist", "inlisti".
253 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
255 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
257 PP/14 fix log_write() format string regression from TF/03.
258 Bugzilla 1152. Patch from Dmitry Isaikin.
264 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
266 PP/02 Harmonised TLS library version reporting; only show if debugging.
267 Layout now matches that introduced for other libraries in 4.74 PP/03.
269 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
271 PP/04 New "dns_use_edns0" global option.
273 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
276 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
277 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
279 TK/01 Updated PolarSSL code to 0.14.2.
280 Bugzilla 1097. Patch from Andreas Metzler.
282 PP/07 Catch divide-by-zero in ${eval:...}.
285 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
288 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
289 format-string attack -- SECURITY: remote arbitrary code execution.
291 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
292 time unintentionally subject to list matching rules, letting the header
293 cause arbitrary Exim lookups (of items which can occur in lists, *not*
294 arbitrary string expansion). This allowed for information disclosure.
296 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
297 INT_MIN/-1 -- value coerced to INT_MAX.
303 NM/01 Workround for PCRE version dependency in version reporting
306 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
307 This fixes portability to compilers other than gcc, notably
308 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
310 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
311 makefiles for portability to HP-UX and POSIX correctness.
313 PP/01 Permit LOOKUP_foo enabling on the make command-line.
314 Also via indented variable definition in the Makefile.
315 (Debugging by Oliver Heesakkers).
317 PP/02 Restore caching of spamd results with expanded spamd_address.
318 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
320 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
321 Improves build reliability. Fix from: Frank Elsner
323 NM/02 Fix wide character breakage in the rfc2047 coding
324 Fixes bug 1064. Patch from Andrey N. Oktyabrski
326 NM/03 Allow underscore in dnslist lookups
327 Fixes bug 1026. Patch from Graeme Fowler
329 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
330 Code patches from Adam Ciarcinski of NetBSD.
332 NM/04 Fixed exiqgrep to cope with mailq missing size issue
335 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
336 is logged, to avoid truncation. Patch from John Horne.
338 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
339 Patch from Jakob Hirsch.
341 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
342 SQL string expansion failure details.
343 Patch from Andrey Oktyabrski.
345 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
346 Patch from Simon Arlott.
348 PP/09 New lookups functionality failed to compile on old gcc which rejects
349 extern declarations in function scope.
350 Patch from Oliver Fleischmann
352 PP/10 Use sig_atomic_t for flags set from signal handlers.
353 Check getgroups() return and improve debugging.
354 Fixed developed for diagnosis in bug 927 (which turned out to be
357 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
358 Patch from Mark Zealey.
360 PP/12 Bugzilla 1056: Improved spamd server selection.
361 Patch from Mark Zealey.
363 PP/13 Bugzilla 1086: Deal with maildir quota file races.
364 Based on patch from Heiko Schlittermann.
366 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
367 Patch from Uwe Doering, sign-off by Michael Haardt.
369 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
370 variable declaration deep within a block. Bug and patch from
373 PP/15 lookups-Makefile IRIX compatibilty coercion.
375 PP/16 Make DISABLE_DKIM build knob functional.
377 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
378 Patch by Simon Arlott
380 TF/03 Fix valgrind.h portability to C89 compilers that do not support
381 variable argument macros. Our copy now differs from upstream.
387 TF/01 Failure to get a lock on a hints database can have serious
388 consequences so log it to the panic log.
390 TF/02 Log LMTP confirmation messages in the same way as SMTP,
391 controlled using the smtp_confirmation log selector.
393 TF/03 Include the error message when we fail to unlink a spool file.
395 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
396 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
397 for maintaining out-of-tree patches for some time.
399 PP/01 Bugzilla 139: Documentation and portability issues.
400 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
401 Handle per-OS dynamic-module compilation flags.
403 PP/02 Let /dev/null have normal permissions.
404 The 4.73 fixes were a little too stringent and complained about the
405 permissions on /dev/null. Exempt it from some checks.
406 Reported by Andreas M. Kirchwitz.
408 PP/03 Report version information for many libraries, including
409 Exim version information for dynamically loaded libraries. Created
410 version.h, now support a version extension string for distributors
411 who patch heavily. Dynamic module ABI change.
413 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
414 privilege escalation vulnerability whereby the Exim run-time user
415 can cause root to append content of the attacker's choosing to
418 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
421 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
422 If dropping privileges for untrusted macros, we disabled normal logging
423 on the basis that it would fail; for the Exim run-time user, this is not
424 the case, and it resulted in successful deliveries going unlogged.
425 Fixed. Reported by Andreas Metzler.
431 PP/01 Date: & Message-Id: revert to normally being appended to a message,
432 only prepend for the Resent-* case. Fixes regression introduced in
433 Exim 4.70 by NM/22 for Bugzilla 607.
435 PP/02 Include check_rfc2047_length in configure.default because we're seeing
436 increasing numbers of administrators be bitten by this.
438 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
440 PP/03 Bugzilla 994: added openssl_options main configuration option.
442 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
444 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
446 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
448 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
449 without a peer certificate, leading to a segfault because of an
450 assumption that peers always have certificates. Be a little more
451 paranoid. Problem reported by Martin Tscholak.
453 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
454 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
455 NB: ClamAV planning to remove STREAM in "middle of 2010".
456 CL also introduces -bmalware, various -d+acl logging additions and
457 more caution in buffer sizes.
459 PP/09 Implemented reverse_ip expansion operator.
461 PP/10 Bugzilla 937: provide a "debug" ACL control.
463 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
465 PP/12 Bugzilla 973: Implement --version.
467 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
469 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
471 PP/15 Bugzilla 816: support multiple condition rules on Routers.
473 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
474 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
475 ignore trailing whitespace.
477 JJ/02 prevent non-panic DKIM error from being sent to paniclog
479 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
482 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
483 Notification from Dr Andrew Aitchison.
485 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
486 ExtendedDetectionInfo response format.
487 Notification from John Horne.
489 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
492 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
493 XSL and documented dependency on system catalogs, with examples of how
496 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
499 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
500 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
501 configuration file which is writeable by the Exim user or group.
503 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
504 of configuration files to cover files specified with the -C option if
505 they are going to be used with root privileges, not just the default
508 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
509 option (effectively making it always true).
511 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
512 files to be used while preserving root privileges.
514 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
515 that rogue child processes cannot use them.
517 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
518 run-time user, instead of root.
520 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
521 Exim run-time user without dropping privileges.
523 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
524 result string, instead of calling string_vformat() twice with the same
527 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
528 for other users. Others should always drop root privileges if they use
529 -C on the command line, even for a whitelisted configure file.
531 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
533 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
539 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
540 $data_path, and $header_path variables; fixed documentation bugs and
543 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
544 exipick to access non-standard spools, including the "frozen" queue
547 NM/01 Bugzilla 965: Support mysql stored procedures.
548 Patch from Alain Williams
550 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
552 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
553 Patch from Andreas Metzler
555 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
556 Patch from Kirill Miazine
558 NM/05 Bugzilla 671: Added umask to procmail example.
560 JJ/03 installed exipick 20100323.0, fixing doc bug
562 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
563 directory. Notification and patch from Dan Rosenberg.
565 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
567 TK/02 Improve log output when DKIM signing operation fails.
569 MH/01 Treat the transport option dkim_domain as a colon separated
570 list, not as a single string, and sign the message with each element,
571 omitting multiple occurences of the same signer.
573 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
574 Bugzilla 985, 986. Patch by Simon Arlott
576 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
577 Patch by Simon Arlott
579 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
580 MBX locking. Notification from Dan Rosenberg.
586 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
588 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
590 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
592 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
594 NM/04 Bugzilla 915: Flag broken perl installation during build.
600 TK/01 Added patch by Johannes Berg that expands the main option
601 "spamd_address" if it starts with a dollar sign.
603 TK/02 Write list of recipients to X-Envelope-Sender header when building
604 the mbox-format spool file for content scanning (suggested by Jakob
607 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
608 (http://www.dcc-servers.net/) support via dccifd. Activated by
609 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
611 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
612 by Mark Daniel Reidel <mr@df.eu>.
614 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
615 When building exim an external PCRE library is now needed -
616 PCRE is a system library on the majority of modern systems.
617 See entry on PCRE_LIBS in EDITME file.
619 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
620 conversation. Added nologin parameter to request.
621 Patch contributed by Kirill Miazine.
623 TF/01 Do not log submission mode rewrites if they do not change the address.
625 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
627 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
628 log files in place. Contributed by Roberto Lima.
630 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
632 TF/03 Bugzilla 615: When checking the local_parts router precondition
633 after a local_part_suffix or local_part_prefix option, Exim now
634 does not use the address's named list lookup cache, since this
635 contains cached lookups for the whole local part.
637 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
638 Robert Millan. Documentation is in experimental-spec.txt.
640 TF/04 Bugzilla 668: Fix parallel build (make -j).
642 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
644 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
645 Patch provided by Jan Srzednicki.
647 TF/05 Leading white space used to be stripped from $spam_report which
648 wrecked the formatting. Now it is preserved.
650 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
651 that they are available at delivery time.
653 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
655 TF/08 TLS error reporting now respects the incoming_interface and
656 incoming_port log selectors.
658 TF/09 Produce a more useful error message if an SMTP transport's hosts
659 setting expands to an empty string.
661 NM/06 Bugzilla 744: EXPN did not work under TLS.
662 Patch provided by Phil Pennock.
664 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
665 Patch provided by Richard Godbee.
667 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
668 acl_smtp_notquit, added index entry.
670 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
671 Patch provided by Eugene Bujak.
673 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
674 accept(). Patch provided by Maxim Dounin.
676 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
677 Patch provided by Phil Pennock.
679 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
681 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
682 Patch provided by Brad "anomie" Jorsch.
684 NM/14 Bugzilla 622: Added timeout setting to transport filter.
685 Patch provided by Dean Brooks.
687 TK/05 Add native DKIM support (does not depend on external libraries).
689 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
690 Patch provided by Graeme Fowler.
692 NM/16 Bugzilla 851: Documentation example syntax fix.
694 NM/17 Changed NOTICE file to remove references to embedded PCRE.
696 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
699 NM/19 Bugzilla 745: TLS version reporting.
700 Patch provided by Phil Pennock.
702 NM/20 Bugzilla 167: bool: condition support.
703 Patch provided by Phil Pennock.
705 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
706 clients. Patch provided by Phil Pennock.
708 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
709 Patch provided by Brad "anomie" Jorsch.
711 NM/23 Bugzilla 687: Fix misparses in eximstats.
712 Patch provided by Heiko Schlittermann.
714 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
715 Patch provided by Heiko Schlittermann.
717 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
718 plus update to original patch.
720 NM/26 Bugzilla 799: Documentation correction for ratelimit.
722 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
723 Patch provided by David Brownlee.
725 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
727 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
729 NM/30 Bugzilla 888: TLS documentation bugfixes.
731 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
733 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
734 Unlike the original bugzilla I have changed all shell scripts in src tree.
736 NM/33 Bugzilla 898: Transport filter timeout fix.
737 Patch by Todd Rinaldo.
739 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
740 Patch by Serge Demonchaux.
742 NM/35 Bugzilla 39: Base64 decode bug fixes.
743 Patch by Jakob Hirsch.
745 NM/36 Bugzilla 909: Correct connect() call in dcc code.
747 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
749 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
751 NM/39 Bugzilla 911: Fixed MakeLinks build script.
757 TK/01 Add preliminary DKIM support. Currently requires a forked version of
758 ALT-N's libdkim that I have put here:
759 http://duncanthrax.net/exim-experimental/
761 Note to Michael Haardt: I had to rename some vars in sieve.c. They
762 were called 'true' and it seems that C99 defines that as a reserved
763 keyword to be used with 'bool' variable types. That means you could
764 not include C99-style headers which use bools without triggering
765 build errors in sieve.c.
767 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
768 as mailq or other aliases. Changed the --help handling significantly
769 to do whats expected. exim_usage() emits usage/help information.
771 SC/01 Added the -bylocaldomain option to eximstats.
773 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
775 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
777 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
783 PH/01 Another patch from the Sieve maintainer.
785 PH/02 When an IPv6 address is converted to a string for single-key lookup
786 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
787 dots are used instead of colons so that keys in lsearch files need not
788 contain colons. This was done some time before quoting was made available
789 in lsearch files. However, iplsearch files do require colons in IPv6 keys
790 (notated using the quote facility) so as to distinguish them from IPv4
791 keys. This meant that lookups for IP addresses in host lists did not work
792 for iplsearch lookups.
794 This has been fixed by arranging for IPv6 addresses to be expressed with
795 colons if the lookup type is iplsearch. This is not incompatible, because
796 previously such lookups could never work.
798 The situation is now rather anomolous, since one *can* have colons in
799 ordinary lsearch keys. However, making the change in all cases is
800 incompatible and would probably break a number of configurations.
802 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
805 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
806 conversion specification without a maximum field width, thereby enabling
807 a rogue spamd server to cause a buffer overflow. While nobody in their
808 right mind would setup Exim to query an untrusted spamd server, an
809 attacker that gains access to a server running spamd could potentially
810 exploit this vulnerability to run arbitrary code as the Exim user.
812 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
813 $primary_hostname instead of what libspf2 thinks the hosts name is.
815 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
816 a directory entry by the name of the lookup key. Previously, if a
817 symlink pointed to a non-existing file or a file in a directory that
818 Exim lacked permissions to read, a lookup for a key matching that
819 symlink would fail. Now it is enough that a matching directory entry
820 exists, symlink or not. (Bugzilla 503.)
822 PH/03 The body_linecount and body_zerocount variables are now exported in the
825 PH/04 Added the $dnslist_matched variable.
827 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
828 This means they are set thereafter only if the connection becomes
831 PH/06 Added the client_condition to authenticators so that some can be skipped
832 by clients under certain conditions.
834 PH/07 The error message for a badly-placed control=no_multiline_responses left
835 "_responses" off the end of the name.
837 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
839 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
840 (without spaces) instead of just copying the configuration text.
842 PH/10 Added the /noupdate option to the ratelimit ACL condition.
844 PH/11 Added $max_received_linelength.
846 PH/12 Added +ignore_defer and +include_defer to host lists.
848 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
849 way in which PCRE > 7.0 is built.
851 PH/14 Implemented queue_only_load_latch.
853 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
854 MAIL command. The effect was to mangle the value on 64-bit systems.
856 PH/16 Another patch from the Sieve maintainer.
858 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
860 PH/18 If a system quota error occurred while trying to create the file for
861 a maildir delivery, the message "Mailbox is full" was not appended to the
862 bounce if the delivery eventually timed out. Change 4.67/27 below applied
863 only to a quota excession during the actual writing of the file.
865 PH/19 It seems that peer DN values may contain newlines (and other non-printing
866 characters?) which causes problems in log lines. The DN values are now
867 passed through string_printing() before being added to log lines.
869 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
870 and InterBase are left for another time.)
872 PH/21 Added message_body_newlines option.
874 PH/22 Guard against possible overflow in moan_check_errorcopy().
876 PH/23 POSIX allows open() to be a macro; guard against that.
878 PH/24 If the recipient of an error message contained an @ in the local part
879 (suitably quoted, of course), incorrect values were put in $domain and
880 $local_part during the evaluation of errors_copy.
886 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
887 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
890 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
891 issue a MAIL command.
893 PH/02 In an ACL statement such as
895 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
897 if a client was not listed at all, or was listed with a value other than
898 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
899 the condition was not true (as it should be), so access was not denied.
900 The bug was that the ! inversion was incorrectly passed on to the second
901 item. This has been fixed.
903 PH/03 Added additional dnslists conditions == and =& which are different from
904 = and & when the dns lookup returns more than one IP address.
906 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
907 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
909 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
910 FSYNC, which compiles an option called disable_fsync that allows for
911 bypassing fsync(). The documentation is heavily laced with warnings.
913 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
915 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
916 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
917 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
918 including adding "make clean"; (3) Added -fPIC when compiling the test
919 dynamically loaded module, to get rid of a warning.
921 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
922 message fails, move_frozen_messages = true and ignore_bounce_errors_after
923 = 0s. The bug is otherwise harmless.
925 PH/07 There was a bug in the dovecot authenticator such that the value of
926 $auth1 could be overwritten, and so not correctly preserved, after a
927 successful authentication. This usually meant that the value preserved by
928 the server_setid option was incorrect.
930 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
932 PH/09 Installed PCRE release 7.0.
934 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
935 run for batched SMTP input. It is now run at the start of every message
936 in the batch. While fixing this I discovered that the process information
937 (output by running exiwhat) was not always getting set for -bs and -bS
938 input. This is fixed, and it now also says "batched" for BSMTP.
940 PH/11 Added control=no_pipelining.
942 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
943 patch, slightly modified), and move the expansion of helo_data till after
944 the connection is made in the smtp transport (so it can use these
947 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
949 PH/14 Added log_selector = +pid.
951 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
953 PH/16 Add ${if forany and ${if forall.
955 PH/17 Added dsn_from option to vary the From: line in DSNs.
957 PH/18 Flush SMTP output before performing a callout, unless control =
958 no_callout_flush is set.
960 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
961 was true (the default) a successful delivery failed to delete the retry
962 item, thus causing premature timeout of the address. The bug is now
965 PH/20 Added hosts_avoid_pipelining to the smtp transport.
967 PH/21 Long custom messages for fakedefer and fakereject are now split up
968 into multiline reponses in the same way that messages for "deny" and
969 other ACL rejections are.
971 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
972 with slight modification.
974 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
975 draft, changing the syntax and factoring some duplicate code".
977 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
978 for deliveries of the second and subsequent messages over the same SMTP
981 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
982 ${reduce, with only minor "tidies".
984 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
986 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
987 expansion side effects.
989 PH/27 When a message times out after an over-quota error from an Exim-imposed
990 quota, the bounce message says "mailbox is full". This message was not
991 being given when it was a system quota that was exceeded. It now should
994 MH/03 Made $recipients available in local_scan(). local_scan() already has
995 better access to the recipient list through recipients_list[], but
996 $recipients can be useful in postmaster-provided expansion strings.
998 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
999 in the case of a MAIL command with additional options following the
1000 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1001 were accidentally chopped off.
1003 PH/29 SMTP synchronization checks are implemented when a command is read -
1004 there is a check that no more input is waiting when there shouldn't be
1005 any. However, for some commands, a delay in an ACL can mean that it is
1006 some time before the response is written. In this time, more input might
1007 arrive, invalidly. So now there are extra checks after an ACL has run for
1008 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1009 pipelining has not been advertised.
1011 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1013 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1014 returned in the SMTP response when smtp_return_error_details was false.
1015 This has been fixed.
1017 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1018 instead of the C I/O that was originally supplied, because problems were
1019 reported on Solaris.
1021 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1022 Exim which did not show up earlier: it was assuming that a call to
1023 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1024 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1025 was a macro that became an assignment, so it seemed to work. This has
1026 changed to a proper function call with a void return, hence the compile
1027 error. Exim's code has been fixed.
1029 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1032 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1034 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1036 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1037 because it turns out to be too big or there is a timeout) there is no
1038 "Completed" line in the log. When some messages of this type were
1039 selected by exigrep, they were listed as "not completed". Others were
1040 picked up by some special patterns. I have improved the selection
1041 criteria to be more general.
1043 PH/38 The host_find_failed option in the manualroute router can now be set
1044 to "ignore", to completely ignore a host whose IP address cannot be
1045 found. If all hosts are ignored, the behaviour is controlled by the new
1046 host_all_ignored option.
1048 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1049 homing or because of multiple MX records with /mx) generated more than
1050 one IP address, and the following item turned out to be the local host,
1051 all the secondary addresses of the first item were incorrectly removed
1052 from the list, along with the local host and any following hosts (which
1053 is what is supposed to happen).
1055 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1056 whoever called Exim into the -H file. In the case of the daemon it was
1057 behaving confusingly. When first started, it used values for whoever
1058 started the daemon, but after a SIGHUP it used the Exim user (because it
1059 calls itself on a restart). I have changed the code so that it now always
1062 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1063 message are rejected with the same error (e.g. no authentication or bad
1064 sender address), and a DATA command is nevertheless sent (as can happen
1065 with PIPELINING or a stupid MUA), the error message that was given to the
1066 RCPT commands is included in the rejection of the DATA command. This is
1067 intended to be helpful for MUAs that show only the final error to their
1070 PH/42 Another patch from the Sieve maintainer.
1072 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1073 Eximstats - Fixed some broken HTML links and added missing column headers
1075 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1076 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1078 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1080 PH/43 Yet another patch from the Sieve maintainer.
1082 PH/44 I found a way to check for a TCP/IP connection going away before sending
1083 the response to the final '.' that terminates a message, but only in the
1084 case where the client has not sent further data following the '.'
1085 (unfortunately, this is allowed). However, in many cases there won't be
1086 any further data because there won't be any more messages to send. A call
1087 to select() can be used: if it shows that the input is "ready", there is
1088 either input waiting, or the socket has been closed. An attempt to read
1089 the next input character can distinguish the two cases. Previously, Exim
1090 would have sent an OK response which the client would never have see.
1091 This could lead to message repetition. This fix should cure that, at
1092 least in a lot of common cases.
1094 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1095 advertised in response to EHLO.
1101 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1102 fixed by 4.65/MH/01 (is this a record?) are fixed:
1104 (i) An empty string was always treated as zero by the numeric comparison
1105 operators. This behaviour has been restored.
1107 (ii) It is documented that the numeric comparison operators always treat
1108 their arguments as decimal numbers. This was broken in that numbers
1109 starting with 0 were being interpreted as octal.
1111 While fixing these problems I realized that there was another issue that
1112 hadn't been noticed. Values of message_size_limit (both the global option
1113 and the transport option) were treated as octal if they started with 0.
1114 The documentation was vague. These values are now always treated as
1115 decimal, and I will make that clear in the documentation.
1121 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1122 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1125 MH/01 Don't check that the operands of numeric comparison operators are
1126 integers when their expansion is in "skipping" mode (fixes bug
1127 introduced by 4.64-PH/07).
1129 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1130 child addresses, Exim now panics and dies. Previously, because the count
1131 is held in a short int, deliveries were likely to be lost. As such a
1132 large number of recipients for a single message is ridiculous
1133 (performance will be very, very poor), I have chosen to impose a limit
1134 rather than extend the field.
1140 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1141 leftover -K file (the existence of which was triggered by #402).
1142 While we were at it, introduced process PID as part of the -K
1143 filename. This should rule out race conditions when creating
1146 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1147 processing considerably. Previous code took too long for large mails,
1148 triggering a timeout which in turn triggers #401.
1150 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1151 in the DK code in transports.c. sendfile() is not really portable,
1152 hence the _LINUX specificness.
1154 TF/01 In the add_headers option to the mail command in an Exim filter,
1155 there was a bug that Exim would claim a syntax error in any
1156 header after the first one which had an odd number of characters
1159 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1160 callout verification, Exim cached a "reject" for the entire domain. This
1161 is correct for most verifications, but it is not correct for a recipient
1162 verification with use_sender or use_postmaster set, because in that case
1163 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1164 case of MAIL FROM:<> rejection from other early rejections (e.g.
1165 rejection of HELO). When verifying a recipient using a non-null MAIL
1166 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1167 Whatever the result of the callout, the value of the domain cache is
1168 left unchanged (for any other kind of callout, getting as far as trying
1169 RCPT means that the domain itself is ok).
1171 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1174 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1175 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1176 session, the daemon ignores ECONNECT errors and logs others; it now
1177 ignores EPIPE as well.
1179 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1180 (quoted-printable decoding).
1182 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1183 later the small subsequent patch to fix an introduced bug.
1185 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1187 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1189 PH/08 An error is now given if message_size_limit is specified negative.
1191 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1192 to be given (somewhat) arbitrary names.
1194 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1197 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1198 miscellaneous code fixes
1200 PH/10 Added the log_reject_target ACL modifier to specify where to log
1203 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1204 hostname. This is wrong, because it relates to the incoming message (and
1205 probably the interface on which it is arriving) and not to the outgoing
1206 callout (which could be using a different interface). This has been
1207 changed to use the value of the helo_data option from the smtp transport
1208 instead - this is what is used when a message is actually being sent. If
1209 there is no remote transport (possible with a router that sets up host
1210 addresses), $smtp_active_hostname is used.
1212 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1213 tweaks were necessary in order to get it to work (see also 21 below):
1214 (a) The code assumed that strncpy() returns a negative number on buffer
1215 overflow, which isn't the case. Replaced with Exim's string_format()
1217 (b) There were several signed/unsigned issues. I just did the minimum
1218 hacking in of casts. There is scope for a larger refactoring.
1219 (c) The code used strcasecmp() which is not a standard C function.
1220 Replaced with Exim's strcmpic() function.
1221 (d) The code set only $1; it now sets $auth1 as well.
1222 (e) A simple test gave the error "authentication client didn't specify
1223 service in request". It would seem that Dovecot has changed its
1224 interface. Fortunately there's a specification; I followed it and
1225 changed what the client sends and it appears to be working now.
1227 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1230 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1231 address A is aliased to B and C, where B exists and C does not. Without
1232 -v the output is "A verified" because verification stops after a
1233 successful redirection if more than one address is generated. However,
1234 with -v the child addresses are also verified. Exim was outputting "A
1235 failed to verify" and then showing the successful verification for C,
1236 with its parentage. It now outputs "B failed to verify", showing B's
1237 parentage before showing the successful verification of C.
1239 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1240 look up a TXT record in a specific list after matching in a combined
1243 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1244 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1245 they consult the DNS. I had assumed they would set it the way they
1246 wanted; and indeed my experiments on Linux seem to show that in some
1247 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1248 To be on the safe side, however, I have now made the interface to
1249 host_find_byname() similar to host_find_bydns(), with an argument
1250 containing the DNS resolver options. The host_find_byname() function now
1251 sets these options at its start, just as host_find_bydns() does. The smtp
1252 transport options dns_qualify_single and dns_search_parents are passed to
1253 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1254 of host_find_byname() use the default settings of RES_DEFNAMES
1255 (qualify_single) but not RES_DNSRCH (search_parents).
1257 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1258 spool_read_header() do less string testing, by means of a preliminary
1259 switch on the second character of optional "-foo" lines. (This is
1260 overdue, caused by the large number of possibilities that now exist.
1261 Originally there were few.) While I was there, I also converted the
1262 str(n)cmp tests so they don't re-test the leading "-" and the first
1263 character, in the hope this might squeeze out yet more improvement.
1265 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1266 flag allowing group syntax was set by the header_syntax check but not
1267 turned off, possible causing trouble later; (2) The flag was not being
1268 set at all for the header_verify test, causing "group"-style headers to
1269 be rejected. I have now set it in this case, and also caused header_
1270 verify to ignore an empty address taken from a group. While doing this, I
1271 came across some other cases where the code for allowing group syntax
1272 while scanning a header line wasn't quite right (mostly, not resetting
1273 the flag correctly in the right place). These bugs could have caused
1274 trouble for malformed header lines. I hope it is now all correct.
1276 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1277 with the "reply" argument non-NULL. The code, however (which originally
1278 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1279 but it didn't always do it. This confused somebody who was copying the
1280 code for some other use. I have removed all the tests.
1282 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1283 feature that was used to support insecure browsers during the U.S. crypto
1284 embargo. It requires special client support, and Exim is probably the
1285 only MTA that supported it -- and would never use it because real RSA is
1286 always available. This code has been removed, because it had the bad
1287 effect of slowing Exim down by computing (never used) parameters for the
1288 RSA_EXPORT functionality.
1290 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1291 authenticator to fail if there's a tab character in the incoming data
1292 (there should never be unless someone is messing about, as it's supposed
1293 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1294 if the connection is using TLS or if the remote IP is the same as the
1295 local IP, and the "valid-client-cert option" if a client certificate has
1298 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1299 authenticators. This can be used for authorization after authentication
1300 succeeds. (In the case of plaintext, it servers for both authentication
1303 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1304 if any retry times were supplied.
1306 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1307 connection, where there is no client IP address to check. In this
1308 situation, the verify now always succeeds.
1310 PH/25 Applied John Jetmore's -Mset patch.
1312 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1314 PH/27 In a string expansion for a processed (not raw) header when multiple
1315 headers of the same name were present, leading whitespace was being
1316 removed from all of them, but trailing whitespace was being removed only
1317 from the last one. Now trailing whitespace is removed from each header
1318 before concatenation. Completely empty headers in a concatenation (as
1319 before) are ignored.
1321 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1322 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1324 PH/29 [Removed. This was a change that I later backed out, and forgot to
1325 correct the ChangeLog entry (that I had efficiently created) before
1326 committing the later change.]
1328 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1329 address errors (4xx response to RCPT) over the same connection as other
1330 messages routed to the same hosts. Such deliveries are always "forced",
1331 so retry times are not inspected. This resulted in far too many retries
1332 for the affected addresses. The effect occurred only when there were more
1333 hosts than the hosts_max_try setting in the smtp transport when it had
1334 the 4xx errors. Those hosts that it had tried were not added to the list
1335 of hosts for which the message was waiting, so if all were tried, there
1336 was no problem. Two fixes have been applied:
1338 (i) If there are any address or message errors in an SMTP delivery, none
1339 of the hosts (tried or untried) are now added to the list of hosts
1340 for which the message is waiting, so the message should not be a
1341 candidate for sending over the same connection that was used for a
1342 successful delivery of some other message. This seems entirely
1343 reasonable: after all the message is NOT "waiting for some host".
1344 This is so "obvious" that I'm not sure why it wasn't done
1345 previously. Hope I haven't missed anything, but it can't do any
1346 harm, as the worst effect is to miss an optimization.
1348 (ii) If, despite (i), such a delivery is accidentally attempted, the
1349 routing retry time is respected, so at least it doesn't keep
1350 hammering the server.
1352 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1353 in ${readsocket because some servers need this prod.
1355 PH/32 Added some extra debug output when updating a wait-xxx database.
1357 PH/33 The hint "could be header name not terminated by colon", which has been
1358 given for certain expansion errors for a long time, was not being given
1359 for the ${if def:h_colon_omitted{... case.
1361 PH/34 The spec says: "With one important exception, whenever a domain list is
1362 being scanned, $domain contains the subject domain." There was at least
1363 one case where this was not true.
1365 PH/35 The error "getsockname() failed: connection reset by peer" was being
1366 written to the panic log as well as the main log, but it isn't really
1367 panic-worthy as it just means the connection died rather early on. I have
1368 removed the panic log writing for the ECONNRESET error when getsockname()
1371 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1372 runs only) independently of the message's sender address. This meant
1373 that, if the 4xx error was in fact related to the sender, a different
1374 message to the same recipient with a different sender could confuse
1375 things. In particualar, this can happen when sending to a greylisting
1376 server, but other circumstances could also provoke similar problems.
1377 I have changed the default so that the retry time for these errors is now
1378 based a combination of the sender and recipient addresses. This change
1379 can be overridden by setting address_retry_include_sender=false in the
1382 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1383 remote server are returned as part of bounce messages. This was not
1384 happening for LMTP over a pipe (the lmtp transport), but now it is the
1385 same for both kinds of LMTP.
1387 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1388 in header lines that were in fact CNAMEs. This is no longer the case.
1390 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1391 and queue runs started by the daemon processed all messages. This has
1392 been fixed so that -R and -S can now usefully be given with -q<time>.
1394 PH/40 Import PCRE release 6.7 (fixes some bugs).
1396 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1398 PH/42 Give an error if -q is specified more than once.
1400 PH/43 Renamed the variables $interface_address and $interface_port as
1401 $received_ip_address and $received_port, to make it clear that these
1402 values apply to message reception, and not to the outgoing interface when
1403 a message is delivered. (The old names remain recognized, of course.)
1405 PH/44 There was no timeout on the connect() call when using a Unix domain
1406 socket in the ${readsocket expansion. There now is.
1408 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1409 be meaningful with "accept".
1411 SC/01 Eximstats V1.43
1412 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1414 SC/02 Eximstats V1.44
1415 Use a glob alias rather than an array ref in the generated
1416 parser. This improves both readability and performance.
1418 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1419 Collect SpamAssassin and rejection statistics.
1420 Don't display local sender or destination tables unless
1421 there is data to show.
1422 Added average volumes into the top table text output.
1424 SC/04 Eximstats V1.46
1425 Collect data on the number of addresses (recipients)
1426 as well as the number of messages.
1428 SC/05 Eximstats V1.47
1429 Added 'Message too big' to the list of mail rejection
1430 reasons (thanks to Marco Gaiarin).
1432 SC/06 Eximstats V1.48
1433 Mainlog lines which have GMT offsets and are too short to
1434 have a flag are now skipped.
1436 SC/07 Eximstats V1.49 (Alain Williams)
1437 Added the -emptyok flag.
1439 SC/08 Eximstats V1.50
1440 Fixes for obtaining the IP address from reject messages.
1442 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1443 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1444 whitesspace changes from 4.64-PH/27
1446 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1449 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1450 are found, allow negative numbers in numeric criteria)
1452 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1454 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1455 to match changes made in 4.64-PH/43
1457 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1459 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1460 "domain" is actually an IP address, and force a failure. This locks out
1461 those revolvers/nameservers that support "A-for-A" lookups, in
1462 contravention of the specifications.
1464 PH/48 When a host name was looked up from an IP address, and the subsequent
1465 forward lookup of the name timed out, the host name was left in
1466 $sender_host_name, contrary to the specification.
1468 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1469 restricted to single-key lookups, Exim was not diagnosing an error if
1470 * or *@ was used with a query-style lookup.
1472 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1474 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1475 long ago, but noone interested enough thought of it. Let's just say that
1476 the "1.1" means that there are some new functions that weren't there at
1477 some point in the past.
1479 PH/51 Error processing for expansion failure of helo_data from an smtp
1480 transport during callout processing was broken.
1482 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1483 tested/used via the -bh/-bhc/-bs options.
1485 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1486 bug, fixed in subsequent PCRE releases).
1488 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1489 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1491 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1497 SC/01 Use a glob alias rather than an array ref in eximstats generated
1498 parser. This improves both readability and performance.
1500 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1501 Don't display local sender or destination tables in eximstats unless
1502 there is data to show.
1503 Added average volumes into the eximstats top table text output.
1505 SC/03 Collect data on the number of addresses (recipients) as well
1506 as the number of messages in eximstats.
1508 TF/01 Correct an error in the documentation for the redirect router. Exim
1509 does (usually) call initgroups() when daemonizing.
1511 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1512 with consistent privilege compared to when running as a daemon.
1514 TF/03 Note in the spec that $authenticated_id is not set for local
1515 submissions from trusted users.
1517 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1518 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1520 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1521 by adding some example configuration directives to the default
1522 configuration file. A little bit of work is required to uncomment the
1523 directives and define how usernames and passwords are checked, but
1524 there is now a framework to start from.
1526 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1527 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1528 without this. I don't know how relevant this is to other LDAP libraries.
1530 PH/02 Add the verb name to the "unknown ACL verb" error.
1532 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1534 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1536 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1537 directory not expanded when it should be if an expanded home directory
1538 was set for the address (which is overridden by the transport).
1540 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1543 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1544 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1545 because it is too late at that time, and has no effect.
1547 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1548 security issue with \' (bugzilla #107). I could not use the
1549 PQescapeStringConn() function, because it needs a PGconn value as one of
1552 PH/08 When testing addresses using -bt, indicate those final addresses that
1553 are duplicates that would not cause an additional delivery. At least one
1554 person was confused, thinking that -bt output corresponded to deliveries.
1555 (Suppressing duplicates isn't a good idea as you lose the information
1556 about possibly different redirections that led to the duplicates.)
1558 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1559 systems where poll() doesn't work, in particular OS X.
1561 PH/10 Added more information to debugging output for retry time not reached.
1563 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1564 operations in malware.c.
1566 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1569 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1570 syslog_duplication set false, log lines that would normally be written
1571 both the the main log and to the reject log were not written to syslog at
1574 PH/14 In the default configuration, change the use of "message" in ACL warn
1575 statements to "add_header".
1577 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1578 not followed by a command (e.g. "seen endif").
1580 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1581 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1584 PH/17 Added extra conditions to the default value of delay_warning_condition
1588 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1589 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1590 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1593 The Auto-Submitted: and various List- headers are standardised, whereas I
1594 don't think Precedence: ever was.
1596 PH/18 Refactored debugging code in route_finduser() to show more information,
1597 in particular, the error code if getpwnam() issues one.
1599 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1600 This is apparently needed in addition to the PH/07 change above to avoid
1601 any possible encoding problems.
1603 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1604 but not after initializing Perl.
1606 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1607 output them only if debugging. By default they are written stderr,
1608 apparently, which is not desirable.
1610 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1613 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1616 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1618 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1619 authenticated or an ident call has been made. Suppress the default
1620 values for $authenticated_id and $authenticated_sender (but permit -oMai
1621 and -oMas) when testing with -bh.
1623 PH/24 Re-jigged the order of the tests in the default configuration so that the
1624 tests for valid domains and recipients precede the DNS black list and CSA
1625 tests, on the grounds that those ones are more expensive.
1627 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1628 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1629 This bug exists in every version of Exim that I still have, right back to
1632 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1633 However, an attempt to turn on case-sensitivity in a regex key by
1634 including (?-i) didn't work because the subject string was already
1635 lowercased, and the effects were non-intuitive. It turns out that a
1636 one-line patch can be used to allow (?-i) to work as expected.
1642 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1643 other effects) broke the use of negated acl sub-conditions.
1645 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1648 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1649 "Deny" causes Exim to reject the incoming connection with a 554 error.
1650 Unfortunately, if there is a major crisis, such as a disk failure,
1651 tcp-wrappers gives "deny", whereas what one would like would be some
1652 kind of temporary error. A kludge has been added to help with this.
1653 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1654 554 error is used if errno is still zero or contains ENOENT (which occurs
1655 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1658 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1660 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1661 errors. Otherwise a message that provokes a temporary error (when other
1662 messages do not) can cause a whole host to time out.
1664 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1665 addresses were routed directly to files or pipes from a redirect router.
1666 File deliveries just didn't batch; pipe deliveries might have suffered
1669 PH/06 A failure to get a lock for a hints database would erroneously always say
1670 "Failed to get write lock", even when it was really a read lock.
1672 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1673 of 0600. This has been changed to use the value of the lockfile_mode
1674 option (which defaults to 0600).
1676 PH/08 Applied small patch from the Sieve maintainer.
1678 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1679 folder from quota calculations, a direct delivery into this folder messed
1680 up the contents of the maildirsize file. This was because the regex was
1681 used only to exclude .Trash (or whatever) when the size of the mailbox
1682 was calculated. There was no check that a delivery was happening into an
1683 excluded directory. This bug has been fixed by ignoring all quota
1684 processing for deliveries into excluded directories.
1686 PH/10 Added the maildirfolder_create_regex option to appendfile.
1692 PH/01 The code for finding all the local interface addresses on a FreeBSD
1693 system running IPv6 was broken. This may well have applied to all BSD
1694 systems, as well as to others that have similar system calls. The broken
1695 code found IPv4 interfaces correctly, but gave incorrect values for the
1696 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1697 that it would not match correctly against @[] and not recognize the IPv6
1700 PH/02 The ipliteral router was not recognizing addresses of the form user@
1701 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1703 PH/03 Added disable_ipv6.
1705 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1706 decoded form, because it is most often used to construct To: headers
1707 lines in autoreplies, and the decoded form may well be syntactically
1708 invalid. However, $reply_address has leading white space removed, and all
1709 newlines turned into spaces so that the autoreply transport does not
1712 PH/05 If group was specified without a user on a router, and no group or user
1713 was specified on a transport, the group from the router was ignored.
1715 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1716 for visible compile-time settings that can be used to change these
1717 numbers, for those that want even more. Backwards compatibility with old
1718 spool files has been maintained. However, going back to a previous Exim
1719 release will lost any variables that are in spool files.
1721 PH/07 Two small changes when running in the test harness: increase delay when
1722 passing a TCP/IP connection to a new process, in case the original
1723 process has to generate a bounce, and remove special handling of
1724 127.0.0.2 (sic), which is no longer necessary.
1726 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1727 be the same on different OS.
1729 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1732 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1733 whether --show-vars was specified or not
1735 JJ/02 exipick: Added support for new ACL variable spool format introduced
1738 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1739 syntactically invalid From: or Reply-to: line, and a filter used this to
1740 generate an autoreply, and therefore failed to obtain an address for the
1741 autoreply, Exim could try to deliver to a non-existent relative file
1742 name, causing unrelated and misleading errors. What now happens is that
1743 it logs this as a hard delivery error, but does not attempt to create a
1746 PH/11 The exinext utility has a -C option for testing purposes, but although
1747 the given file was scanned by exinext itself; it wasn't being passed on
1748 when Exim was called.
1750 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1751 an end-of-file indication when reading a command response.
1753 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1754 compiled. In many other places in Exim, IPv6 addresses are always
1755 recognized, so I have changed this. It also means that IPv4 domain
1756 literals of the form [IPV4:n.n.n.n] are now always recognized.
1758 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1759 used if the router is not running as root, for example, when verifying at
1760 ACL time, or when using -bh. The debugging output from this situation was
1761 non-existent - all you got was a failure to exec. I have made two
1764 (a) Failures to set uid/gid, the current directory, or a process leader
1765 in a subprocess such as that created by queryprogram now generate
1766 suitable debugging ouput when -d is set.
1768 (b) The queryprogram router detects when it is not running as root,
1769 outputs suitable debugging information if -d is set, and then runs
1770 the subprocess without attempting to change uid/gid.
1772 PH/15 Minor change to Makefile for building test_host (undocumented testing
1775 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1776 additional section of a DNS packet that returns MX or SRV records.
1777 Instead, it always explicitly searches for A/AAAA records. This avoids
1778 major problems that occur when a DNS server includes only records of one
1779 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1780 fixed another bug: if SRV records were looked up and the corresponding
1781 address records were *not* found in the additional section, the port
1782 values from the SRV records were lost.
1784 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1785 using the correct key (the original address) when searching the retry
1786 rules in order to find which one to use for generating the retry hint.
1788 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1789 adding the default one. Similarly, if it contains a Reply-To: header, the
1790 errors_reply_to option, if set, is not used.
1792 PH/19 When calculating a retry time, Exim used to measure the "time since
1793 failure" by looking at the "first failed" field in the retry record. Now
1794 it does not use this if it is later than than the arrival time of the
1795 message. Instead it uses the arrival time. This makes for better
1796 behaviour in cases where some deliveries succeed, thus re-setting the
1797 "first failed" field. An example is a quota failure for a huge message
1798 when small messages continue to be delivered. Without this change, the
1799 "time since failure" will always be short, possible causing more frequent
1800 delivery attempts for the huge message than are intended.
1801 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1803 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1804 $1, $2, $3) because the numerical variables can be reset during some
1805 expansion items (e.g. "match"), thereby losing the authentication data.
1807 PH/21 Make -bV show the size of off_t variables so that the test suite can
1808 decide whether to run tests for quotas > 2G.
1810 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1811 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1812 filecount value is greater than 2G or if a quota value is greater than 2G
1813 on a system where the size of off_t is not greater than 4, a panic error
1816 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1817 never match. The debug and -bh output now contains an explicit error
1818 message indicating a malformed IPv4 address or mask.
1820 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1821 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1822 PH/23 above applies.
1824 PH/25 Do not write to syslog when running in the test harness. The only
1825 occasion when this arises is a failure to open the main or panic logs
1826 (for which there is an explicit test).
1828 PH/26 Added the /no_tell option to "control=freeze".
1830 PH/27 If a host name lookup failed very early in a connection, for example, if
1831 the IP address matched host_lookup and the reverse lookup yielded a name
1832 that did not have a forward lookup, an error message of the form "no IP
1833 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1834 could be logged. Now it outputs the IP address instead of "NULL".
1836 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1837 allows the sender and the authenticated sender to be set when
1838 submitting a message from within Exim. Since child_open_exim() is
1839 documented for local_scan(), the new function should be too.
1841 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1842 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1843 results in an empty string is now treated as unset.
1845 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1847 PH/31 Added sender_verify_fail logging option.
1849 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1850 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1851 tidied the source and removed it altogether.
1853 PH/33 When a queue run was abandoned because the load average was too high, a
1854 log line was always written; now it is written only if the queue_run log
1855 selector is set. In addition, the log line for abandonment now contains
1856 information about the queue run such as the pid. This is always present
1857 in "start" and "stop" lines but was omitted from the "abandon" line.
1859 PH/34 Omit spaces between a header name and the colon in the error message that
1860 is given when verify = headers_syntax fails (if there are lots of them,
1861 the message gets confusing).
1863 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1864 names, as there are now some PTR records that contain slashes. This check
1865 is only to protect against broken name servers that fall over on strange
1866 characters, so the fact that it applies to all lookups doesn't matter.
1868 PH/36 Now that the new test suite is complete, we can remove some of the
1869 special code in Exim that was needed for the old test suite. For example,
1870 sorting DNS records because real resolvers return them in an arbitrary
1871 order. The new test suite's fake resolver always returns records in the
1874 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1875 bounces) except when queue_only is set, to avoid logging races between
1876 the different processes.
1878 PH/38 Panic-die if .include specifies a non-absolute path.
1880 PH/39 A tweak to the "H" retry rule from its user.
1882 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1883 a label. They prevented compilation on older perls.
1885 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1886 a warning to be raised on newish perls.
1888 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1889 on queue. Changes to match documented behaviour of showing count of
1890 messages matching specified criteria.
1892 PH/40 Changed the default ident timeout from 30s to 5s.
1894 PH/41 Added support for the use of login_cap features, on those BSD systems
1895 that have them, for controlling the resources used by pipe deliveries.
1897 PH/42 The content-scanning code uses fopen() to create files in which to put
1898 message data. Previously it was not paying any attention to the mode of
1899 the files. Exim runs with umask(0) because the rest of the code creates
1900 files with open(), and sets the required mode explicitly. Thus, these
1901 files were ending up world-writeable. This was not a big issue, because,
1902 being within the spool directory, they were not world-accessible. I have
1903 created a function called modefopen, which takes an additional mode
1904 argument. It sets umask(777), creates the file, chmods it to the required
1905 mode, then resets the umask. All the relevant calls to fopen() in the
1906 content scanning code have been changed to use this function.
1908 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1909 to 24 hours. This avoids potential overflow problems when processing G
1910 and H retry rules. I suspect nobody ever tinkers with this value.
1912 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1914 PH/45 When the plaintext authenticator is running as a client, the server's
1915 challenges are checked to ensure they are valid base64 strings. By
1916 default, the authentication attempt is cancelled if an invalid string is
1917 received. Setting client_ignore_invalid_base64 true ignores these errors.
1918 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1919 they are received. Thus, the responses can be made to depend on the
1920 challenges. If an invalid string is ignored, an empty string is placed in
1923 PH/46 Messages that are created by the autoreply transport now contains a
1924 References: header, in accordance with RFCs 2822 and 3834.
1926 PH/47 Added authenticated_sender_force to the smtp transport.
1928 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1930 PH/49 Installed latest patch from the Sieve maintainer.
1932 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1933 was also set, the appendfile transport was unnecessarily scanning a
1934 directory of message files (e.g. for maildir delivery) to find the count
1935 of files (along with the size), even though it did not need this
1936 information. It now does the scan only if it needs to find either the
1937 size of the count of files.
1939 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1941 PH/52 Two bugs concerned with error handling when the smtp transport is
1944 (i) Exim was not creating retry information for temporary errors given
1945 for individual recipients after the DATA command when the smtp transport
1946 was used in LMTP mode. This meant that they could be retried too
1947 frequently, and not timed out correctly.
1949 (ii) Exim was setting the flag that allows error details to be returned
1950 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1951 recipients that were returned after the DATA command.
1953 PH/53 This is related to PH/52, but is more general: for any failing address,
1954 when detailed error information was permitted to be returned to the
1955 sender, but the error was temporary, then after the final timeout, only
1956 "retry timeout exceeded" was returned. Now it returns the full error as
1957 well as "retry timeout exceeded".
1959 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1960 do this, and (what is worse) MTAs that accept it.
1962 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1963 will now be deprecated.
1965 PH/56 New os.c-cygwin from the Cygwin maintainer.
1967 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1968 formats (previously only available in exim formats via -bpr, -bpru,
1969 and -bpra. Now also available in native and exiqgrep formats)
1971 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1972 with very large, slow to parse queues
1974 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1976 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1978 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1979 responses to authentication challenges, though it was showing the
1980 challenges; (ii) I've removed the CR characters from the debug output for
1983 PH/58 Allow for the insertion of a newline as well as a space when a string
1984 is turned into more than one encoded-word during RFC 2047 encoding. The
1985 Sieve code now uses this.
1987 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1988 data_4xx, lost_connection, tls_required.
1990 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1991 message was being sent as an SMTP response.
1993 PH/61 Add -l and -k options to exicyclog.
1995 PH/62 When verifying, if an address was redirected to one new address, so that
1996 verification continued, and the new address failed or deferred after
1997 having set something in $address_data, the value of $address_data was not
1998 passed back to the ACL. This was different to the case when no
1999 redirection occurred. The value is now passed back in both cases.
2001 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2002 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2003 use login_cap.h, so on its own it isn't the distinguishing feature. The
2004 new name refers directly to the setclassresources() function.
2006 PH/65 Added configuration files for NetBSD3.
2008 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2010 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2011 is preferred over IPv4.
2013 PH/68 The bounce_return_message and bounce_return_body options were not being
2014 honoured for bounces generated during the reception of non-SMTP messages.
2015 In particular, this applied to messages rejected by the ACL. This bug has
2016 been fixed. However, if bounce_return_message is true and bounce_return_
2017 body is false, the headers that are returned for a non-SMTP message
2018 include only those that have been read before the error was detected.
2019 (In the case of an ACL rejection, they have all been read.)
2021 PH/69 The HTML version of the specification is now built in a directory called
2022 spec_html instead of spec.html, because the latter looks like a path with
2023 a MIME-type, and this confuses some software.
2025 PH/70 Catch two compiler warnings in sieve.c.
2027 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2028 function verify_get_ident() calls ip_connect() to connect a socket, but
2029 if the "connect()" function timed out, ip_connect() used to close the
2030 socket. However, verify_get_ident() also closes the socket later, and in
2031 between Exim writes to the log, which may get opened at this point. When
2032 the socket was closed in ip_connect(), the log could get the same file
2033 descriptor number as the socket. This naturally causes chaos. The fix is
2034 not to close the socket in ip_connect(); the socket should be closed by
2035 the function that creates it. There was only one place in the code where
2036 this was missing, in the iplookup router, which I don't think anybody now
2037 uses, but I've fixed it anyway.
2039 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2040 well as to direct DNS lookups. Otherwise the handling of names in host
2041 lists is inconsistent and therefore confusing.
2047 PH/01 Two changes to the default runtime configuration:
2049 (1) Move the checks for relay_from_hosts and authenticated clients from
2050 after to before the (commented out) DNS black list checks.
2052 (2) Add control=submission to the relay_from_hosts and authenticated
2053 clients checks, on the grounds that messages accepted by these
2054 statements are most likely to be submissions.
2056 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2058 (1) Generate an error if the third argument for the ${prvs expansion is
2061 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2064 (3) Reset the variables that are obtained from the first argument of
2065 ${prvscheck and used in the second argument before leaving the code,
2066 because their memory is reclaimed, so using them afterwards may do
2069 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2070 one (it's much easier than Tom thought :-).
2072 (5) Because of (4), we can now allow for the use of $prvscheck_result
2073 inside the third argument.
2075 PH/03 For some reason, the default setting of PATH when running a command from
2076 a pipe transport was just "/usr/bin". I have changed it to
2079 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2080 anything to be listed in the output from -bV.
2082 PH/05 When a filter generated an autoreply, the entire To: header line was
2083 quoted in the delivery log line, like this:
2085 => >A.N.Other <ano@some.domain> <original@ddress> ...
2087 This has been changed so that it extracts the operative address. There
2088 may be more than one such address. If so, they are comma-separated, like
2091 => >ano@some.domain,ona@other.domain <original@ddress> ...
2093 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2094 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2095 not being looked up in the rDNS to get a host name, Exim was showing the
2096 IP address twice in Received: lines, even though the IP addresses were
2097 identical. For example:
2099 Received: from [1.2.3.4] (helo=[1.2.3.4])
2101 However, if the real host name was known, it was omitting the HELO data
2102 if it matched the actual IP address. This has been tidied up so that it
2103 doesn't show the same IP address twice.
2105 PH/07 When both +timestamp and +memory debugging was on, the value given by
2106 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2107 called by the debug printing, thereby overwriting the timestamp buffer.
2108 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2110 PH/08 When the original message was included in an autoreply transport, it
2111 always said "this is a copy of the message, including all the headers",
2112 even if body_only or headers_only was set. It now gives an appropriate
2115 PH/09 Applied a patch from the Sieve maintainer which:
2117 o fixes some comments
2118 o adds the (disabled) notify extension core
2119 o adds some debug output for the result of if/elsif tests
2120 o points to the current vacation draft in the documentation
2121 and documents the missing references header update
2125 o fixes a bug in processing the envelope test (when testing
2126 multiple envelope elements, the last element determinted the
2129 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2130 Electronic Mail") by including:
2132 Auto-submitted: auto-generated
2134 in the messages that it generates (bounce messages and others, such as
2135 warnings). In the case of bounce messages for non-SMTP mesages, there was
2136 also a typo: it was using "Auto_submitted" (underscore instead of
2137 hyphen). Since every message generated by Exim is necessarily in response
2138 to another message, thes have all been changed to:
2140 Auto-Submitted: auto-replied
2142 in accordance with these statements in the RFC:
2144 The auto-replied keyword:
2146 - SHOULD be used on messages sent in direct response to another
2147 message by an automatic process,
2149 - MUST NOT be used on manually-generated messages,
2151 - MAY be used on Delivery Status Notifications (DSNs) and Message
2152 Disposition Notifications (MDNs),
2154 - MUST NOT be used on messages generated by automatic or periodic
2155 processes, except for messages which are automatic responses to
2158 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2159 to the default Received: header definition.
2161 PH/12 Added log selector acl_warn_skipped (default on).
2163 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2164 variables because (a) they are in the wrong storage pool and (b) even if
2165 they were copied, it wouldn't work properly because of the caching.
2167 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2168 checking when decoding. Apparently there are clients that generate
2169 overlong encoded strings. Why am I not surprised?
2171 PH/15 If the first argument of "${if match_address" was not empty, but did not
2172 contain an "@" character, Exim crashed. Now it writes a panic log message
2173 and treats the condition as false.
2175 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2177 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2178 extension "envelope-auth". The code is finished and in agreement with
2179 other implementations, but there is no documentation so far and in fact,
2180 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2181 not changing the active code.
2183 Print executed "if" and "elsif" statements when debugging is used. This
2184 helps a great deal to understand what a filter does.
2186 Document more things not specified clearly in RFC3028. I had all this
2187 sorted out, when out of a sudden new issues came to my mind. Oops."
2189 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2192 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2193 canonical form (as documented). However, after a host name lookup from
2194 the IP address, check_host() was doing a simple string comparison with
2195 addresses acquired from the DNS when checking that the found name did
2196 have the original IP as one of its addresses. Since any found IPv6
2197 addresses are likely to be in abbreviated form, the comparison could
2198 fail. Luckily, there already exists a function for doing the comparison
2199 by converting both addresses to binary, so now that is used instead of
2200 the text comparison.
2202 PH/20 There was another similar case to PH/19, when a complete host name was
2203 given in a host list; looking up its IP address could give an abbreviated
2204 form, whereas the current host's name might or might not be abbreviated.
2205 The same fix has been applied.
2211 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2212 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2215 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2216 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2218 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2220 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2221 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2222 Also, for "Auto-Submitted", treat anything other than "no" as
2223 non-personal, in accordance with RFC 3834. (Previously it treated
2224 anything starting "auto-" as non-personal.)
2226 TF/01 The control=submission/name=... option had a problem with syntax
2227 errors if the name included a slash character. The /name= option
2228 now slurps the rest of the string, so it can include any characters
2229 but it must come last in the list of options (after /sender_retain
2232 PH/05 Some modifications to the interface to the fake nameserver for the new
2240 TK/01 Added the "success_on_redirect" address verification option. See
2241 NewStuff for rationale and an example.
2243 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2245 PH/02 Patch to exigrep to allow it to work on syslog lines.
2247 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2248 fread() to skip over the body file's header line, because in Cygwin the
2249 header line is locked and is inaccessible.
2251 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2252 co-exist for some time) to make it clear that it is the Exim ID that is
2253 referenced, not the Message-ID: header line.
2255 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2256 string_format() function, because snprintf() does not exist on all
2259 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2260 ${stat: expansion item.
2262 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2263 protocol synchronization error", to keep the pedants happy.
2265 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2266 well as for IRIX systems, when gcc is being used. See the host.c source
2269 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2271 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2274 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2275 transport and to the smtp transport in LMTP mode.
2277 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2279 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2280 run a filter in a subprocess. This could lead to confusion in subsequent
2281 lookups in the parent process. There should also be a search_tidyup() at
2282 the end of the subprocess.
2284 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2285 only if the host matched helo_try_verify_hosts, which caused the
2286 verification to occur when the EHLO/HELO command was issued. The ACL just
2287 tested the remembered result. Now, if a previous verification attempt has
2288 not happened, "verify = helo" does it there and then.
2290 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2292 TK/03 Fix log output including CR from clamd.
2294 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2295 exist provoked a memory error which could cause a segfault.
2297 PH/15 Installed PCRE 6.2
2299 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2301 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2302 of the problem. Specifically, suggested +O2 rather than +O1 for the
2305 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2307 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2308 record, the port was not correctly logged when the outgoing_port log
2309 selector was set (it logged the transort's default port).
2311 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2312 fallback_hosts, and "hosts" in the smtp transport.
2314 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2315 host errors such as "Connection refused".
2317 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2318 authentication with radiusclient 0.4.9:
2320 - Error returned from rc_read_config was caught wrongly
2321 - Username/password not passed on to radius server due to wrong length.
2323 The presumption is that some radiusclient API changes for 4.51/PH/17
2324 were not taken care of correctly. The code is still untested by me (my
2325 Linux distribution still has 0.3.2 of radiusclient), but it was
2326 contributed by a Radius user.
2328 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2329 expanding the "port" option of the smtp transport.
2331 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2332 while reading a MIME header. Thanks to Tom Hughes for a patch.
2334 PH/24 Include config.h inside local_scan.h so that configuration settings are
2337 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2338 that in an ACL for RCPT (for example), you can examine exactly what was
2341 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2342 commands, but it was not correctly comparing the address with the actual
2343 client host address. Thus, it would show the EHLO address in Received:
2344 header lines when this was not necessary.
2346 PH/27 Added the % operator to ${eval:}.
2348 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2349 it should be ignoring failures (because with -C, for example, it has lost
2350 privilege). It wasn't ignoring creation failures other than "already
2353 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2356 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2357 because an input line was too long, either on its own, or by virtue of
2358 too many continuations, the temporary file was not being removed, and the
2359 return code was incorrect.
2361 PH/31 Missing "BOOL" in function definition in filtertest.c.
2363 PH/32 Applied Sieve patches from the maintainer.
2365 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2367 PH/33 Added "verify = not_blind".
2369 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2370 Local/Makefile (with some defaults set). These are used in built scripts
2371 such as exicyclog, but they have never been used in the exim_install
2372 script (though there are many overriding facilities there). I have
2373 arranged that the exim_install script now takes note of these two
2376 PH/35 Installed configuration files for Dragonfly.
2378 PH/36 When a locally submitted message by a trusted user did not contain a
2379 From: header, and the sender address was obtained from -f or from an SMTP
2380 MAIL command, and the trusted user did not use -F to supply a sender
2381 name, $originator_name was incorrectly used when constructing a From:
2382 header. Furthermore, $originator_name was used for submission mode
2383 messages from external hosts without From: headers in a similar way,
2384 which is clearly wrong.
2386 PH/37 Added control=suppress_local_fixups.
2388 PH/38 When log_selector = +received_sender was set, and the addition of the
2389 sender made the log line's construction buffer exactly full, or one byte
2390 less than full, an overflow happened when the terminating "\n" was
2393 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2394 when the result of a list match is failure because a DNS lookup failed.
2396 PH/40 RM_COMMAND is now used in the building process.
2398 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2399 the "build-* directories that it finds.
2401 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2402 address in a domain literal was a prefix of an interface address.
2404 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2405 when verifying a sender address, unless rewrite_headers is false.
2407 PH/44 Wrote a long comment about why errors_to addresses are verified as
2408 recipients, not senders.
2410 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2411 the ratelimit ACL was added.
2413 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2415 PH/46 Added extra information about PostgreSQL errors to the error string.
2417 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2418 avoiding the need to install special zones in a real server. This is
2419 backwards compatible; if it can't find the fake resolver, it drops back.
2420 Thus, both old and new test suites can be run.
2422 TF/02 Added util/ratelimit.pl
2424 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2425 clock is set back in time.
2427 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2428 Candler <B.Candler@pobox.com>.
2430 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2431 OK for addresses that are the result of redirections.
2433 PH/48 A number of further additions for the benefit of the new test suite,
2434 including a fake gethostbyname() that interfaces to the fake DNS resolver
2437 TF/06 The fix for widen_domains has also been applied to qualify_single and
2438 search_parents which are the other dnslookup options that can cause
2441 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2444 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2446 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2447 service" if it encountered an MX record with an empty target hostname.
2448 The message is now "an MX or SRV record indicated no SMTP service".
2450 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2451 if the DNS of the sending site is misconfigured. This is quite a
2452 common situation. This change restores the behaviour of treating a
2453 helo verification defer as a failure.
2455 PH/51 If self=fail was set on a router, the bounce message did not include the
2456 actual error message.
2462 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2464 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2465 command itself ended in error, the underlying message about the transport
2466 filter timeout was being overwritten with the pipe command error. Now the
2467 underlying error message should be appended to the second error message.
2469 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2471 PH/02 Reduce the amount of output that "make" produces by default. Full output
2472 can still be requested.
2474 PH/03 The warning log line about a condition test deferring for a "warn" verb
2475 was being output only once per connection, rather than after each
2476 occurrence (because it was using the same function as for successful
2477 "warn" verbs). This seems wrong, so I have changed it.
2479 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2480 it should not have, which might have caused a crash in the right
2481 circumstances, but probably never did.
2483 PH/04 Installed a modified version of Tony Finch's patch to make submission
2484 mode fix the return path as well as the Sender: header line, and to
2485 add a /name= option so that you can make the user's friendly name appear
2488 TF/03 Added the control = fakedefer ACL modifier.
2490 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2491 Mark Lowes for thorough testing.
2493 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2495 TK/03 Merged latest SRS patch from Miles Wilton.
2497 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2498 with the definition in sysexits.h (which is #included earlier).
2499 Fortunately, Exim does not actually use EX_OK. The code used to try to
2500 preserve the sysexits.h value, by assumimg that macro definitions were
2501 scanned for macro replacements. I have been disabused of this notion,
2502 so now the code just undefines EX_OK before #including unistd.h.
2504 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2505 in the smtp transport. When a block could not be written in a single
2506 write() function, the timeout was being re-applied to each part-write.
2507 This seems wrong - if the receiver was accepting one byte at a time it
2508 would take for ever. The timeout is now adjusted when this happens. It
2509 doesn't have to be particularly precise.
2511 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2512 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2514 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2515 without a domain if the check to <postmaster@domain> fails.
2517 SC/01 Eximstats: added -xls and the ability to specify output files
2518 (patch written by Frank Heydlauf).
2520 SC/02 Eximstats: use FileHandles for outputing results.
2522 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2524 SC/04 Eximstats: fixed display of large numbers with -nvr option
2526 SC/05 Eximstats: fixed merging of reports with empty tables.
2528 SC/06 Eximstats: added the -include_original_destination flag
2530 SC/07 Eximstats: removed tabs and trailing whitespace.
2532 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2534 TK/06 MBOX spool code: Add real "From " MBOX separator line
2535 so the .eml file is really in mbox format (even though
2536 most programs do not really care). Patch from Alex Miller.
2538 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2539 The latter is generated from $received_to and is only set if the
2540 message has one envelope recipient. SA can use these headers,
2541 obviously out-of-the-box. Patch from Alex Miller.
2543 PH/08 The ${def test on a variable was returning false if the variable's
2544 value was "0", contrary to what the specification has always said!
2545 The result should be true unless the variable is empty.
2547 PH/09 The syntax error of a character other than { following "${if
2548 def:variable_name" (after optional whitespace) was not being diagnosed.
2549 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2550 accidental colon was present, for example, could give incorrect results.
2552 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2553 result is used (not including appendfile, where other changes are about
2556 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2557 This involved changing a lot of size variables from int to off_t. It
2558 should work with maildirs and everything.
2560 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2561 spamd dying while we are connected to it.
2563 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2566 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2567 function for BDB 4.3.
2569 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2571 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2572 the sources for PCRE have been re-arranged and more files are now
2575 PH/15 The code I had for printing potentially long long variables in PH/11
2576 above was not the best (it lost precision). The length of off_t variables
2577 is now inspected at build time, and an appropriate printing format (%ld
2578 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2579 to be "long long int" or "long int". This is needed for the internal
2580 formatting function string_vformat().
2582 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2583 the configuration file to be ":syslog", then the script "guesses" where
2584 the logs files are, rather than using the compiled in default. In our
2585 case the guess is not the same as the compiled default, so the script
2586 suddenly stopped working when I started to use syslog. The patch checks
2587 to see if log_file_path is "". If so, it attempts to read it from exim
2588 with no configuration file to get the compiled in version, before it
2589 falls back to the previous guessing code."
2591 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2592 implementing BATV in an Exim configuration. See NewStuff for the gory
2595 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2596 Makefile that are specific to HP-UX.
2598 PH/18 If the "use_postmaster" option was set for a recipient callout together
2599 with the "random" option, the postmaster address was used as the MAIL
2600 FROM address for the random test, but not for the subsequent recipient
2601 test. It is now used for both.
2603 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2604 patch removes a few documentation additions to RFC 3028, because the
2605 latest draft now contains them. It adds the new en;ascii-case comparator
2606 and a new error check for 8bit text in MIME parts. Comparator and
2607 require names are now matched exactly. I enabled the subaddress
2608 extension, but it is not well tested yet (read: it works for me)."
2610 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2611 rework some of the code of TK/09 above to avoid the hardwired use of
2612 "%lld" and "long long". Replaced the call to snprintf() with a call to
2615 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2616 records point to non-existent hosts", "retry timeout exceeded", and
2617 "retry time not reached for any host after a long failure period".
2619 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2620 experimental DomainKeys support:
2622 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2623 (2) On an error such as an illegally used "control", the wrong name for
2624 the control was given.
2626 These problems did NOT occur unless DomainKeys support was compiled.
2628 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2630 PH/24 Added ${if match_ip condition.
2632 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2633 databases so that it will be absolutely obvious if a crash occurs in the
2634 DB library. This is a regular occurrence (often caused by mis-matched
2637 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2638 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2639 picked up on a user's system that detects such things. There doesn't seem
2640 to be a gcc warning option for this - only an attribute that has to be
2641 put on the function's prototype. It seems that in Fedora Core 4 they have
2642 set this on a number of new functions. No doubt there will be more in due
2645 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2646 specify a transport. However, if an address that was verified by such a
2647 router was the subject of a callout, Exim crashed because it tried to
2648 read the rcpt_include_affixes from the non-existent transport. Now it
2649 just assumes that the setting of that option is false. This bug was
2650 introduced by 4.51/PH/31.
2652 PH/28 Changed -d+all to exclude +memory, because that information is very
2653 rarely of interest, but it makes the output a lot bigger. People tend to
2654 do -d+all out of habit.
2656 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2657 code in os-type was giving problems when libc.so lives in lib64, like on
2660 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2661 aren't the modern standard, and it seems that some systems' include files
2662 don't always have them. Exim was already checking for some of the newer
2663 ones like T_AAAA, and defining it itself. I've added checks for all the
2664 record types that Exim uses.
2666 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2667 not automatically generating a new one, as it is supposed to. This
2668 prevented TLS from working. If the file did exist, but contained invalid
2669 data, a new version was generated, as expected. It was only the case of a
2670 non-existent file that was broken.
2672 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2673 with a change in libdomainkeys > 0.64.
2675 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2676 from DNS. If the selector record carries the flag, it now has
2677 precedence over the domain-wide flag.
2679 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2681 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2682 the use of a port name that isn't defined in /etc/services) occurred, the
2683 message was deferred as in a normal delivery, and thus remained on the
2684 spool, instead of being failed because of the mua_wrapper setting. This
2685 is now fixed, and I tidied up some of the mua_wrapper messages at the
2688 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2689 the messages in a hash of arrays rather than using individual hashes.
2690 This is a bit cleaner and results in dramatic memory savings, albeit
2691 at a slight CPU cost.
2693 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2694 as requested by Marc Sherman.
2696 SC/10 Eximstats: added histograms for user specified patterns as requested
2699 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2701 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2702 fopen() in the content-scanning modules that did not already have it.
2708 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2709 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2711 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2713 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2715 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2716 to test Sieve filters that use "vacation".
2718 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2719 that changes the way the GnuTLS parameters are stored in the cache file.
2720 The new format can be generated externally. For backward compatibility,
2721 if the data in the cache doesn't make sense, Exim assumes it has read an
2722 old-format file, and it generates new data and writes a new file. This
2723 means that you can't go back to an older release without removing the
2726 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2727 work if there are any delivery delays because "one_time" forces the
2728 parent to be marked "delivered", so its unseen clone is never tried
2729 again. For this reason, Exim now forbids the simultaneous setting of
2732 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2733 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2734 entry if you want to know the details. The fix, however, neglected to
2735 consider the case when local delivery batching is involved. The test for
2736 "previously delivered" was not happening when checking to see if an
2737 address could be batched with a previous (undelivered) one; under
2738 certain circumstances this could lead to multiple deliveries to the same
2741 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2742 in its include files, and this causes problems building Exim.
2744 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2745 header_syntax) but Exim was just ignoring anything given after a slash.
2746 In particular, this caused confusion with an attempt to use "verify =
2747 reverse_host_lookup/defer_ok". An error is now given when options are
2748 supplied for verify items that do not have them. (Maybe reverse_host_
2749 lookup should have a defer_ok option, but that's a different point.)
2751 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2752 defined by RFC 821) to 2048, because there were problems with some AUTH
2753 commands, and RFC 1869 says the size should be increased for extended
2754 SMTP commands that take arguments.
2756 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2759 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2760 "unknown" error; now it says that the functionality isn't in the binary.
2762 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2763 an address' error message when a string expansion fails (syntax or
2764 whatever). Otherwise the password may appear in the log. Following change
2765 PH/42 below, there is no longer a chance of it appearing in a bounce
2768 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2770 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2771 of its addresses was ever tried. (Bugzilla bug #2.)
2773 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2774 the result incorrectly in the debug output. (It correctly added a newline
2775 to what was transported.)
2777 TF/01 Added $received_time.
2779 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2780 commented out examples of how to interface to a virus scanner and to
2781 SpamAssassin. Also added commented examples of av_scanner and
2782 spamd_address settings.
2784 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2785 and controls are allowed in which ACLs. There were a couple of minor
2786 errors. Some of the entries in the conditions table (which is a table of
2787 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2788 negation of where the condition IS allowed.
2790 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2792 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2793 header file does not have a version number, so I've had to invent a new
2794 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2795 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2796 radiusclient), but it was contributed by a Radius user.
2798 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2799 files or directories, for OpenSSL.
2801 PH/19 When an Exim process that is running as root has to create an Exim log
2802 file, it does so in a subprocess that runs as exim:exim so as to get the
2803 ownership right at creation (otherwise, other Exim processes might see
2804 the file with the wrong ownership). There was no test for failure of this
2805 fork() call, which would lead to the process getting stuck as it waited
2806 for a non-existent subprocess. Forks do occasionally fail when resources
2807 run out. I reviewed all the other calls to fork(); they all seem to check
2810 PH/20 When checking for unexpected SMTP input at connect time (before writing
2811 the banner), Exim was not dealing correctly with a non-positive return
2812 from the read() function. If the client had disconnected by this time,
2813 the result was a log entry for a synchronization error with an empty
2814 string after "input=" when read() returned zero. If read() returned -1
2815 (an event I could not check), uninitialized data bytes were printed.
2816 There were reports of junk text (parts of files, etc) appearing after
2819 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2821 PH/22 Added support for macro redefinition, and (re)definition in between
2822 driver and ACL definitions.
2824 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2825 forgetting to use the resulting value; it was using the unexpanded value.
2827 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2828 hadn't been configured. The fix is from Juergen Kreileder, who
2829 understands it better than I do:
2831 "Here's what I see happening with three configured cyrus_sasl
2832 authenticators configured (plain, login, cram-md5):
2834 On startup auth_cyrus_sasl_init() gets called for each of these.
2835 This means three calls to sasl_listmech() without a specified mech_list.
2836 => SASL tests which mechs of all available mechs actually work
2837 => three warnings about OTP not working
2838 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2840 With the patch, sasl_listmech() also gets called three times. But now
2841 SASL's mech_list option is set to the server_mech specified in the the
2842 authenticator. Or in other words, the answer from sasl_listmech()
2843 gets limited to just the mech you're testing for (which is different
2845 => the return list contains just 'plain' or 'login', 'cram-md5' or
2846 nothing depending on the value of ob->server_mech.
2848 I've just tested the patch: Authentication still works fine,
2849 unavailable mechs specified in the exim configuration are still
2850 caught, and the auth.log warnings about OTP are gone."
2852 PH/25 When debugging is enabled, the contents of the command line are added
2853 to the debugging output, even when log_selector=+arguments is not
2856 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2857 answer is "GNU", and only if the return is "GNU/something" is the answer
2860 PH/27 $acl_verify_message is now set immediately after the failure of a
2861 verification in an ACL, and so is available in subsequent modifiers. In
2862 particular, the message can be preserved by coding like this:
2864 warn !verify = sender
2865 set acl_m0 = $acl_verify_message
2867 Previously, $acl_verify_message was set only while expanding "message"
2868 and "log_message" when a very denied access.
2870 PH/28 Modified OS/os.c-Linux with
2872 -#ifndef OS_LOAD_AVERAGE
2873 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2875 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2876 nomenclature these days.)
2878 PH/29 Installed patch from the Sieve maintainer that adds the options
2879 sieve_useraddress and sieve_subaddress to the redirect router.
2881 PH/30 In these circumstances:
2882 . Two addresses routed to the same list of hosts;
2883 . First host does not offer TLS;
2884 . First host accepts first address;
2885 . First host gives temporary error to second address;
2886 . Second host offers TLS and a TLS session is established;
2887 . Second host accepts second address.
2888 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2889 and peerdn, if requested) that were in fact used only for the second
2892 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2893 attention to any local part prefix or suffix that was matched by the
2894 router that accepted the address. It now behaves in the same way as it
2895 does for delivery: the affixes are removed from the local part unless
2896 rcpt_include_affixes is set on the transport.
2898 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2899 timeout during the DATA phase of an incoming message.
2901 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2902 applied a patch sanctioned by the Sieve maintainer.
2904 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2905 the uid or gid is negative. A case of a negative gid caused this to be
2906 noticed. The fix allows for either to be negative.
2908 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2909 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2912 PH/36 The change PH/12 above was broken. Fixed it.
2914 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2915 the grounds that routing the same address twice would always produce the
2916 same answer. This might have been true once, but it is certainly no
2917 longer true now. Routing a child address may depend on the previous
2918 routing that produced that child. Some complicated redirection strategies
2919 went wrong when messages had multiple recipients, and made Exim's
2920 behaviour dependent on the order in which the addresses were given.
2922 I have moved the duplicate checking until after the routing is complete.
2923 Exim scans the addresses that are assigned to local and remote
2924 transports, and removes any duplicates. This means that more work will be
2925 done, as duplicates will always all be routed, but duplicates are
2926 presumably rare, so I don't expect this is of any significance.
2928 For deliveries to pipes, files, and autoreplies, the duplicate checking
2929 still happens during the routing process, since they are not going to be
2932 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2933 It corrects a timeout issue with spamd. This is Ian's comment: "The
2934 background is that sometimes spamd either never reads data from a
2935 connection it has accepted, or it never writes response data. The exiscan
2936 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2937 blindly assumes that writes won't block so it may never time out."
2939 PH/39 Allow G after quota size as well as K and M.
2941 PH/40 The value set for $authenticated_id in an authenticator may not contain
2942 binary zeroes or newlines because the value is written to log lines and
2943 to spool files. There was no check on this. Now the value is run through
2944 the string_printing() function so that such characters are converted to
2945 printable escape sequences.
2947 PH/41 $message_linecount is a new variable that contains the total number of
2948 lines in the message. Compare $body_linecount, which is the count for the
2951 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2952 bounce and delay warning messages, except in certain special cases, which
2955 (a) An SMTP error message from a remote host;
2956 (b) A message specified in a :fail: redirection;
2957 (c) A message specified in a "fail" command in a system filter;
2958 (d) A message specified in a FAIL return from the queryprogram router;
2959 (e) A message specified by the cannot_route_message router option.
2961 In these cases only, Exim does include the error details in bounce and
2962 warning messages. There are also a few cases where bland messages such
2963 as "unrouteable address" or "local delivery error" are given.
2965 PH/43 $value is now also set for the "else" part of a ${run expansion.
2967 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2968 being worked on, but at least Exim now implements the latest version to
2971 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2972 process to complete was treated as a delivery failure, a timeout while
2973 writing the message to the pipe was logged, but erroneously treated as a
2974 successful delivery. Such timeouts include transport filter timeouts. For
2975 consistency with the overall process timeout, these timeouts are now
2976 treated as errors, giving rise to delivery failures by default. However,
2977 there is now a new Boolean option for the pipe transport called
2978 timeout_defer, which, if set TRUE, converts the failures into defers for
2979 both kinds of timeout. A transport filter timeout is now identified in
2982 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2983 systems where "make" and "gmake" are different, calling "gmake" at top
2984 level broke things. I've arranged for the value of $(MAKE) to be passed
2985 from the Makefile to this script so that it can call the same version of
2989 A note about Exim versions 4.44 and 4.50
2990 ----------------------------------------
2992 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2993 changes of various kinds. As a consequence, a big documentation update was
2994 needed. This delayed the release for rather longer than seemed good, especially
2995 in the light of a couple of (minor) security issues. Therefore, the changes
2996 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2997 release. So 4.44 and 4.50 are in effect two different branches that both start
3000 I have left the 4.50 change log unchanged; it contains all the changes since
3001 4.43. The change log for 4.44 is below; many of its items are identical to
3002 those for 4.50. This seems to be the most sensible way to preserve the
3003 historical information.
3009 1. Minor wording change to the doc/README.SIEVE file.
3011 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3012 computation of the current number of files was incorrect.
3014 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3015 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3018 4. Give more explanation in the error message when the command for a transport
3019 filter fails to execute.
3021 5. There are several places where Exim runs a non-Exim command in a
3022 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3023 was being done only for the command run by the queryprogram router. It is
3024 now done for all such subprocesses. The other cases are: ${run, transport
3025 filters, and the commands run by the lmtp and pipe transports.
3027 6. Added CONFIGURE_GROUP build-time option.
3029 7. Some older OS have a limit of 256 on the maximum number of file
3030 descriptors. Exim was using setrlimit() to set 1000 as a large value
3031 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3032 systems. I've change it so that if it can't get 1000, it tries for 256.
3034 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3035 was an oversight, and furthermore, ever since the addition of extra
3036 controls (e.g. 4.43/32), the checks on when to allow different forms of
3037 "control" were broken. There should now be diagnostics for all cases when a
3038 control that does not make sense is encountered.
3040 9. Added the /retain_sender option to "control=submission".
3042 10. $recipients is now available in the predata ACL (oversight).
3044 11. Tidy the search cache before the fork to do a delivery from a message
3045 received from the command line. Otherwise the child will trigger a lookup
3046 failure and thereby defer the delivery if it tries to use (for example) a
3047 cached ldap connection that the parent has called unbind on.
3049 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3050 of $address_data from the recipient verification was clobbered by the
3051 sender verification.
3053 13. The value of address_data from a sender verification is now available in
3054 $sender_address_data in subsequent conditions in the ACL statement.
3056 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3058 15. Added a new option "connect=<time>" to callout options, to set a different
3061 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3062 was its contents. (It was OK if the option was not defined at all.)
3064 17. A "Completed" log line is now written for messages that are removed from
3065 the spool by the -Mrm option.
3067 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3068 information about exactly what failed.
3070 19. Added -dd to debug only the daemon process.
3072 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3073 handles timeouts, both on the server side and network timeouts. Renamed the
3074 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3076 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3077 It is now set to "smtps".
3079 22. $host_address is now set to the target address during the checking of
3080 ignore_target_hosts.
3082 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3083 being passed; this would have caused $sender_host_name to have been used if
3084 matching the list had actually called for a host name (not very likely,
3085 since this list is usually IP addresses). A host name is now passed as
3088 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3089 code that specifies a non-restarting handler (typically sigaction() in
3090 modern systems) in an attempt to fix a rare and obscure crash bug.
3092 25. Narrowed the window for a race in the daemon that could cause it to ignore
3093 SIGCHLD signals. This is not a major problem, because they are used only to
3094 wake it up if nothing else does.
3096 26. A malformed maildirsize file could cause Exim to calculate negative values
3097 for the mailbox size or file count. Odd effects could occur as a result.
3098 The maildirsize information is now recalculated if the size or filecount
3101 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3102 support for a long time. Removed HAVE_SYS_VFS_H.
3104 28. Installed the latest version of exipick from John Jetmore.
3106 29. In an address list, if the pattern was not a regular expression, an empty
3107 subject address (from a bounce message) matched only if the pattern was an
3108 empty string. Non-empty patterns were not even tested. This was the wrong
3109 because it is perfectly reasonable to use an empty address as part of a
3110 database query. An empty address is now tested by patterns that are
3111 lookups. However, all the other forms of pattern expect the subject to
3112 contain a local part and a domain, and therefore, for them, an empty
3113 address still always fails if the pattern is not itself empty.
3115 30. Exim went into a mad DNS loop when attempting to do a callout where the
3116 host was specified on an smtp transport, and looking it up yielded more
3117 than one IP address.
3119 31. Re-factored the code for checking spool and log partition space into a
3120 function that finds that data and another that does the check. The former
3121 is then used to implement four new variables: $spool_space, $log_space,
3122 $spool_inodes, and $log_inodes.
3124 32. The RFC2047 encoding function was originally intended for short strings
3125 such as real names; it was not keeping to the 75-character limit for
3126 encoded words that the RFC imposes. It now respects the limit, and
3127 generates multiple encoded words if necessary. To be on the safe side, I
3128 have increased the buffer size for the ${rfc2047: expansion operator from
3131 33. It is now permitted to omit both strings after an "if" condition; if the
3132 condition is true, the result is "true". As before, when the second string
3133 is omitted, a false condition yields an empty string. This makes it less
3134 cumbersome to write custom ACL and router conditions.
3136 34. Failure to deliver a bounce message always caused it to be frozen, even if
3137 there was an errors_to setting on the router. The errors_to setting is now
3140 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3141 canonical form (fully expanded) before being placed in
3142 $sender_host_address.
3144 36. The table in the code that translates DNS record types into text (T_A to
3145 "A" for instance) was missing entries for NS and CNAME. It is just possible
3146 that this could have caused confusion if both these types were looked up
3147 for the same domain, because the text type is used as part of Exim's
3148 per-process caching. But the chance of anyone hitting this buglet seems
3151 37. The dnsdb lookup has been extended in a number of ways.
3153 (1) There is a new type, "zns", which walks up the domain tree until it
3154 finds some nameserver records. It should be used with care.
3156 (2) There is a new type, "mxh", which is like "mx" except that it returns
3157 just the host names, not the priorities.
3159 (3) It is now possible to give a list of domains (or IP addresses) to be
3160 looked up. The behaviour when one of the lookups defers can be
3161 controlled by a keyword.
3163 (4) It is now possible to specify the separator character for use when
3164 multiple records are returned.
3166 38. The dnslists ACL condition has been extended: it is now possible to supply
3167 a list of IP addresses and/or domains to be looked up in a particular DNS
3170 39. Added log_selector=+queue_time_overall.
3172 40. When running the queue in the test harness, wait just a tad after forking a
3173 delivery process, to get repeatability of debugging output.
3175 41. Include certificate and key file names in error message when GnuTLS fails
3176 to set them up, because the GnuTLS error message doesn't include the name
3177 of the failing file when there is a problem reading it.
3179 42. Allow both -bf and -bF in the same test run.
3181 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3183 44. The "Exiscan patch" is now merged into the mainline Exim source.
3185 45. Sometimes the final signoff response after QUIT could fail to get
3186 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3187 < 0 before doing a fflush(). This bug looks as though it goes back to the
3188 introduction of TLS in release 3.20, but "sometimes" must have been rare
3189 because the tests only now provoked it.
3191 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3192 (this can affect the format of dates).
3194 47. exim_tidydb, when checking for the continued existence of a message for
3195 which it has found a message-specific retry record, was not finding
3196 messages that were in split spool directories. Consequently, it was
3197 deleting retry records that should have stayed in existence.
3199 48. Steve fixed some bugs in eximstats.
3201 49. The SPA authentication driver was not abandoning authentication and moving
3202 on to the next authenticator when an expansion was forced to fail,
3203 contradicting the general specification for all authenticators. Instead it
3204 was generating a temporary error. It now behaves as specified.
3206 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3207 (the order specifies the preference for clients). The order is now AES256,
3208 AES128, 3DES, ARCFOUR128.
3210 51. Small patch to Sieve code - explicitly set From: when generating an
3213 52. Exim crashed if a remote delivery caused a very long error message to be
3214 recorded - for instance if somebody sent an entire SpamAssassin report back
3215 as a large number of 550 error lines. This bug was coincidentally fixed by
3216 increasing the size of one of Exim's internal buffers (big_buffer) that
3217 happened as part of the Exiscan merge. However, to be on the safe side, I
3218 have made the code more robust (and fixed the comments that describe what
3221 53. Now that there can be additional text after "Completed" in log lines (if
3222 the queue_time_overall log selector is set), a one-byte patch to exigrep
3223 was needed to allow it to recognize "Completed" as not the last thing in
3226 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3227 patch that reportedly fixes this has been added. I am not expert enough to
3228 create a test for it. This is what the patch creator wrote:
3230 "I found a little strange behaviour of ldap code when working with
3231 Windows 2003 AD Domain, where users was placed in more than one
3232 Organization Units. When I tried to give exim partial DN, the exit code
3233 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3234 But simultaneously result of request was absolutely normal ldap result,
3235 so I produce this patch..."
3237 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3238 REFERENCE, so I have modified the code to exclude the patch when that macro
3241 55. Some experimental protocols are using DNS PTR records for new purposes. The
3242 keys for these records are domain names, not reversed IP addresses. The
3243 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3244 leaves it alone. Component reversal etc. now happens only for IP addresses.
3247 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3249 57. Double the size of the debug message buffer (to 2048) so that more of very
3250 long debug lines gets shown.
3252 58. The exicyclog utility now does better if the number of log files to keep
3253 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3255 59. Two changes related to the smtp_active_hostname option:
3257 (1) $smtp_active_hostname is now available as a variable.
3258 (2) The default for smtp_banner uses $smtp_active_hostname instead
3259 of $primary_hostname.
3261 60. The host_aton() function is supposed to be passed a string that is known
3262 to be a valid IP address. However, in the case of IPv6 addresses, it was
3263 not checking this. This is a hostage to fortune. Exim now panics and dies
3264 if the condition is not met. A case was found where this could be provoked
3265 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3266 components; fortuitously, this particular loophole had already been fixed
3267 by change 4.50/55 above.
3269 If there are any other similar loopholes, the new check in host_aton()
3270 itself should stop them being exploited. The report I received stated that
3271 data on the command line could provoke the exploit when Exim was running as
3272 exim, but did not say which command line option was involved. All I could
3273 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3274 running as the user.
3277 61. There was a buffer overflow vulnerability in the SPA authentication code
3278 (which came originally from the Samba project). I have added a test to the
3279 spa_base64_to_bits() function which I hope fixes it.
3282 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3283 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3285 63. The daemon start-up calls getloadavg() while still root for those OS that
3286 need the first call to be done as root, but it missed one case: when
3287 deliver_queue_load_max is set with deliver_drop_privilege. This is
3288 necessary for the benefit of the queue runner, because there is no re-exec
3289 when deliver_drop_privilege is set.
3291 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3292 This has been fixed.
3294 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3295 was in use, was not putting the data itself into the right store pool;
3296 consequently, it could be overwritten for a subsequent message in the same
3297 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3300 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3302 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3303 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3304 the function were treating the return as a boolean value, which happened to
3305 work because 0=false and not-0=true, but is not correct code.
3307 68. The host_aton() function was not handling scoped IPv6 addresses (those
3308 with, for example, "%eth0" on the end) correctly.
3310 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3311 negated items (that is, ~something) in unsigned ints. Some compilers
3312 apparently mutter when there is no cast.
3314 70. If an address verification called from an ACL failed, and did not produce a
3315 user-specific message (i.e. there was only a "system" message), nothing was
3316 put in $acl_verify_message. In this situation, it now puts the system
3319 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3320 session; change 4.31/43 added the unwanted input to the log line - except
3321 that it did not do this in the start of session case. It now does.
3323 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3324 This is wrong and can cause the other end to generate a synchronization
3325 error if it is another Exim or anything else that does the synchronization
3326 check. A QUIT command is no longer sent after a timeout.
3328 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3329 during host lookups.
3331 74. The defer_ok option of callout verification was not working if it was used
3332 when verifying addresses in header lines, that is, for this case:
3334 verify = header_sender/callout=defer_ok
3336 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3337 those file descriptors could be used for SMTP connections. If anything
3338 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3339 could be sent to the SMTP client, causing chaos. The daemon now opens
3340 stdin, stdout, and stderr to /dev/null when it puts itself into the
3343 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3344 log by default. The user can override this with suitable Perl magic.
3346 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3347 the log message when discard triggers, was not working for the DATA ACL or
3348 for the non-SMTP ACL.
3350 78. Error message wording change in sieve.c.
3352 79. If smtp_accept_max_per_host was set, the number of connections could be
3353 restricted to fewer than expected, because the daemon was trying to set up
3354 a new connection before checking whether the processes handling previous
3355 connections had finished. The check for completed processes is now done
3356 earlier. On busy systems, this bug wouldn't be noticed because something
3357 else would have woken the daemon, and it would have reaped the completed
3360 80. If a message was submitted locally by a user whose login name contained one
3361 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3362 It caused a spool format error. I have fixed the spool reading code. A
3363 related problem was that the "from" clause in the Received: line became
3364 illegal because of the space(s). It is now covered by ${quote_local_part.
3366 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3369 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3371 83. Patch from Sieve maintainer to fix unterminated string problem in
3372 "vacation" handling.
3374 84. Some minor changes to the Linux configuration files to help with other
3375 OS variants using glibc.
3377 85. One more patch for Sieve to update vacation handling to latest spec.
3380 ----------------------------------------------------
3381 See the note above about the 4.44 and 4.50 releases.
3382 ----------------------------------------------------
3388 1. Change 4.43/35 introduced a bug that caused file counts to be
3389 incorrectly computed when quota_filecount was set in an appendfile
3392 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3393 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3396 3. Give more explanation in the error message when the command for a transport
3397 filter fails to execute.
3399 4. There are several places where Exim runs a non-Exim command in a
3400 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3401 was being done only for the command run by the queryprogram router. It is
3402 now done for all such subprocesses. The other cases are: ${run, transport
3403 filters, and the commands run by the lmtp and pipe transports.
3405 5. Some older OS have a limit of 256 on the maximum number of file
3406 descriptors. Exim was using setrlimit() to set 1000 as a large value
3407 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3408 systems. I've change it so that if it can't get 1000, it tries for 256.
3410 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3411 was an oversight, and furthermore, ever since the addition of extra
3412 controls (e.g. 4.43/32), the checks on when to allow different forms of
3413 "control" were broken. There should now be diagnostics for all cases when a
3414 control that does not make sense is encountered.
3416 7. $recipients is now available in the predata ACL (oversight).
3418 8. Tidy the search cache before the fork to do a delivery from a message
3419 received from the command line. Otherwise the child will trigger a lookup
3420 failure and thereby defer the delivery if it tries to use (for example) a
3421 cached ldap connection that the parent has called unbind on.
3423 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3424 of $address_data from the recipient verification was clobbered by the
3425 sender verification.
3427 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3428 was its contents. (It was OK if the option was not defined at all.)
3430 11. A "Completed" log line is now written for messages that are removed from
3431 the spool by the -Mrm option.
3433 12. $host_address is now set to the target address during the checking of
3434 ignore_target_hosts.
3436 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3437 being passed; this would have caused $sender_host_name to have been used if
3438 matching the list had actually called for a host name (not very likely,
3439 since this list is usually IP addresses). A host name is now passed as
3442 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3443 code that specifies a non-restarting handler (typically sigaction() in
3444 modern systems) in an attempt to fix a rare and obscure crash bug.
3446 15. Narrowed the window for a race in the daemon that could cause it to ignore
3447 SIGCHLD signals. This is not a major problem, because they are used only to
3448 wake it up if nothing else does.
3450 16. A malformed maildirsize file could cause Exim to calculate negative values
3451 for the mailbox size or file count. Odd effects could occur as a result.
3452 The maildirsize information is now recalculated if the size or filecount
3455 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3456 support for a long time. Removed HAVE_SYS_VFS_H.
3458 18. Updated exipick to current release from John Jetmore.
3460 19. Allow an empty sender to be matched against a lookup in an address list.
3461 Previously the only cases considered were a regular expression, or an
3464 20. Exim went into a mad DNS lookup loop when doing a callout where the
3465 host was specified on the transport, if the DNS lookup yielded more than
3468 21. The RFC2047 encoding function was originally intended for short strings
3469 such as real names; it was not keeping to the 75-character limit for
3470 encoded words that the RFC imposes. It now respects the limit, and
3471 generates multiple encoded words if necessary. To be on the safe side, I
3472 have increased the buffer size for the ${rfc2047: expansion operator from
3475 22. Failure to deliver a bounce message always caused it to be frozen, even if
3476 there was an errors_to setting on the router. The errors_to setting is now
3479 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3480 canonical form (fully expanded) before being placed in
3481 $sender_host_address.
3483 24. Updated eximstats to version 1.33
3485 25. Include certificate and key file names in error message when GnuTLS fails
3486 to set them up, because the GnuTLS error message doesn't include the name
3487 of the failing file when there is a problem reading it.
3489 26. Expand error message when OpenSSL has problems setting up cert/key files.
3492 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3493 (this can affect the format of dates).
3495 28. exim_tidydb, when checking for the continued existence of a message for
3496 which it has found a message-specific retry record, was not finding
3497 messages that were in split spool directories. Consequently, it was
3498 deleting retry records that should have stayed in existence.
3500 29. eximstats updated to version 1.35
3501 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3502 1.35 - bugfix such that pie charts by volume are generated correctly
3504 30. The SPA authentication driver was not abandoning authentication and moving
3505 on to the next authenticator when an expansion was forced to fail,
3506 contradicting the general specification for all authenticators. Instead it
3507 was generating a temporary error. It now behaves as specified.
3509 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3510 (the order specifies the preference for clients). The order is now AES256,
3511 AES128, 3DES, ARCFOUR128.
3513 31. Small patch to Sieve code - explicitly set From: when generating an
3516 32. Exim crashed if a remote delivery caused a very long error message to be
3517 recorded - for instance if somebody sent an entire SpamAssassin report back
3518 as a large number of 550 error lines. This bug was coincidentally fixed by
3519 increasing the size of one of Exim's internal buffers (big_buffer) that
3520 happened as part of the Exiscan merge. However, to be on the safe side, I
3521 have made the code more robust (and fixed the comments that describe what
3524 33. Some experimental protocols are using DNS PTR records for new purposes. The
3525 keys for these records are domain names, not reversed IP addresses. The
3526 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3527 leaves it alone. Component reversal etc. now happens only for IP addresses.
3530 34. The host_aton() function is supposed to be passed a string that is known
3531 to be a valid IP address. However, in the case of IPv6 addresses, it was
3532 not checking this. This is a hostage to fortune. Exim now panics and dies
3533 if the condition is not met. A case was found where this could be provoked
3534 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3535 components; fortuitously, this particular loophole had already been fixed
3536 by change 4.50/55 or 4.44/33 above.
3538 If there are any other similar loopholes, the new check in host_aton()
3539 itself should stop them being exploited. The report I received stated that
3540 data on the command line could provoke the exploit when Exim was running as
3541 exim, but did not say which command line option was involved. All I could
3542 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3543 running as the user.
3546 35. There was a buffer overflow vulnerability in the SPA authentication code
3547 (which came originally from the Samba project). I have added a test to the
3548 spa_base64_to_bits() function which I hope fixes it.
3551 36. The daemon start-up calls getloadavg() while still root for those OS that
3552 need the first call to be done as root, but it missed one case: when
3553 deliver_queue_load_max is set with deliver_drop_privilege. This is
3554 necessary for the benefit of the queue runner, because there is no re-exec
3555 when deliver_drop_privilege is set.
3557 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3558 was in use, was not putting the data itself into the right store pool;
3559 consequently, it could be overwritten for a subsequent message in the same
3560 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3563 38. Sometimes the final signoff response after QUIT could fail to get
3564 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3565 < 0 before doing a fflush(). This bug looks as though it goes back to the
3566 introduction of TLS in release 3.20, but "sometimes" must have been rare
3567 because the tests only now provoked it.
3573 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3574 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3575 Now it returns an integer. A number of calls were still expecting a T/F
3576 return. Fortuitously, in all cases, the tests worked in OK situations,
3577 which is the norm. However, things would have gone wrong on any write
3578 failures on the smtp file descriptor. This function is used when sending
3579 messages over SMTP and also when doing verify callouts.
3581 2. When Exim is called to do synchronous delivery of a locally submitted
3582 message (the -odf or -odi options), it no longer closes stderr before doing
3585 3. Implemented the mua_wrapper option.
3587 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3589 5. Implemented the functions header_remove(), header_testname(),
3590 header_add_at_position(), and receive_remove_recipient(), and exported them
3593 6. If an ACL "warn" statement specified the addition of headers, Exim already
3594 inserted X-ACL-Warn: at the start if there was no header name. However, it
3595 was not making this test for the second and subsequent header lines if
3596 there were newlines in the string. This meant that an invalid header could
3597 be inserted if Exim was badly configured.
3599 7. Allow an ACL "warn" statement to add header lines at the start or after all
3600 the Received: headers, as well as at the end.
3602 8. Added the rcpt_4xx retry error code.
3604 9. Added postmaster_mailfrom=xxx to callout verification option.
3606 10. Added mailfrom=xxxx to the callout verification option, for verify=
3609 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3610 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3612 12. Inserted some casts to stop certain compilers warning when using pointer
3613 differences as field lengths or precisions in printf-type calls (mostly
3614 affecting debugging statements).
3616 13. Added optional readline() support for -be (dynamically loaded).
3618 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3619 same clock tick as a message's arrival, so that its received time was the
3620 same as the "first fail" time on the retry record, and that message
3621 remained on the queue past the ultimate address timeout, every queue runner
3622 would try a delivery (because it was past the ultimate address timeout) but
3623 after another failure, the ultimate address timeout, which should have then
3624 bounced the address, did not kick in. This was a "< instead of <=" error;
3625 in most cases the first failure would have been in the next clock tick
3626 after the received time, and all would be well.
3628 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3629 being recognized when the domain list was tested by the match_domain
3630 condition in an expansion string.
3632 16. Added the ${str2b64: operator.
3634 17. Exim was always calling setrlimit() to set a large limit for the number of
3635 processes, without checking whether the existing limit was already
3636 adequate. (It did check for the limit on file descriptors.) Furthermore,
3637 errors from getrlimit() and setrlimit() were being ignored. Now they are
3638 logged to the main and panic logs, but Exim does carry on, to try to do its
3639 job under whatever limits there are.
3641 18. Imported PCRE 5.0.
3643 19. Trivial typo in log message " temporarily refused connection" (the leading
3646 20. If the log selector return_path_on_delivery was set and an address was
3647 redirected to /dev/null, the delivery process crashed because it assumed
3648 that a return path would always be set for a "successful" delivery. In this
3649 case, the whole delivery is bypassed as an optimization, and therefore no
3652 21. Internal re-arrangement: the function for sending a challenge and reading
3653 a response while authentication was assuming a zero-terminated challenge
3654 string. It's now changed to take a pointer and a length, to allow for
3655 binary data in such strings.
3657 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3659 23. Exim was not respecting finduser_retries when seeking the login of the
3660 uid under which it was called; it was always trying 10 times. (The default
3661 setting of finduser_retries is zero.) Also, it was sleeping after the final
3662 failure, which is pointless.
3664 24. Implemented tls_on_connect_ports.
3666 25. Implemented acl_smtp_predata.
3668 26. If the domain in control=submission is set empty, Exim assumes that the
3669 authenticated id is a complete email address when it generates From: or
3670 Sender: header lines.
3672 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3673 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3674 chown and chgrp in /bin and hostname in /usr/bin.
3676 28. Exim was keeping the "process log" file open after each use, just as it
3677 does for the main log. This opens the possibility of it remaining open for
3678 long periods when the USR1 signal hits a daemon. Occasional processlog
3679 errors were reported, that could have been caused by this. Anyway, it seems
3680 much more sensible not to leave this file open at all, so that is what now
3683 29. The long-running daemon process does not normally write to the log once it
3684 has entered its main loop, and it closes the log before doing so. This is
3685 so that log files can straightforwardly be renamed and moved. However,
3686 there are a couple of unusual error situations where the daemon does write
3687 log entries, and I had neglected to close the log afterwards.
3689 30. The text of an SMTP error response that was received during a remote
3690 delivery was being truncated at 512 bytes. This is too short for some of
3691 the long messages that one sometimes sees. I've increased the limit to
3694 31. It is now possible to make retry rules that apply only when a message has a
3695 specific sender, in particular, an empty sender.
3697 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3698 it possible to be selective about when SMTP synchronization is enforced.
3700 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3702 32. Implemented hosts_connection_nolog.
3704 33. Added an ACL for QUIT.
3706 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3709 35. Added mailbox_size and mailbox_filecount to appendfile.
3711 36. Added control = no_multiline_responses to ACLs.
3713 37. There was a bug in the logic of the code that waits for the clock to tick
3714 in the case where the clock went backwards by a substantial amount such
3715 that the microsecond fraction of "now" was more than the microsecond
3716 fraction of "then" (but the whole seconds number was less).
3718 38. Added support for the libradius Radius client library this is found on
3719 FreeBSD (previously only the radiusclient library was supported).
3725 1. When certain lookups returned multiple values in the form name=value, the
3726 quoting of the values was not always being done properly. Specifically:
3727 (a) If the value started with a double quote, but contained no whitespace,
3729 (b) If the value contained whitespace other than a space character (i.e.
3730 tabs or newlines or carriage returns) it was not quoted.
3731 This fix has been applied to the mysql and pgsql lookups by writing a
3732 separate quoting function and calling it from the lookup code. The fix
3733 should probably also be applied to nisplus, ibase and oracle lookups, but
3734 since I cannot test any of those, I have not disturbed their existing code.
3736 2. A hit in the callout cache for a specific address caused a log line with no
3737 reason for rejecting RCPT. Now it says "Previous (cached) callout
3738 verification failure".
3740 3. There was an off-by-one bug in the queryprogram router. An over-long
3741 return line was truncated at 256 instead of 255 characters, thereby
3742 overflowing its buffer with the terminating zero. As well as fixing this, I
3743 have increased the buffer size to 1024 (and made a note to document this).
3745 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3746 when Exim is waiting for an SMTP response from a remote server, Exim
3747 restarts its select() call on the socket, thereby resetting its timeout.
3748 This is not a problem when such interrupts are rare. Somebody set up a cron
3749 job to run exiwhat every 2 minutes, which is less than the normal select()
3750 timeout (5 or 10 minutes). This meant that the select() timeout never
3751 kicked in because it was always reset. I have fixed this by comparing the
3752 time when an interrupt arrives with the time at the start of the first call
3753 to select(). If more time than the timeout has elapsed, the interrupt is
3754 treated as a timeout.
3756 5. Some internal re-factoring in preparation for the addition of Sieve
3757 extensions (by MH). In particular, the "personal" test is moved to a
3758 separate function, and given an option for scanning Cc: and Bcc: (which is
3759 not set for Exim filters).
3761 6. When Exim created an email address using the login of the caller as the
3762 local part (e.g. when creating a From: or Sender: header line), it was not
3763 quoting the local part when it contained special characters such as @.
3765 7. Installed new OpenBSD configuration files.
3767 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3768 try to make them clearer.
3770 9. Callout options, other than the timeout value, were being ignored when
3771 verifying sender addresses in header lines. For example, when using
3773 verify = header_sender/callout=no_cache
3775 the cache was (incorrectly) being used.
3777 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3778 only the Cygwin environment.
3780 11. When return_path_on_delivery was set as a log selector, if different remote
3781 addresses in the same message used different return paths and parallel
3782 remote delivery occurred, the wrong values would sometimes be logged.
3783 (Whenever a remote delivery process finished, the return path value from
3784 the most recently started remote delivery process was logged.)
3786 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3787 lines when AUTH and/or TLS are in use. This is the "received protocol"
3788 field. Exim used to use "asmtp" for authenticated SMTP, without any
3789 indication (in the protocol name) for TLS use. Now it follows the RFC and
3790 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3791 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3792 names appear in log lines as well as in Received: header lines.
3794 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3795 extensions, and comparison tests, and to fix some bugs.
3797 14. Changes to the "personal" filter test:
3799 (1) The test was buggy in that it was just doing the equivalent of
3800 "contains" tests on header lines. For example, if a user's address was
3801 anne@some.where, the "personal" test would incorrectly be true for
3803 To: susanne@some.where
3805 This test is now done by extracting each address from the header in turn,
3806 and checking the entire address. Other tests that are part of "personal"
3807 are now done using regular expressions (for example, to check local parts
3808 of addresses in From: header lines).
3810 (2) The list of non-personal local parts in From: addresses has been
3811 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3812 taken from the Sieve specification recommendations.
3814 (3) If the message contains any header line starting with "List-" it is
3815 treated as non-personal.
3817 (4) The test for "circular" in the Subject: header line has been removed
3818 because it now seems ill-conceived.
3820 15. Minor typos in src/EDITME comments corrected.
3822 16. Installed latest exipick from John Jetmore.
3824 17. If headers_add on a router specified a text string that was too long for
3825 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3826 of string_sprintf() is now avoided.
3828 18. $message_body_size was not set (it was always zero) when running the DATA
3829 ACL and the local_scan() function.
3831 19. For the "mail" command in an Exim filter, no default was being set for
3832 the once_repeat time, causing a random time value to be used if "once" was
3833 specified. (If the value happened to be <= 0, no repeat happened.) The
3834 default is now 0s, meaning "never repeat". The "vacation" command was OK
3835 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3836 (I found it when inspecting the code).
3838 20. There is now an overall timeout for performing a callout verification. It
3839 defaults to 4 times the callout timeout, which applies to individual SMTP
3840 commands during the callout. The overall timeout applies when there is more
3841 than one host that can be tried. The timeout is checked before trying the
3842 next host. This prevents very long delays if there are a large number of
3843 hosts and all are timing out (e.g. when the network connections are timing
3844 out). The value of the overall timeout can be changed by specifying an
3845 additional sub-option for "callout", called "maxwait". For example:
3847 verify = sender/callout=5s,maxwait=20s
3849 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3850 to the end before writing, but this should make it even safer).
3852 22. Exim was forgetting that it had advertised PIPELINING for the second and
3853 subsequent messages on an SMTP connection. It was also not resetting its
3854 memory on STARTTLS and an internal HELO.
3856 23. When Exim logs an SMTP synchronization error within a session, it now
3857 records whether PIPELINING has been advertised or not.
3859 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3860 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3861 rather than long int.
3863 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3865 26. Added the never_mail option to autoreply.
3871 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3872 crash if the getsockname() call failed; this can happen if a connection is
3873 closed very soon after it is established. The problem was simply in the
3874 order in which certain operations were done, causing Exim to try to write
3875 to the SMTP stream before it had set up the file descriptor. The bug has
3876 been fixed by making things happen in the correct order.
3882 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3883 before the connection was closed, thus losing the rejection response.
3885 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3886 some early Solaris releases, but causes trouble in current releases where
3887 socklen_t is defined.
3889 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3892 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3895 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3896 that releasing the top of it at the end releases what was used for sub-
3897 expansions (unless the block got too big). However, discard this block if
3898 the first thing is a variable or header, so that we can use its block when
3899 it is dynamic (useful for very large $message_headers, for example).
3901 6. Lookups now cache *every* query, not just the most recent. A new, separate
3902 store pool is used for this. It can be recovered when all lookup caches are
3903 flushed. Lookups now release memory at the end of their result strings.
3904 This has involved some general refactoring of the lookup sources.
3906 7. Some code has been added to the store_xxx() functions to reduce the amount
3907 of flapping under certain conditions.
3909 8. log_incoming_interface used to affect only the <= reception log lines. Now
3910 it causes the local interface and port to be added to several more SMTP log
3911 lines, for example "SMTP connection from", and rejection lines.
3913 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3915 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3917 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3918 could be overwritten at the end of the current message (or the start of a
3919 new message if it was set in a HELO ACL). The value is now preserved for
3920 the duration of the SMTP connection.
3922 12. If a transport had a headers_rewrite setting, and a matching header line
3923 contained an unqualified address, that address was qualified, even if it
3924 did not match any rewriting rules. The underlying bug was that the values
3925 of the flags that permit the existence of unqualified sender and recipient
3926 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3927 non-local messages, and by -bnq for local messages) were not being
3928 preserved with the message after it was received.
3930 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3931 "next input=" as part of the text comprising the host identity instead of
3932 the correct text. The code was using the same buffer for two different
3933 strings. However, depending on which order the printing function evaluated
3934 its arguments, the bug did not always show up. Under Linux, for example, my
3935 test suite worked just fine.
3937 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3938 doesn't work with some older versions of Perl. It has been changed to "my",
3939 which in any case is probably the better facility to use.
3941 15. A really picky compiler found some instances of statements for creating
3942 error messages that either had too many or two few arguments for the format
3945 16. The size of the buffer for calls to the DNS resolver has been increased
3946 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3947 for addresses that have a lot of PTR records. This alleviates a problem; it
3948 does not fully solve it.
3950 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3951 buffer now truncates the list and logs the incident, which is the same
3952 action as happens when Exim is looking up a host name and its aliases.
3953 Previously in this situation something unpredictable would happen;
3954 sometimes it was "internal error: store_reset failed".
3956 18. If a server dropped the connection unexpectedly when an Exim client was
3957 using GnuTLS and trying to read a response, the client delivery process
3958 crashed while trying to generate an error log message.
3960 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3961 string, for example:
3963 warn message = H1: something\nH2: something
3965 the text was added as a single header line from Exim's point of view
3966 though it ended up OK in the delivered message. However, searching for the
3967 second and subsequent header lines using $h_h2: did not work. This has been
3968 fixed. Similarly, if a system filter added multiple headers in this way,
3969 the routers could not see them.
3971 20. Expanded the error message when iplsearch is called with an invalid key to
3972 suggest using net-iplsearch in a host list.
3974 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3975 ACLs are no longer actually imposed (and a message to that effect is
3978 22. If a "gecos" field in a passwd entry contained escaped characters, in
3979 particular, if it contained a \" sequence, Exim got it wrong when building
3980 a From: or a Sender: header from that name. A second bug also caused
3981 incorrect handling when an unquoted " was present following a character
3982 that needed quoting.
3984 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3985 was not being matched caselessly.
3987 24. Arranged for all hyphens in the exim.8 source to be escaped with
3990 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3991 properly. Recipient callout cache records were still being keyed to include
3992 the sender, even when use_sender was set false. This led to far more
3993 callouts that were necessary. The sender is no longer included in the key
3994 when use_sender is false.
3996 26. Added "control = submission" modifier to ACLs.
3998 27. Added the ${base62d: operator to decode base 62 numbers.
4000 28. dnsdb lookups can now access SRV records.
4002 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4003 the configuration file.
4005 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4006 (-v) mode. This makes the output for a verbose queue run more intelligible.
4008 31. Added a use_postmaster feature to recipient callouts.
4010 32. Added the $body_zerocount variable, containing the number of binary zero
4011 bytes in the message body.
4013 33. The time of last modification of the "new" subdirectory is now used as the
4014 "mailbox time last read" when there is a quota error for a maildir
4017 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4019 35. Added +ignore_unknown as a special item in host lists.
4021 36. Code for decoding IPv6 addresses in host lists is now included, even if
4022 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4023 address was recognized as an IP address, but was then not correctly decoded
4024 into binary, causing unexpected and incorrect effects when compared with
4031 1. Very minor rewording of debugging text in manualroute to say "list of
4032 hosts" instead of "hostlist".
4034 2. If verify=header_syntax was set, and a header line with an unqualified
4035 address (no domain) and a large number of spaces between the end of the
4036 name and the colon was received, the reception process suffered a buffer
4037 overflow, and (when I tested it) crashed. This was caused by some obsolete
4038 code that should have been removed. The fix is to remove it!
4040 3. When running in the test harness, delay a bit after writing a bounce
4041 message to get a bit more predictability in the log output.
4043 4. Added a call to search_tidyup() just before forking a reception process. In
4044 theory, someone could use a lookup in the expansion of smtp_accept_max_
4045 per_host which, without the tidyup, could leave open a database connection.
4047 5. Added the variables $recipient_data and $sender_data which get set from a
4048 lookup success in an ACL "recipients" or "senders" condition, or a router
4049 "senders" option, similar to $domain_data and $local_part_data.
4051 6. Moved the writing of debug_print from before to after the "senders" test
4054 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4055 problems for message scanning, either using a data ACL, or using
4056 local_scan() because the Received: header was not generated till after they
4057 were called (in order to set the time as the time of reception completion).
4058 I have revised the way this works. The header is now generated after the
4059 body is received, but before the ACL or local_scan() are called. After they
4060 are run, the timestamp in the header is updated.
4066 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4067 before starting a queue runner without re-exec. This happened only when
4068 deliver_drop_privilege was set or when the Exim user was set to root. The
4069 effect of the bug was that timeouts during subsequent deliveries caused
4070 crashes instead of being properly handled. The handler is now left at its
4071 default (and expected) setting.
4073 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4074 message, again when deliver_drop_privilege is set or Exim is run as root.
4075 The bug described in (1) was not present in this case, but the tidying up
4076 of the other signals was missing. I have made the two cases consistent.
4078 3. The ignore_target_hosts setting on a manualroute router was being ignored
4079 for hosts that were looked up using the /MX notation.
4081 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4084 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4085 operated on the sender address. After changing the $sender_address to <>
4086 for the sender address verify, Exim was re-instated it as the original
4087 (before rewriting) address, but remembering that it had rewritten it, so it
4088 wasn't rewriting it again. This bug also had the effect of breaking the
4089 sender address verification caching when the sender address was rewritten.
4091 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4092 This has been changed so that if the ip literal address matches
4093 ignore_target_hosts, the router declines.
4095 7. Added expansion conditions match_domain, match_address, and match_local_
4096 part (NOT match_host).
4098 8. The placeholder for the Received: header didn't have a length field set.
4100 9. Added code to Exim itself and to exim_lock to test for a specific race
4101 condition that could lead to file corruption when using MBX delivery. The
4102 issue is with the lockfile that is created in /tmp. If this file is removed
4103 after a process has opened it but before that process has acquired a lock,
4104 there is the potential for a second process to recreate the file and also
4105 acquire a lock. This could lead to two Exim processes writing to the file
4106 at the same time. The added code performs the same test as UW imapd; it
4107 checks after acquiring the lock that its file descriptor still refers to
4108 the same named file.
4110 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4111 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4114 11. Added the smtp_active_hostname option. If used, this will typically be made
4115 to depend on the incoming interface address. Because $interface_address is
4116 not set up until the daemon has forked a reception process, error responses
4117 that can happen earlier (such as "too many connections") no longer contain
4120 12. If an expansion in a condition on a "warn" statement fails because a lookup
4121 defers, the "warn" statement is abandoned, and the next ACL statement is
4122 processed. Previously this caused the whole ACL to be aborted.
4124 13. Added the iplsearch lookup type.
4126 14. Added ident_timeout as a log selector.
4128 15. Added tls_certificate_verified as a log selector.
4130 16. Added a global option tls_require_ciphers (compare the smtp transport
4131 option of the same name). This controls incoming TLS connections.
4133 17. I finally figured out how to make tls_require_ciphers do a similar thing
4134 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4135 before starting the TLS session.
4137 18. Tabs are now shown as \t in -bP output.
4139 19. If the log selector return_path_on_delivery was set, Exim crashed when
4140 bouncing a message because it had too many Received: header lines.
4142 20. If two routers both had headers_remove settings, and the first one included
4143 a superfluous trailing colon, the final name in the first list and the
4144 first name in the second list were incorrectly joined into one item (with a
4145 colon in the middle).
4151 1. Added -C and -D options to the exinext utility, mainly to make it easier
4152 to include in the automated testing, but these could be helpful when
4153 multiple configurations are in use.
4155 2. The exinext utility was not formatting the output nicely when there was
4156 an alternate port involved in the retry record key, nor when there was a
4157 message id as well (for retries that were specific to a specific message
4158 and a specific host). It was also confused by IPv6 addresses, because of
4159 the additional colons they contain. I have fixed the IPv4 problem, and
4160 patched it up to do a reasonable job for IPv6.
4162 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4163 delivery, the log line now contains "pipelined" if PIPELINING was used.
4165 4. An SMTP transport process used to panic and die if the bind() call to set
4166 an explicit outgoing interface failed. This has been changed; it is now
4167 treated in the same way as a connect() failure.
4169 5. A reference to $sender_host_name in the part of a conditional expansion
4170 that was being skipped was still causing a DNS lookup. This no longer
4173 6. The def: expansion condition was not recognizing references to header lines
4174 that used bh_ and bheader_.
4176 7. Added the _cache feature to named lists.
4178 8. The code for checking quota_filecount in the appendfile transport was
4179 allowing one more file than it should have been.
4181 9. For compatibility with Sendmail, the command line option
4189 and sets the incoming protocol and host name (for trusted callers). The
4190 host name and its colon can be omitted when only the protocol is to be set.
4191 Note the Exim already has two private options, -pd and -ps, that refer to
4192 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4193 "s", but I don't think that's a major issue.
4195 10. A number of refactoring changes to the code, none of which should affect
4198 (a) The number of logging options was getting close to filling up the
4199 32-bit word that was used as a bit map. I have split them into two classes:
4200 those that are passed in the argument to log_write(), and those that are
4201 only ever tested independently outside of that function. These are now in
4202 separate 32-bit words, so there is plenty of room for expansion again.
4203 There is no change in the user interface or the logging behaviour.
4205 (b) When building, for example, log lines, the code previously used a
4206 macro that called string_cat() twice, in order to add two strings. This is
4207 not really sufficiently general. Furthermore, there was one instance where
4208 it was actually wrong because one of the argument was used twice, and in
4209 one call a function was used. (As it happened, calling the function twice
4210 did not affect the overall behaviour.) The macro has been replaced by a
4211 function that can join an arbitrary number of extra strings onto a growing
4214 (c) The code for expansion conditions now uses a table and a binary chop
4215 instead of a serial search (which was left over from when there were very
4216 few conditions). Also, it now recognizes conditions like "pam" even when
4217 the relevant support is not compiled in: a suitably worded error message is
4218 given if an attempt is made to use such a condition.
4220 11. Added ${time_interval:xxxxx}.
4222 12. A bug was causing one of the ddress fields not to be passed back correctly
4223 from remote delivery subprocesses. The field in question was not being
4224 subsequently used, so this caused to problems in practice.
4226 13. Added new log selectors queue_time and deliver_time.
4228 14. Might have fixed a bug in maildirsizefile handling that threw up
4229 "unexpected character" debug warnings, and recalculated the data
4230 unnecessarily. In any case, I expanded the warning message to give more
4233 15. Added the message "Restricted characters in address" to the statements in
4234 the default ACL that block characters like @ and % in local parts.
4236 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4237 Three changes have been made:
4239 (a) There was a serious bug; a negative response to MAIL caused the whole
4240 recipient domain to be cached as invalid, thereby blocking all messages
4241 to all local parts at the same domain, from all senders. This bug has
4242 been fixed. The domain is no longer cached after a negative response to
4243 MAIL if the sender used is not empty.
4245 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4248 (c) A new callout option, "use_sender" has been added for people who want
4249 the modified behaviour.
4255 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4258 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4259 indeed breaks things for older releases.
4261 3. Added additional logging to the case where there is a problem reading data
4262 from a filter that is running in a subprocess using a pipe, in order to
4263 try to track down a specific problem.
4265 4. Testing facility fudge: when running in the test harness and attempting
4266 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4267 getting "No route to host". Convert this to a timeout.
4269 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4272 6. Some OS don't have socklen_t but use size_t instead. This affects the
4273 fifth argument of getsockopt() amongst other things. This is now
4274 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4275 can be set for individual OS. I have set it for SunOS5, OSF1, and
4276 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4277 some earlier ones do not.
4279 7. Change 4.30/15 was not doing the test caselessly.
4281 8. The standard form for an IPv6 address literal was being rejected by address
4282 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4283 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4284 this, as well as the form without the "IPv6" on the front (but only when
4285 address literals are enabled, of course).
4287 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4289 10. Exim crashed if a message with an empty sender address specified by -f
4290 encountered a router with an errors_to setting. This could be provoked only
4291 by a command such as
4295 where an empty string was supplied; "<>" did not hit this bug.
4297 11. Installed PCRE release 4.5.
4299 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4300 remained set. It is now erased.
4302 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4303 times from message ids (which are base 36 rather than the normal 62).
4305 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4306 were being counted as actual protocol errors, and logged if the log
4307 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4308 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4309 and DATA following a set of rejected RCPTs do not count as protocol errors.
4310 In other words, Exim assumes they were pipelined, though this may not
4311 actually be the case. Of course, in all cases the client gets an
4312 appropriate error code.
4314 15. If a lookup fails in an ACL condition, a message about the failure may
4315 be available; it is used if testing the ACL cannot continue, because most
4316 such messages specify what the cause of the deferral is. However, some
4317 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4318 that caused an old message to be retained and used if a later statement
4319 caused a defer, replacing the real cause of the deferral.
4321 16. If an IP address had so many PTR records that the DNS lookup buffer
4322 was not large enough to hold them, Exim could crash while trying to process
4323 the truncated data. It now detects and logs this case.
4325 17. Further to 4.21/58, another change has been made: if (and only if) the
4326 first line of a message (the first header line) ends with CRLF, a bare LF
4327 in a subsequent header line has a space inserted after it, so as not to
4328 terminate the header.
4330 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4331 unnecessarily, used atoi() instead of strtol(), and didn't check the
4332 termination when getting file sizes from file names by regex.
4334 19. Completely re-implemented the support for maildirsize files, in the light
4335 of a number of problems with the previous contributed implementation
4336 (4.30/29). In particular:
4338 . If the quota is zero, the maildirsize file is maintained, but no quota is
4341 . If the maildir directory does not exist, it is created before any attempt
4342 to write a maildirsize file.
4344 . The quota value in the file is just a cache; if the quota is changed in
4345 the transport, the new value overrides.
4347 . A regular expression is available for excluding directories from the
4350 20. The autoreply transport checks the characters in options that define the
4351 message's headers; it allows continued headers, but it was checking with
4352 isspace() after an embedded newline instead of explicitly looking for a
4355 21. If all the "regular" hosts to which an address was routed had passed their
4356 expiry times, and had not reached their retry times, the address was
4357 bounced, even if fallback hosts were defined. Now Exim should go on to try
4360 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4361 equivalent code in the SMTP transport. Some hosts send humungous responses
4362 to HELO/EHLO, more than 1024 it seems.
4364 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4365 gives compiler warnings in some environments. I've now done it "properly",
4368 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4369 (because of problems with the built-in one) was declared to return uschar *
4370 instead of char *, causing compiler failure.
4372 25. Fixed a file descriptor leak when processing alias/forward files.
4374 26. Fixed a minor format string issue in dbfn.c.
4376 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4378 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4379 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4382 29. When a sender address is verified, it is cached, to save repeating the test
4383 when there is more than one recipient in a message. However, when the
4384 verification involves a callout, it is possible for different callout
4385 options to be set for different recipients. It is too complicated to keep
4386 track of this in the cache, so now Exim always runs a verification when a
4387 callout is required, relying on the callout cache for the optimization.
4388 The overhead is duplication of the address routing, but this should not be
4391 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4392 to be verified with callout=postmaster, and the main callout worked but the
4393 postmaster check failed, the verification correctly failed. However, if a
4394 subsequent RCPT command asked for sender verification *without* the
4395 postmaster check, incorrect caching caused this verification also to fail,
4398 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4399 it was not caching the DNS options (qualify_single, search_parents) that
4400 were used when the lookup failed. A subsequent lookup with different
4401 options therefore always gave the same answer, though there were cases
4402 where it should not have. (Example: a "domains = !$mx_any" option on a
4403 dnslookup router: the "domains" option is always processed without any
4404 widening, but the router might have qualify_single set.) Now Exim uses the
4405 cached value only when the same options are set.
4407 32. Added John Jetmore's "exipick" utility to the distribution.
4409 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4410 than a timeout (e.g. a certificate is required, and is not provided), an
4411 Exim server now closes the connection immediately. Previously it waited for
4412 the client to close - but if the client is SSL, it seems that they each
4413 wait for each other, leading to a delay before one of them times out.
4415 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4416 maintained 0.8.x compatibility because I don't think many are using it, and
4417 it is clearly obsolete.
4419 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4422 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4423 client certificate was expired. A simple patch fixes this, though I don't
4424 understand the full logic of why the verify callback is called multiple
4427 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4428 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4429 which causes problems with some clients (such as the Certicom SSL Plus
4430 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4431 disables the coutermeasure allowing Eudora to connect."
4433 38. Exim was not checking that a write() to a log file succeeded. This could
4434 lead to Bad Things if a log got too big, in particular if it hit a file
4435 size limit. Exim now panics and dies if it cannot write to a log file, just
4436 as it does if it cannot open a log file.
4438 39. Modified OS/Makefile-Linux so that it now contains
4440 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4442 The two -D definitions ensure that Exim is compiled with large file
4443 support, which makes it possible to handle log files that are bigger than
4446 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4447 instance) a domain was checked against a named list that involved a lookup,
4448 causing $domain_data to be set, then another domain was checked against the
4449 same list, then the first domain was re-checked, the value of $domain_data
4450 after the final check could be wrong. In particular, if the second check
4451 failed, it could be set empty. This bug probably also applied to
4454 41. The strip_trailing_dot option was not being applied to the address given
4455 with the -f command-line option.
4457 42. The code for reading a message's header from the spool was incrementing
4458 $received_count, but never initializing it. This meant that the value was
4459 incorrect (doubled) while delivering a message in the same process in which
4460 it was received. In the most common configuration of Exim, this never
4461 happens - a fresh exec is done - but it can happen when
4462 deliver_drop_privilege is set.
4464 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4465 it now includes up to 150 characters of the unexpected data in the log
4468 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4469 and building data strings. The size of both of these buffers was 10 000
4470 bytes - far larger than anybody would *ever* want, thought I. Needless to
4471 say, somebody hit the limit. I have increased the maximum line length to
4472 20 000 and the maximum data length of concatenated lines to 100 000. I have
4473 also fixed two bugs, because there was no checking on these buffers. Tsk,
4474 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4475 buffer is too small.
4477 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4478 lsearch lookups. Now it does.
4480 46. When parsing a route_list item in a manualroute router, a fixed-length
4481 buffer was used for the list of hosts. I made this 1024 bytes long,
4482 thinking that nobody would ever have a list of hosts that long. Wrong.
4483 Somebody had a whole pile of complicated expansion conditions, and the
4484 string was silently truncated, leading to an expansion error. It turns out
4485 that it is easier to change to an unlimited length (owing to other changes
4486 that have happened since this code was originally written) than to build
4487 structure for giving a limitation error. The length of the item that
4488 expands into the list of hosts is now unlimited.
4490 47. The lsearch lookup could not handle data where the length of text line was
4491 more than 4095 characters. Such lines were truncated, leading to shortened
4492 data being returned. It should now handle lines of any length.
4494 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4495 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4498 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4499 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4500 no longer happens in comment lines. A list of replacements is now placed
4501 at the head of all of the source files, except those whose only change is
4502 to replace PERL_COMMAND in the very first #! line.
4504 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4505 messages on the queue, with a bottom-up merge sort, using code contributed
4506 by Michael Haardt. This should make operations like -bp somewhat faster on
4507 large queues. It won't affect queue runners, except when queue_run_in_order
4510 51. Installed eximstats 1.31 in the distribution.
4512 52. Added support for SRV lookups to the dnslookup router.
4514 53. If an ACL referred to $message_body or $message_body_end, the value was not
4515 reset for any messages that followed in the same SMTP session.
4517 54. The store-handling optimization for building very long strings was not
4518 differentiating between the different store pools. I don't think this
4519 actually made any difference in practice, but I've tidied it.
4521 55. While running the routers to verify a sender address, $sender_address
4522 was still set to the sender address. This is wrong, because when routing to
4523 send a bounce to the sender, it would be empty. Therefore, I have changed
4524 it so that, while verifying a sender address, $sender_address is set to <>.
4525 (There is no change to what happens when verifying a recipient address.)
4527 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4528 target A or AAAA records (if not already returned) without resetting the
4529 qualify_single or search_parents options of the DNS resolver. These are
4530 inappropriate in this case because the targets of MX and SRV records must
4531 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4532 target that, when qualified, matched something in the local domain. These
4533 two options are now turned off when doing these lookups.
4535 57. It seems that at least some releases of Reiserfs (which does not have the
4536 concept of a fixed number of inodes) returns zero and not -1 for the
4537 number of available inodes. This interacted badly with check_spool_inodes,
4538 which assumed that -1 was the "no such thing" setting. What I have done is
4539 to check that the total number of inodes is greater than zero before doing
4540 the test of how many are available.
4542 58. When a "warn" ACL statement has a log_message modifier, the message is
4543 remembered, and not repeated. This is to avoid a lot of repetition when a
4544 message has many recipients that cause the same warning to be written.
4545 Howewer, Exim was preserving the list of already written lines for an
4546 entire SMTP session, which doesn't seem right. The memory is now reset if a
4547 new message is started.
4549 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4550 debugging output unless log_rewrite was also set.
4552 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4553 of (int)(handle) when we know that handle contains (void *)(-1).
4555 61. The Exim daemon panic-logs an error return when it closes the incoming
4556 connection. However "connection reset by peer" seems to be common, and
4557 isn't really an error worthy of noting specially, so that particular error
4560 62. When Exim is trying to find all the local interfaces, it used to panic and
4561 die if the ioctl to get the interface flags failed. However, it seems that
4562 on at least one OS (Solaris 9) it is possible to have an interface that is
4563 included in the list of interfaces, but for which you get a failure error
4564 for this call. This happens when the interface is not "plumbed" into a
4565 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4566 failure of the "get flags" call assumes that the interface is down.
4568 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4569 makes life easier for people who are doing arithmetic on fields extracted
4570 from dates, where you often get leading zeros that should not be
4571 interpreted as octal.
4573 64. Added qualify_domain to the redirect router, to override the global
4576 65. If a pathologically long header line contained very many addresses (the
4577 report of this problem mentioned 10 000) and each of them was rewritten,
4578 Exim could use up a very large amount of memory. (It kept on making new
4579 copies of the header line as it rewrote, and never released the old ones.)
4580 At the expense of a bit more processing, the header rewriting function has
4581 been changed so that it no longer eats memory in this way.
4583 66. The generation of the Received: header has been moved from the time that a
4584 message starts to be received, to the time that it finishes. The timestamp
4585 in the Received: header should now be very close to that of the <= log
4586 line. There are two side-effects of this change:
4588 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4589 logged header lines no longer include the local Received: line, because
4590 it has not yet been created. The same applies to a copy of the message
4591 that is returned to a non-SMTP sender when a message is rejected.
4593 (b) When a filter file is tested using -bf, no additional Received: header
4594 is added to the test message. After some thought, I decided that this
4597 This change does not affect the value of $received_for. It is still set
4598 after address rewriting, but before local_scan() is called.
4600 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4602 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4603 gave an unhelpful panic error message, and a defer error. I have managed to
4604 change this behaviour so that it now rejects any supplied certificate,
4605 which seems right, as the list of acceptable certificates is empty.
4607 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4608 gave an unhelpful defer error. I have not managed to make this reject any
4609 supplied certificates, but the error message it gives is "no certificate
4610 supplied", which is not helpful.
4612 70. exigrep's output now also includes lines that are not associated with any
4613 message, but which match the given pattern. Implemented by a patch from
4614 Martin Sluka, which also tidied up the Perl a bit.
4616 71. Recipient callout verification, like sender verification, was using <> in
4617 the MAIL FROM command. This isn't really the right thing, since the actual
4618 sender may affect whether the remote host accepts the recipient or not. I
4619 have changed it to use the actual sender in the callout; this means that
4620 the cache record is now keyed on a recipient/sender pair, not just the
4621 recipient address. There doesn't seem to be a real danger of callout loops,
4622 since a callout by the remote host to check the sender would use <>.
4623 [SEE ABOVE: changed after hitting problems.]
4625 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4626 temporary errors. However, in the case of such a code being given after
4627 the end of a data transmission (i.e. after ".") Exim was failing to write
4628 a retry record for the message. (Yes, there was some broken host that was
4629 actually sending 8xx at this point.)
4631 73. An unknown lookup type in a host list could cause Exim to panic-die when
4632 the list was checked. (An example that provoked this was putting <; in the
4633 middle of a list instead of at the start.) If this happened during a DATA
4634 ACL check, a -D file could be left lying around. This kind of configuration
4635 error no longer causes Exim to die; instead it causes a defer errror. The
4636 incident is still logged to the main and panic logs.
4638 74. Buglet left over from Exim 3 conversion. The message "too many messages
4639 in one connection" was written to the rejectlog but not the mainlog, except
4640 when address rewriting (yes!) was being logged.
4642 75. Added write_rejectlog option.
4644 76. When a system filter was run not as root (that is, when system_filter_user
4645 was set), the values of the $n variables were not being returned to the
4646 main process; thus, they were not subsequently available in the $sn
4649 77. Added +return_path_on_delivery log selector.
4651 78. A connection timeout was being treated differently from recipients deferred
4652 when testing hosts_max_try with a message that was older than the host's
4653 retry timeout. (The host should not be counted, thus allowing all hosts to
4654 be tried at least once before bouncing.) This may have been the cause of an
4655 occasionally reported bug whereby a message would remain on the queue
4656 longer than the retry timeout, but would be bounced if a delivery was
4657 forced. I say "may" because I never totally pinned down the problem;
4658 setting up timeout/retry tests is difficult. See also the next item.
4660 79. The ultimate address timeout was not being applied to errors that involved
4661 a combination of host plus message (for example, a timeout on a MAIL
4662 command). When an address resolved to a number of possible hosts, and they
4663 were not all tried for each delivery (e.g. because of hosts_max_try), a
4664 message could remain on the queue longer than the retry timeout.
4666 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4669 81. Fixed an obscure SMTP outgoing bug which required at least the following
4670 conditions: (a) there was another message waiting for the same server;
4671 (b) the server returned 5xx to all RCPT commands in the first message so
4672 that the message was not completed; (c) the server dropped the connection
4673 or gave a negative response to the RSET that Exim sends to abort the
4674 transaction. The observed case was a dropped connection after DATA that had
4675 been sent in pipelining mode. That is, the server had advertised PIPELINING
4676 but was not implementing it correctly. The effect of the bug was incorrect
4677 behaviour, such as trying another host, and this could lead to a crash.
4683 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4684 and daemon.c were passed as pointers to ints; they should have been
4685 pointers to socklen_t variables (which are typically unsigned ints).
4687 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4690 3. Fixed a really odd bug that affected only the testing scheme; patching a
4691 certain fixed string in the binary changed the value of another string that
4692 happened to be identical to the end of the original first string.
4694 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4695 name", it returns that address as the IP address. On some operating
4696 systems (e.g. Solaris), it also passes back the IP address string as the
4697 "host name". However, on others (e.g. Linux), it passes back an empty
4698 string. Exim wasn't checking for this, and was changing the host name to an
4699 empty string, assuming it had been canonicized.
4701 5. Although rare, it is permitted to have more than one PTR record for a given
4702 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4703 all the names associated with an address, because they do in Solaris.
4704 However, it seems that they do not in Linux for data that comes from the
4705 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4706 I found this out when I moved to a new Linux workstation and tried to run
4707 the Exim test suite.
4709 To get round this problem I have changed the code so that it now does its
4710 own call to the DNS to look up PTR records when searching for a host name.
4711 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4712 addresses that are only in /etc/hosts are still found.
4714 This behaviour is, however, controlled by an option called host_lookup_
4715 order, which defaults to "bydns:byaddr". If people want to use the other
4716 order, or indeed, just use one or the other means of lookup, they can
4717 specify it in this variable.
4719 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4720 some operating systems, this comes back from gethostbyaddr() as an empty
4721 string, and this is what Exim used to test for. However, it seems that in
4722 other systems, "." is yielded. Exim now tests for this case too.
4724 7. The values of check_spool_space and check_log_space are now held internally
4725 as a number of kilobytes instead of an absolute number of bytes. If a
4726 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4727 kilobyte. This means that much larger values can be stored.
4729 8. Exim monitor: an attempt to get the action menu when not actually pointing
4730 at a message produces an empty menu entitled "No message selected". This
4731 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4732 no entries in it ("Shell widget menu has zero width and/or height"). So I
4733 have added a single, blank menu entry in this case.
4735 9. Added ${quote_local_part.
4737 10. MIME decoding is now applied to the contents of Subject: header lines when
4740 11. Now that a reference to $sender_host_address automatically causes a reverse
4741 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4742 host lookup before query-style lookups in lists that might use this
4743 variable. This has therefore been abolished, and the "net-" prefix is no
4744 longer necessary for query-style lookups.
4746 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4747 have been a typo for LFLAGS, so it has been changed.
4749 13. The install script calls Exim with "-C /dev/null" in order to find the
4750 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4751 to be output. Howeve, since Exim outputs its version number before the
4752 error, it didn't break the script. It just looked ugly. I fixed this by
4753 always allowing "-C /dev/null" if the caller is root.
4755 14. Ignore overlarge ACL variable number when reading spool file - insurance
4756 against a later release with more variables having written the file.
4758 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4759 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4760 this, as well as the form without the "IPv6" on the front.
4762 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4763 OS/Makefile-Darwin file.
4765 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4766 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4768 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4769 variable. However, some OS use ints for inodes. I've added cast to long int
4770 to get rid of the compiler warning.
4772 19. I had forgotten to lock out "/../" in configuration file names when
4773 ALT_CONFIG_PREFIX was set.
4775 20. Routers used for verification do not need to specify transports. However,
4776 if such a router generated a host list, and callout was configured, Exim
4777 crashed, because it could not find a port number from the (non-existent)
4778 transport. It now assumes port 25 in this circumstance.
4780 21. Added the -t option to exigrep.
4782 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4783 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4784 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4785 them is set, LOOKUP_LSEARCH is forced.
4787 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4789 24. Added sender and host information to the "rejected by local_scan()" log
4790 line; previously there was no indication of these.
4792 25. Added .include_if_exists.
4794 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4795 Linux. It turns out that not all file systems support this. Apparently some
4796 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4797 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4798 not supported on the file descriptor, is now ignored when Exim is trying to
4799 sync a directory. This applies only to Linux.
4801 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4803 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4804 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4807 29. Added support for maildirsize files from supplied patch (modified a bit).
4809 30. The use of :fail: followed by an empty string could lead Exim to respond to
4810 sender verification failures with (e.g.):
4812 550 Verification failed for <xxx>
4813 550 Sender verify failed
4815 where the first response line was missing the '-' that indicates it is not
4816 the final line of the response.
4818 31. The loop for finding the name of the user that called Exim had a hardwired
4819 limit of 10; it now uses the value of finduser_retries, which is used for
4820 all other user lookups.
4822 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4825 33. Exim was neglecting to zero errno before one call of strtol() when
4826 expanding a string and expecting an integer value. On some systems this
4827 resulted in spurious "integer overflow" errors. Also, it was casting the
4828 result into an int without checking.
4830 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4831 did not work. The code looks as if it has *never* worked, though it appears
4832 to have been documented since at least releast 1.62. I have made it work.
4834 35. The "timeout_DNS" error in retry rules, also documented since at least
4835 1.62, also never worked. As it isn't clear exactly what this means, and
4836 clearly it isn't a major issue, I have abolished the feature by treating it
4837 as "timeout", and writing a warning to the main and panic logs.
4839 36. The display of retry rules for -brt wasn't always showing the error code
4842 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4843 timeout_connect_A, timeout_connect_MX.
4845 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4846 to the empty sender.
4848 39. The daemon was not analysing the content of -oX till after it had closed
4849 stderr and disconnected from the controlling terminal. This meant that any
4850 syntax errors were only noted on the panic log, and the return code from
4851 the command was 0. By re-arranging the code a little, I've made the
4852 decoding happen first, so such errors now appear on stderr, and the return
4853 code is 1. However, the actual setting up of the sockets still happens in
4854 the disconnected process, so errors there are still only recorded on the
4857 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4858 connections (as happens on some IP stacks) was logged at start up time as
4859 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4860 it from "IPv6 and IPv4", which means that two separate sockets are being
4863 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4864 says whether AF_INET or AF_INET6 was passed as an argument.
4866 42. Exiwhat output was messed up when time zones were included in log
4869 43. Exiwhat now gives more information about the daemon's listening ports,
4870 and whether -tls-on-connect was used.
4872 44. The "port" option of the smtp transport is now expanded.
4874 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4875 silently ignored. Now an error message is written to the main and panic
4878 46. There's a new ACL modifier called "logwrite" which writes to a log file
4879 as soon as it is encountered.
4881 47. Added $local_user_uid and $local_user_gid at routing time.
4883 48. Exim crashed when trying to verify a sender address that was being
4886 49. Exim was recognizing only a space character after ".include". It now also
4887 recognizes a tab character.
4889 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4890 extracting the relevant information from the specification. The man page no
4891 longer contains scrambled data for the -d option, and I've added a section
4892 at the front about calling Exim under different names.
4894 51. Added "extra_headers" argument to the "mail" command in filter files.
4896 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4899 53. Installed eximstats 1.29.
4901 54. Added transport_filter_timeout as a generic transport option.
4903 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4904 Cc: header lines. This was required by RFC 822, but it not required by RFC
4907 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4908 incoming messages that did not have them. Now it does so only if the
4909 message originates locally, that is, if there is no associated remote host
4910 address. When Resent- header lines are present, this applies to the Resent-
4911 lines rather than the non-Resent- lines.
4913 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4914 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4916 58. Messages for configuration errors now include the name of the main
4917 configuration file - useful now that there may be more than one file in a
4918 list (.included file names were always shown).
4920 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4921 for those rare installations that do not start the daemon as root or run it
4922 setuid root. I've cut out the call to initgroups() if the daemon is not
4925 60. The Exim user and group can now be bound into the binary as text strings
4926 that are looked up at the start of Exim's processing.
4928 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4930 62. Added $mailstore_basename variable.
4932 63. Installed patch to sieve.c from Michael Haardt.
4934 64. When Exim failed to open the panic log after failing to open the main log,
4935 the original message it was trying to log was written to stderr and debug
4936 output, but if they were not available (the usual case in production), it
4937 was lost. Now it is written to syslog before the two lines that record the
4938 failures to open the logs.
4940 65. Users' Exim filters run in subprocesses under the user's uid. It is
4941 possible for a "deliver" command or an alias in a "personal" command to
4942 provoke an address rewrite. If logging of address rewriting is configured,
4943 this fails because the process is not running as root or exim. There may be
4944 a better way of dealing with this, but for the moment (because 4.30 needs
4945 to be released), I have disabled address rewrite logging when running a
4946 filter in a non-root, non-exim process.
4952 1. The buildconfig auxiliary program wasn't quoting the value set for
4953 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4954 not defined. This bug was masked in 4.22 by the effect that was fixed in
4957 2. Some messages that were rejected after a message id was allocated were
4958 shown as "incomplete" by exigrep. It no longer does this for messages that
4959 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4961 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4962 have allow_domain_literals set, the ID did not get logged in the <= line.
4963 Domain literals are now always recognized in Message-ID: header lines.
4965 4. The first argument for a ${extract expansion item is the key name or field
4966 number. Leading and trailing spaces in this item were not being ignored,
4967 causing some misleading effects.
4969 5. When deliver_drop_privilege was set, single queue runner processes started
4970 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4971 same command in the process it spins off) were not dropping privilege.
4973 6. When the daemon running as "exim" started a queue runner, it always
4974 re-executed Exim in the spun-off process. This is a waste of effort when
4975 deliver_drop_privilege is set. The new process now just calls the
4976 queue-runner function directly.
4982 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4985 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4986 ignored. Though the use of -oP was forcing the writing of a pid file, it
4987 was always written to the default place.
4989 3. If the message "no IP address found for host xxxx" is generated during
4990 incoming verification, it is now followed by identification of the incoming
4991 connection (so you can more easily find what provoked it).
4993 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4995 5. Added some features to "harden" Exim a bit more against certain attacks:
4997 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4998 be put in Local/Makefile. This is like the never_users runtime option,
4999 but it cannot be overridden. The default setting is "root".
5001 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5002 prefix string with which any file named in a -C command line option
5005 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5006 is retained for -C and -D only if the caller of Exim is root. Without
5007 it, the exim user may also use -C and -D and retain privilege.
5009 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5010 command line option is disabled.
5012 6. Macro names set by the -D option must start with an upper case letter, just
5013 like macro names defined in the configuration file.
5015 7. Added "dereference=" facility to LDAP.
5017 8. Two instances of the typo "uknown" in the source files are fixed.
5019 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5020 the Configure-Makefile script screwed up while processing it.
5022 10. Incorporated PCRE 4.4.
5024 11. The SMTP synchronization check was not operating right at the start of an
5025 SMTP session. For example, it could not catch a HELO sent before the client
5026 waited for the greeting. There is now a check for outstanding input at the
5027 point when the greeting is written. Because of the duplex, asynchronous
5028 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5029 way, but not yet received, when the check is performed.
5031 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5032 on TCP/IP sockets, because this apparently causes some broken clients to
5035 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5036 unchanged) from the Cygwin maintainer.
5038 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5039 was supported instead of testing for mbx. Effectively a typo.
5041 15. The spa authenticator server code was not checking that the input it
5042 received was valid base64.
5044 16. The debug output line for the "set" modifier in ACLs was not showing the
5045 name of the variable that was being set.
5047 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5049 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5050 lookup on its own. Something else was needed to trigger the lookup. For
5051 example, a match in host_lookup or the need for a host name in a host list.
5052 Now, if $sender_host_name is referenced and the host name has not yet been
5053 looked up, a lookup is performed. If the lookup fails, the variable remains
5054 empty, and $host_lookup_failed is set to "1".
5056 19. Added "eqi" as a case-independent comparison operator.
5058 20. The saslauthd authentication condition could segfault if neither service
5059 nor realm was specified.
5061 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5062 error message that was logged was misleading, and incoming SMTP
5063 connections were dropped. The message is now more accurate, and temporary
5064 errors are given to SMTP connections.
5066 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5067 (except RSET) with a 421 error, until QUIT is received. However, it was
5068 failing to send a response to QUIT.
5070 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5071 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5072 ACL is now run after the helo_try_verify_hosts code.
5074 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5075 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5076 case-independent, so other case variants are also recognized). Apparently
5077 some systems use these upper case variants.
5079 25. If more than two messages were waiting for the same host, and a transport
5080 filter was specified for the transport, Exim sent two messages over the
5081 same TCP/IP connection, and then failed with "socket operation on non-
5082 socket" when it tried to send the third.
5084 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5086 27. The extern definition of crypt16() in expand.c was not being excluded when
5087 the OS had its own crypt16() function.
5089 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5090 as a preferred synonym for return_size_limit, both as an option and as an
5093 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5095 30. Changed the default configuration ACL to relax the local part checking rule
5096 for addresses that are not in any local domains. For these addresses,
5097 slashes and pipe symbols are allowed within local parts, but the sequence
5098 /../ is explicitly forbidden.
5100 31. SPA server authentication was not clearing the challenge buffer before
5103 32. log_message in a "warn" ACL statement was writing to the reject log as
5104 well as to the main log, which contradicts the documentation and doesn't
5105 seem right (because no rejection is happening). So I have stopped it.
5107 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5108 However, I am unable to do any testing of this.
5110 34. Fixed an infelicity in the appendfile transport. When checking directories
5111 for a mailbox, to see if any needed to be created, it was accidentally
5112 using path names with one or more superfluous leading slashes; tracing
5113 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5115 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5116 contents are added to the log line that is written for every discarded
5117 recipient. (Previously a log_message setting was ignored.)
5119 36. The ${quote: operator now quotes the string if it is empty.
5121 37. The install script runs exim in order to find its version number. If for
5122 some reason other than non-existence or emptiness, which it checks, it
5123 could not run './exim', it was installing it with an empty version number,
5124 i.e. as "exim-". This error state is now caught, and the installation is
5127 38. An argument was missing from the function that creates an error message
5128 when Exim fails to connect to the socket for saslauthd authentication.
5129 This could cause Exim to crash, or give a corrupted message.
5131 39. Added isip, isip4, and isip6 to ${if conditions.
5133 40. The ACL variables $acl_xx are now saved with the message, and can be
5134 accessed later in routers, transports, and filters.
5136 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5137 strings in the file are not string-expanded.
5139 42. If a MAIL command specified a SIZE value that was too large to fit into an
5140 int variable, the check against message_size_limit failed. Such values are
5141 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5142 day this will have to be increased, but I don't think I want to be around
5143 when emails are that large.
5150 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5151 iconv() is not standard in FreeBSD.
5153 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5154 IPv6 enabled. The observed symptom was a segmentation fault on return from
5155 the function os_common_find_running_interfaces() in src/os.c.
5157 3. In the check_special_case() function in daemon.c I had used "errno" as an
5158 argument name, which causes warnings on some systems. This was basically a
5159 typo, since it was named "eno" in the comments!
5161 4. The code that waits for the clock to tick (at a resolution of some fraction
5162 of a second) so as to ensure message-id uniqueness was always waiting for
5163 at least one whole tick, when it could have waited for less. [This is
5164 almost certainly not relevant at current processor speeds, where it is
5165 unlikely to ever wait at all. But we try to future-proof.]
5167 5. The function that sleeps for a time interval that includes fractions of a
5168 second contained a race. It did not block SIGALRM between setting the
5169 timer, and suspending (a couple of lines later). If the interval was short
5170 and the sigsuspend() was delayed until after it had expired, the suspension
5171 never ended. On busy systems this could lead to processes getting stuck for
5174 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5175 process, before it forks any delivery processes. The open lookup caching
5176 mechanism meant that the open file or database connection was passed into
5177 the delivery process. The problem was that delivery processes always tidy
5178 up cached lookup data. This could cause a problem for the next delivery
5179 process started by the queue runner, because the external queue runner
5180 process does not know about the closure. So the next delivery process
5181 still has data in the lookup cache. In the case of a file lookup, there was
5182 no problem because closing a file descriptor in a subprocess doesn't affect
5183 the parent. However, if the lookup was caching a connection to a database,
5184 the connection was closed, and the second delivery process was likely to
5185 see errors such as "PGSQL: query failed: server closed the connection
5186 unexpectedly". The problem has been fixed by closing all cached lookups
5187 in a queue runner before running a delivery process.
5189 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5190 seem to have the "const" qualifier which it has on other OS. I've
5193 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5194 *of the same type* (client or server) with the same public name that an
5195 error should be diagnosed.
5197 9. When Exim looked up a host name for an IP address, but failed to find the
5198 original IP address when looking up the host name (a safety check), it
5199 output the message "<ip address> does not match any IP for NULL", which was
5200 confusing, to say the least. The bug was that the host name should have
5201 appeared instead of "NULL".
5203 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5204 user that is built into the binary, and the -C or -D options is used, root
5205 privilege is dropped before the configuration file is read. In addition,
5206 logging is switched to stderr instead of the normal log files. If the
5207 configuration then re-defines the Exim user, the unprivileged environment
5208 is probably not what is expected, so Exim logs a panic warning message (but
5211 However, if deliver_drop_privilege is set, the unprivileged state may well
5212 be exactly what is intended, so the warning has been cut out in that case,
5213 and Exim is allowed to try to write to its normal log files.
5219 1. smtp_return_error_details was not giving details for temporary sender
5220 or receiver verification errors.
5222 2. Diagnose a configuration error if two authenticators have the same public
5225 3. Exim used not to create the message log file for a message until the first
5226 delivery attempt. This could be confusing when incoming messages were held
5227 for policy or load reasons. The message log file is now created at the time
5228 the message is received, and an initial "Received" line is written to it.
5230 4. The automatically generated man page for command line options had a minor
5231 bug that caused no ill effects; however, a more serious problem was that
5232 the procedure for building the man page automatically didn't always
5233 operate. Consequently, release 4.20 contains an out-of-date version. This
5234 shouldn't happen again.
5236 5. When building Exim with embedded Perl support, the script that builds the
5237 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5238 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5240 6. The freeze_tell option was not being used for messages that were frozen on
5241 arrival, either by an ACL or by local_scan().
5243 7. Added the smtp_incomplete_transaction log selector.
5245 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5246 was accepting AUTH without a new EHLO.
5248 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5249 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5252 10. Logging of TCP/IP connections (when configured) now happens in the main
5253 daemon process instead of the child process, so that the TCP/IP connection
5254 count is more accurate (but it can never be perfect).
5256 11. The use of "drop" in a nested ACL was not being handled correctly in the
5257 outer ACL. Now, if condition failure induced by the nested "drop" causes
5258 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5259 or "require"), the connection is dropped.
5261 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5262 that yield "discard" can now be used with an "accept" or a "discard" verb,
5263 but an error is generated for any others (because I can't see a useful way
5264 to define what should happen).
5266 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5267 are now processed in the same way as lines in the Exim configuration file.
5268 In particular, continuation lines are supported.
5270 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5272 15. Added -ti meaning -t -i.
5274 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5275 domains, and warn by logging if others are found.
5277 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5278 returned from GIFCONF when Exim is trying to find the list of interfaces on
5279 a host. The code in os.c has been modified to copy each ifreq to an aligned
5280 structure in all cases.
5282 Also, in some cases, the returned ifreq's were being copied to a 'struct
5283 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5284 means the last couple of bytes of an IPv6 address could be chopped if the
5285 ifreq contained only a normal sockaddr (14 bytes storage).
5287 18. Named domain lists were not supported in the hosts_treat_as_local option.
5288 An entry such as +xxxx was not recognized, and was treated as a literal
5291 19. Ensure that header lines added by a DATA ACL are included in the reject log
5292 if the ACL subsequently rejects the message.
5294 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5295 MD5 (which is deprecated).
5297 21. When testing a filter file using -bf, Exim was writing a message when it
5298 took the sender from a "From " line in the message, but it was not doing so
5299 when it took $return_path from a Return-Path: header line. It now does.
5301 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5302 with a valid header line field name (a series of printing characters
5303 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5305 23. Changed "disc" in the source to "disk" to conform to the documentation and
5306 the book and for uniformity.
5308 24. Ignore Sendmail's -Ooption=value command line item.
5310 25. When execve() failed while trying to run a command in a pipe transport,
5311 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5312 could be confused with a return value of 69 from the command itself. This
5313 has been changed to 127, the value the shell returns if it is asked to run
5314 a non-existent command. The wording for the related log line suggests a
5315 non-existent command as the problem.
5317 26. If received_header_text expands to an empty string, do not add a Received:
5318 header line to the message. (Well, it adds a token one on the spool, but
5319 marks it "old" so that it doesn't get used or transmitted.)
5321 27. Installed eximstats 1.28 (addition of -nt option).
5323 28. There was no check for failure on the call to getsockname() in the daemon
5324 code. This can fail if there is a shortage of resources on the system, with
5325 ENOMEM, for example. A temporary error is now given on failure.
5327 29. Contrary to the C standard, it seems that in some environments, the
5328 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5329 program. Exim now does this explicitly; it affects the formatting of
5330 timestamps using strftime().
5332 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5333 complaints. I've now initialized all the variables, to avoid this.
5335 32. Header lines added by a system filter were not being "seen" during
5336 transport-time rewrites.
5338 33. The info_callback() function passed to OpenSSL is set up with type void
5339 (*)(SSL *, int, int), as described somewhere. However, when calling the
5340 function (actually a macro) that sets it up, the type void(*)() is
5341 expected. I've put in a cast to prevent warnings from picky compilers.
5343 34. If a DNS black list lookup found a CNAME record, but there were no A
5344 records associated with the domain it pointed at, Exim crashed.
5346 35. If a DNS black list lookup returned more than one A record, Exim ignored
5347 all but the first. It now scans all returned addresses if a particular IP
5348 value is being sought. In this situation, the contents of the
5349 $dnslist_value variable are a list of all the addresses, separated by a
5352 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5353 to accept a host name and all its aliases if the forward lookup for any of
5354 them yielded the IP address of the incoming connection. Now it accepts only
5355 those names whose forward lookup yields the correct IP address. Any other
5356 names are discarded. This closes a loophole whereby a rogue DNS
5357 administrator could create reverse DNS records to break through a
5358 wildcarded host restriction in an ACL.
5360 37. If a user filter or a system filter that ran in a subprocess used any of
5361 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5362 the wrong values were passed to the pipe command ($thisaddress had the
5363 value of $0, $0 had the value of $1, etc). This bug was introduced by
5364 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5366 38. Improved the line breaking for long SMTP error messages from ACLs.
5367 Previously, if there was no break point between 40 and 75 characters, Exim
5368 left the rest of the message alone. Two changes have been made: (a) I've
5369 reduced the minimum length to 35 characters; (b) if it can't find a break
5370 point between 35 and 75 characters, it looks ahead and uses the first one
5371 that it finds. This may give the occasional overlong line, but at least the
5372 remaining text gets split now.
5374 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5375 file descriptors might be low, and that setting 1000 would always raise it.
5376 It turns out that in some environments, the limit is already over 1000 and
5377 that lowering it causes trouble. So now Exim takes care not to decrease it.
5379 40. When delivering a message, the value of $return_path is set to $sender_
5380 address at the start of routing (routers may change the value). By an
5381 oversight, this default was not being set up when an address was tested by
5382 -bt or -bv, which affected the outcome if any router or filter referred to
5385 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5386 log only when "message" or "log_message" are set. However, if one of the
5387 conditions was an address verification, or a call to a nested ACL, the
5388 messages generated by the underlying test were being passed through. This
5389 no longer happens. The underlying message is available in $acl_verify_
5390 message for both "message" and "log_message" expansions, so it can be
5391 passed through if needed.
5393 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5394 new expansion $bh_ to give the encoded byte string without charset
5395 translation. Translation happens only if iconv() is available; HAVE_ICONV
5396 indicates this at build time. HEADERS_CHARSET gives the charset to
5397 translate to; headers_charset can change it in the configuration, and
5398 "headers charset" can change it in an individual filter file.
5400 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5401 that creates RFC 2047 encoded "words" labels them as that charset instead
5402 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5403 expansion operator; (ii) when Exim creates a From: line for a local
5404 message; (iii) when a header line is rewritten to include a "phrase" part.
5406 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5407 buggy, causing it to skip the first lines of messages whose message ID
5408 ended in 'D'. This would not have bitten before Exim release 4.14, because
5409 message IDs were unlikely to end in 'D' before then. The effect was to have
5410 incorrect size information for certain domains.
5412 45. #include "config.h" was missing at the start of the crypt16.c module. This
5413 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5416 46. If there was a timeout during a "random" callout check, Exim treated it as
5417 a failure of the random address, and carried on sending RSET and the real
5418 address. If the delay was just some slowness somewhere, the response to the
5419 original RCPT would be taken as a response to RSET and so on, causing
5420 mayhem of various kinds.
5422 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5423 when I implemented it. It didn't allow for the fact that some option values
5424 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5425 the right test for positive values.
5427 48. Domain names in DNS records are case-independent. Exim always looks them up
5428 in lower case. Some resolvers return domain names in exactly the case they
5429 appear in the zone file, that is, they may contain uppercase letters. Not
5430 all resolvers do this - some return always lower case. Exim was treating a
5431 change of case by a resolver as a change of domain, similar to a widening
5432 of a domain abbreviation. This triggered its re-routing code and so it was
5433 trying to route what was effectively the same domain again. This normally
5434 caused routing to fail (because the router wouldn't handle the domain
5435 twice). Now Exim checks for this case specially, and just changes the
5436 casing of the domain that it ultimately uses when it transmits the message
5439 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5442 50. If a filter generated a file delivery with a non-absolute name (possible if
5443 no home directory exists for the router), the forbid_file option was not
5446 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5447 the existing equality matching.
5449 52. Exim was using ints instead of ino_t variables in some places where it was
5450 dealing with inode numbers.
5452 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5453 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5454 called TMPDIR, and if it finds it is different, it changes its value.
5456 54. The smtp_printf() function is now made available to local_scan() so
5457 additional output lines can be written before returning. There is also an
5458 smtp_fflush() function to enable the detection of a dropped connection.
5459 The variables smtp_input and smtp_batched_input are exported to
5462 55. Changed the default runtime configuration: the message "Unknown user"
5463 has been removed from the ACL, and instead placed on the localuser router,
5464 using the cannot_route_message feature. This means that any verification
5465 failures that generate their own messages won't get overridden. Similarly,
5466 the "Unrouteable address" message that was in the ACL for unverifiable
5467 relay addresses has also been removed.
5469 56. Added hosts_avoid_esmtp to the smtp transport.
5471 57. The exicyclog script was not checking for the esoteric option
5472 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5473 will work only if exicyclog is run under the appropriate euid.
5475 58. Following a discussion on the list, the rules by which Exim recognises line
5476 endings on incoming messages have been changed. The -dropcr and drop_cr
5477 options are now no-ops, retained only for backwards compatibility. The
5478 following line terminators are recognized: LF CRLF CR. However, special
5479 processing applies to CR:
5481 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5482 nor a local message in the state where . is a terminator.
5484 (ii) If a bare CR is encountered in a header line, an extra space is added
5485 after the line terminator so as not to end the header. The reasoning
5486 behind this is that bare CRs in header lines are most likely either
5487 to be mistakes, or people trying to play silly games.
5489 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5490 was being incorrectly given as 18 bytes larger than it should have been.
5491 This is a VOB (very old bug).
5493 60. This may never have affected anything current, but just in case it has:
5494 When the local host is found other than at the start of a list of hosts,
5495 the local host, those with the same MX, and any that follow, are discarded.
5496 When the list in question was part of a longer list of hosts, the following
5497 hosts (not currently being processed) were also being discarded. This no
5498 longer happens. I'm not sure if this situation could ever has previously
5501 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5504 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5505 header. This is something that is recommended in a new Internet Draft, and
5506 is something that is documented as being done by Sendmail. There are two
5507 possible values. For messages generated by the autoreply transport, Exim
5510 Auto-Submitted: auto-replied
5512 whereas for all other generated messages (e.g. bounces) it adds
5514 Auto-Submitted: auto-generated
5516 63. The "personal" condition in filters now includes a test for the
5517 Auto-Submitted: header. If it contains the string "auto-" the message it
5518 not considered personal.
5520 64. Added rcpt_include_affixes as a generic transport option.
5522 65. Added queue_only_override (default true).
5524 66. Added the syslog_duplication option.
5526 67. If what should have been the first header line of a message consisted of
5527 a space followed by a colon, Exim was mis-interpreting it as a header line.
5528 It isn't of course - it is syntactically invalid and should therefore be
5529 treated as the start of the message body. The misbehaviour could have
5530 caused a number of strange effects, including loss of data in subsequent
5531 header lines, and spool format errors.
5533 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5534 client host had authenticated. This control can now be exercised by an ACL
5535 for more flexibility.
5537 69. By default, callouts do not happen when testing with -bh. There is now a
5538 variant, -bhc, which does actually run the callout code, including
5539 consulting and updating the callout cache.
5541 70. Added support for saslauthd authentication, courtesy of Alexander
5544 71. If statvfs() failed on the spool or log directories while checking their
5545 size for availability, Exim confusingly gave the error "space shortage".
5546 Furthermore, in debugging mode it crashed with a floating point exception.
5547 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5548 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5549 serious problem, Exim now writes to the main and panic logs when this
5550 happens, with details of the failure. It then refuses to accept the
5551 incoming message, giving the message "spool directory problem" or "log
5552 directory problem" with a 421 code for SMTP messages.
5554 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5555 0, 1, and 2 exist, because some OS complain for execs without them (see
5556 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5557 descriptors. However, the code omitted to check that the open succeeded,
5558 causing mysterious errors if for some reason the permissions on /dev/null
5559 got screwed. Now Exim writes a message to the main and panic logs, and
5560 bombs out if it can't open /dev/null.
5562 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5563 interact so that it is all more flexible. It is supposed to remain
5564 backwards compatible. Also added extra_local_interfaces.
5566 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5567 to bomb out with an assertion failure - to the client this appears as a
5568 connection drop. This problem occurs in the part of the code that was taken
5569 from the Samba project. Fortunately, the assertion is in a very simple
5570 function, so I have fixed this by reproducing the function inline in the
5571 one place where it is called, and arranging for authentication to fail
5572 instead of killing the process with assert().
5574 75. The SPA client code was not working when the server requested OEM rather
5575 than Unicode encoding.
5577 76. Added code to make require_files with a specific uid setting more usable in
5578 the case where statting the file as root fails - usually a non-root-mounted
5579 NFS file system. When this happens and the failure is EACCES, Exim now
5580 forks a subprocess and does the per-uid checking as the relevant uid.
5582 77. Added process_log_path.
5584 78. If log_file_path was not explicitly set, a setting of check_log_space or
5585 check_log_inodes was ignored.
5587 79. If a space check for the spool or log partitions fails, the incident is now
5588 logged. Of course, in the latter case the data may get lost...
5590 80. Added the %p formatting code to string_format() so that it can be used to
5591 print addresses in debug_print(). Adjusted all the address printing in the
5592 debugging in store.c to use %p rather than %d.
5594 81. There was a concern that a line of code in smtp_in.c could overflow a
5595 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5596 initially expressed, the concern was not well-founded, because trailing
5597 spaces are removed early. However, if the trailing spaces were followed by
5598 a NULL, they did not get removed, so the overflow was possible. Two fixes
5601 (a) I re-wrote the offending code in a cleaner fashion.
5602 (b) If an incoming SMTP command contains a NULL character, it is rejected
5605 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5606 runs initgroups(), so that if the Exim user is in any additional groups,
5607 they will be used during message reception.
5613 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob