1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
23 PP/01 Handle short writes when writing local log-files.
24 In practice, only affects FreeBSD (8 onwards).
25 Bugzilla 1053, with thanks to Dmitry Isaikin.
27 NM/01 Bugzilla 949 - Documentation tweak
29 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
32 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
34 PP/02 Implemented gsasl authenticator.
36 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
38 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
39 `pkg-config foo` for cflags/libs.
41 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
42 with rest of GSASL and with heimdal_gssapi.
44 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
45 `pkg-config foo` for cflags/libs for the TLS implementation.
47 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
48 properties get this fed in as external SSF. A number of robustness
49 and debugging improvements to the cyrus_sasl authenticator.
51 PP/08 cyrus_sasl server now expands the server_realm option.
53 PP/09 Bugzilla 1214 - Log authentication information in reject log.
54 Patch by Jeremy Harris.
56 PP/10 Added dbmjz lookup type.
58 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
60 PP/12 MAIL args handles TAB as well as SP, for better interop with
61 non-compliant senders.
62 Analysis and variant patch by Todd Lyons.
64 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
65 Bug report from Lars Müller <lars@samba.org> (via SUSE),
66 Patch from Dirk Mueller <dmueller@suse.com>
68 PP/13 tls_peerdn now print-escaped for spool files.
69 Observed some $tls_peerdn in wild which contained \n, which resulted
70 in spool file corruption.
72 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
73 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
74 or write after TLS renegotiation, which otherwise led to messages
77 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
78 as a tracking header (ie: a signed header comes before the signature).
79 Patch from Wolfgang Breyha.
81 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
82 comma-sep list; embedded commas doubled.
84 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
86 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
88 Report and patch from Dmitry Banschikov.
90 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
91 Removed SSL_clear() after SSL_new() which led to protocol negotiation
92 failures. We appear to now support TLS1.1+ with Exim.
94 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
95 lets Exim select keys and certificates based upon TLS SNI from client.
96 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
97 before an outbound SMTP session. New log_selector, +tls_sni.
99 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
100 NULL dereference. Report and patch from Alun Jones.
102 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
103 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
104 Not seeing resolver debug output on NetBSD, but suspect this is a
105 resolver implementation change.
107 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
108 Left warnings. Added "eximon gdb" invocation mode.
110 PP/21 Defaulting "accept_8bitmime" to true, not false.
112 PP/22 Added -bw for inetd wait mode support.
114 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
115 locate the relevant includes and libraries. Made this the default.
117 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
118 Bugzilla 1246, report and most of solution from Tomasz Kusy.
120 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
121 This may cause build issues on older platforms.
123 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
124 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
125 gnutls_require_mac & gnutls_require_protocols (no longer supported).
126 Added SNI support via GnuTLS too.
127 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
129 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
131 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
132 Applied second patch from Janne, implementing suggestion to default
133 multiple-strings-in-record handling to match SPF spec.
135 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
137 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
138 read-only, out of scope).
139 Patch from Wolfgang Breyha, report from Stuart Northfield.
141 PP/29 Fix three issues highlighted by clang analyser static analysis.
142 Only crash-plausible issue would require the Cambridge-specific
143 iplookup router and a misconfiguration.
144 Report from Marcin Mirosław.
146 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
148 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
149 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
150 As part of this, removing so much warning spew let me fix some minor
151 real issues in debug logging.
153 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
154 assignment on my part. Fixed.
156 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
157 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
158 Janne Snabb (who went above and beyond: thank you).
160 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
161 string otherwise requires a connection and a bunch more work and it's
162 relatively easy to get wrong. Should also expose TLS library linkage
165 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
166 64-bit ${eval} (JH/03).
168 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
169 GNU libc to support some of the 64-bit stuff, should not lead to
170 conflicts. Defined before os.h is pulled in, so if a given platform
171 needs to override this, it can.
173 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
174 protection layer was required, which is not implemented.
175 Bugzilla 1254, patch from Wolfgang Breyha.
177 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
178 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
179 tls_dhparam take prime identifiers. Also unbreak combination of
180 OpenSSL+DH_params+TLSSNI.
182 PP/39 Disable SSLv2 by default in OpenSSL support.
188 PP/01 Solaris build fix for Oracle's LDAP libraries.
189 Bugzilla 1109, patch from Stephen Usher.
191 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
193 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
196 TF/02 Fix a couple more cases where we did not log the error message
197 when unlink() failed. See also change 4.74-TF/03.
199 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
200 lock up or crash if it happened to be inside a call to libc when it
201 got a SIGUSR1 from exiwhat.
203 The SIGUSR1 handler appends the current process status to the process
204 log which is later printed by exiwhat. It used to use the general
205 purpose logging code to do this, but several functions it calls are
206 not safe for signals.
208 The new output code in the SIGUSR1 handler is specific to the process
209 log, and simple enough that it's easy to inspect for signal safety.
210 Removing some special cases also simplifies the general logging code.
211 Removing the spurious timestamps from the process log simplifies
214 TF/04 Improved ratelimit ACL condition.
216 The /noupdate option has been deprecated in favour of /readonly which
217 has clearer semantics. The /leaky, /strict, and /readonly update modes
218 are mutually exclusive. The update mode is no longer included in the
219 database key; it just determines when the database is updated. (This
220 means that when you upgrde Exim will forget old rate measurements.)
222 Exim now checks that the per_* options are used with an update mode that
223 makes sense for the current ACL. For example, when Exim is processing a
224 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
225 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
226 must specify per_mail/readonly. If you omit the update mode it defaults to
227 /leaky where that makes sense (as before) or /readonly where required.
229 The /noupdate option is now undocumented but still supported for
230 backwards compatibility. It is equivalent to /readonly except that in
231 ACLs where /readonly is required you may specify /leaky/noupdate or
232 /strict/noupdate which are treated the same as /readonly.
234 A useful new feature is the /count= option. This is a generalization
235 of the per_byte option, so that you can measure the throughput of other
236 aggregate values. For example, the per_byte option is now equivalent
237 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
239 The per_rcpt option has been generalized using the /count= mechanism
240 (though it's more complicated than the per_byte equivalence). When it is
241 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
242 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
243 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
244 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
245 Note that using per_rcpt with a non-readonly update mode in more than
246 one ACL will cause the recipients to be double-counted. (The per_mail
247 and per_byte options don't have this problem.)
249 The handling of very low rates has changed slightly. If the computed rate
250 is less than the event's count (usually one) then this event is the first
251 after a long gap. In this case the rate is set to the same as this event's
252 count, so that the first message of a spam run is counted properly.
254 The major new feature is a mechanism for counting the rate of unique
255 events. The new per_addr option counts the number of different
256 recipients that someone has sent messages to in the last time period. It
257 behaves like per_rcpt if all the recipient addresses are different, but
258 duplicate recipient addresses do not increase the measured rate. Like
259 the /count= option this is a general mechanism, so the per_addr option
260 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
261 example, measure the rate that a client uses different sender addresses
262 with the options per_mail/unique=$sender_address. There are further
263 details in the main documentation.
265 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
267 TF/06 Removed a few PCRE remnants.
269 TF/07 Automatically extract Exim's version number from tags in the git
270 repository when doing development or release builds.
272 PP/02 Raise smtp_cmd_buffer_size to 16kB.
273 Bugzilla 879. Patch from Paul Fisher.
275 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
276 Heavily based on revision 40f9a89a from Simon Arlott's tree.
279 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
281 PP/05 Variable $av_failed, true if the AV scanner deferred.
282 Bugzilla 1078. Patch from John Horne.
284 PP/06 Stop make process more reliably on build failure.
285 Bugzilla 1087. Patch from Heiko Schlittermann.
287 PP/07 Make maildir_use_size_file an _expandable_ boolean.
288 Bugzilla 1089. Patch from Heiko Schlittermann.
290 PP/08 Handle ${run} returning more data than OS pipe buffer size.
291 Bugzilla 1131. Patch from Holger Weiß.
293 PP/09 Handle IPv6 addresses with SPF.
294 Bugzilla 860. Patch from Wolfgang Breyha.
296 PP/10 GnuTLS: support TLS 1.2 & 1.1.
298 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
301 PP/11 match_* no longer expand right-hand-side by default.
302 New compile-time build option, EXPAND_LISTMATCH_RHS.
303 New expansion conditions, "inlist", "inlisti".
305 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
307 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
309 PP/14 fix log_write() format string regression from TF/03.
310 Bugzilla 1152. Patch from Dmitry Isaikin.
316 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
318 PP/02 Harmonised TLS library version reporting; only show if debugging.
319 Layout now matches that introduced for other libraries in 4.74 PP/03.
321 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
323 PP/04 New "dns_use_edns0" global option.
325 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
328 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
329 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
331 TK/01 Updated PolarSSL code to 0.14.2.
332 Bugzilla 1097. Patch from Andreas Metzler.
334 PP/07 Catch divide-by-zero in ${eval:...}.
337 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
340 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
341 format-string attack -- SECURITY: remote arbitrary code execution.
343 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
344 time unintentionally subject to list matching rules, letting the header
345 cause arbitrary Exim lookups (of items which can occur in lists, *not*
346 arbitrary string expansion). This allowed for information disclosure.
348 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
349 INT_MIN/-1 -- value coerced to INT_MAX.
355 NM/01 Workround for PCRE version dependency in version reporting
358 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
359 This fixes portability to compilers other than gcc, notably
360 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
362 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
363 makefiles for portability to HP-UX and POSIX correctness.
365 PP/01 Permit LOOKUP_foo enabling on the make command-line.
366 Also via indented variable definition in the Makefile.
367 (Debugging by Oliver Heesakkers).
369 PP/02 Restore caching of spamd results with expanded spamd_address.
370 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
372 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
373 Improves build reliability. Fix from: Frank Elsner
375 NM/02 Fix wide character breakage in the rfc2047 coding
376 Fixes bug 1064. Patch from Andrey N. Oktyabrski
378 NM/03 Allow underscore in dnslist lookups
379 Fixes bug 1026. Patch from Graeme Fowler
381 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
382 Code patches from Adam Ciarcinski of NetBSD.
384 NM/04 Fixed exiqgrep to cope with mailq missing size issue
387 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
388 is logged, to avoid truncation. Patch from John Horne.
390 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
391 Patch from Jakob Hirsch.
393 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
394 SQL string expansion failure details.
395 Patch from Andrey Oktyabrski.
397 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
398 Patch from Simon Arlott.
400 PP/09 New lookups functionality failed to compile on old gcc which rejects
401 extern declarations in function scope.
402 Patch from Oliver Fleischmann
404 PP/10 Use sig_atomic_t for flags set from signal handlers.
405 Check getgroups() return and improve debugging.
406 Fixed developed for diagnosis in bug 927 (which turned out to be
409 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
410 Patch from Mark Zealey.
412 PP/12 Bugzilla 1056: Improved spamd server selection.
413 Patch from Mark Zealey.
415 PP/13 Bugzilla 1086: Deal with maildir quota file races.
416 Based on patch from Heiko Schlittermann.
418 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
419 Patch from Uwe Doering, sign-off by Michael Haardt.
421 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
422 variable declaration deep within a block. Bug and patch from
425 PP/15 lookups-Makefile IRIX compatibilty coercion.
427 PP/16 Make DISABLE_DKIM build knob functional.
429 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
430 Patch by Simon Arlott
432 TF/03 Fix valgrind.h portability to C89 compilers that do not support
433 variable argument macros. Our copy now differs from upstream.
439 TF/01 Failure to get a lock on a hints database can have serious
440 consequences so log it to the panic log.
442 TF/02 Log LMTP confirmation messages in the same way as SMTP,
443 controlled using the smtp_confirmation log selector.
445 TF/03 Include the error message when we fail to unlink a spool file.
447 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
448 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
449 for maintaining out-of-tree patches for some time.
451 PP/01 Bugzilla 139: Documentation and portability issues.
452 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
453 Handle per-OS dynamic-module compilation flags.
455 PP/02 Let /dev/null have normal permissions.
456 The 4.73 fixes were a little too stringent and complained about the
457 permissions on /dev/null. Exempt it from some checks.
458 Reported by Andreas M. Kirchwitz.
460 PP/03 Report version information for many libraries, including
461 Exim version information for dynamically loaded libraries. Created
462 version.h, now support a version extension string for distributors
463 who patch heavily. Dynamic module ABI change.
465 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
466 privilege escalation vulnerability whereby the Exim run-time user
467 can cause root to append content of the attacker's choosing to
470 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
473 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
474 If dropping privileges for untrusted macros, we disabled normal logging
475 on the basis that it would fail; for the Exim run-time user, this is not
476 the case, and it resulted in successful deliveries going unlogged.
477 Fixed. Reported by Andreas Metzler.
483 PP/01 Date: & Message-Id: revert to normally being appended to a message,
484 only prepend for the Resent-* case. Fixes regression introduced in
485 Exim 4.70 by NM/22 for Bugzilla 607.
487 PP/02 Include check_rfc2047_length in configure.default because we're seeing
488 increasing numbers of administrators be bitten by this.
490 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
492 PP/03 Bugzilla 994: added openssl_options main configuration option.
494 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
496 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
498 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
500 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
501 without a peer certificate, leading to a segfault because of an
502 assumption that peers always have certificates. Be a little more
503 paranoid. Problem reported by Martin Tscholak.
505 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
506 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
507 NB: ClamAV planning to remove STREAM in "middle of 2010".
508 CL also introduces -bmalware, various -d+acl logging additions and
509 more caution in buffer sizes.
511 PP/09 Implemented reverse_ip expansion operator.
513 PP/10 Bugzilla 937: provide a "debug" ACL control.
515 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
517 PP/12 Bugzilla 973: Implement --version.
519 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
521 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
523 PP/15 Bugzilla 816: support multiple condition rules on Routers.
525 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
526 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
527 ignore trailing whitespace.
529 JJ/02 prevent non-panic DKIM error from being sent to paniclog
531 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
534 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
535 Notification from Dr Andrew Aitchison.
537 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
538 ExtendedDetectionInfo response format.
539 Notification from John Horne.
541 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
544 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
545 XSL and documented dependency on system catalogs, with examples of how
548 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
551 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
552 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
553 configuration file which is writeable by the Exim user or group.
555 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
556 of configuration files to cover files specified with the -C option if
557 they are going to be used with root privileges, not just the default
560 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
561 option (effectively making it always true).
563 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
564 files to be used while preserving root privileges.
566 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
567 that rogue child processes cannot use them.
569 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
570 run-time user, instead of root.
572 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
573 Exim run-time user without dropping privileges.
575 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
576 result string, instead of calling string_vformat() twice with the same
579 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
580 for other users. Others should always drop root privileges if they use
581 -C on the command line, even for a whitelisted configure file.
583 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
585 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
591 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
592 $data_path, and $header_path variables; fixed documentation bugs and
595 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
596 exipick to access non-standard spools, including the "frozen" queue
599 NM/01 Bugzilla 965: Support mysql stored procedures.
600 Patch from Alain Williams
602 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
604 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
605 Patch from Andreas Metzler
607 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
608 Patch from Kirill Miazine
610 NM/05 Bugzilla 671: Added umask to procmail example.
612 JJ/03 installed exipick 20100323.0, fixing doc bug
614 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
615 directory. Notification and patch from Dan Rosenberg.
617 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
619 TK/02 Improve log output when DKIM signing operation fails.
621 MH/01 Treat the transport option dkim_domain as a colon separated
622 list, not as a single string, and sign the message with each element,
623 omitting multiple occurences of the same signer.
625 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
626 Bugzilla 985, 986. Patch by Simon Arlott
628 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
629 Patch by Simon Arlott
631 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
632 MBX locking. Notification from Dan Rosenberg.
638 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
640 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
642 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
644 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
646 NM/04 Bugzilla 915: Flag broken perl installation during build.
652 TK/01 Added patch by Johannes Berg that expands the main option
653 "spamd_address" if it starts with a dollar sign.
655 TK/02 Write list of recipients to X-Envelope-Sender header when building
656 the mbox-format spool file for content scanning (suggested by Jakob
659 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
660 (http://www.dcc-servers.net/) support via dccifd. Activated by
661 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
663 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
664 by Mark Daniel Reidel <mr@df.eu>.
666 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
667 When building exim an external PCRE library is now needed -
668 PCRE is a system library on the majority of modern systems.
669 See entry on PCRE_LIBS in EDITME file.
671 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
672 conversation. Added nologin parameter to request.
673 Patch contributed by Kirill Miazine.
675 TF/01 Do not log submission mode rewrites if they do not change the address.
677 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
679 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
680 log files in place. Contributed by Roberto Lima.
682 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
684 TF/03 Bugzilla 615: When checking the local_parts router precondition
685 after a local_part_suffix or local_part_prefix option, Exim now
686 does not use the address's named list lookup cache, since this
687 contains cached lookups for the whole local part.
689 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
690 Robert Millan. Documentation is in experimental-spec.txt.
692 TF/04 Bugzilla 668: Fix parallel build (make -j).
694 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
696 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
697 Patch provided by Jan Srzednicki.
699 TF/05 Leading white space used to be stripped from $spam_report which
700 wrecked the formatting. Now it is preserved.
702 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
703 that they are available at delivery time.
705 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
707 TF/08 TLS error reporting now respects the incoming_interface and
708 incoming_port log selectors.
710 TF/09 Produce a more useful error message if an SMTP transport's hosts
711 setting expands to an empty string.
713 NM/06 Bugzilla 744: EXPN did not work under TLS.
714 Patch provided by Phil Pennock.
716 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
717 Patch provided by Richard Godbee.
719 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
720 acl_smtp_notquit, added index entry.
722 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
723 Patch provided by Eugene Bujak.
725 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
726 accept(). Patch provided by Maxim Dounin.
728 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
729 Patch provided by Phil Pennock.
731 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
733 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
734 Patch provided by Brad "anomie" Jorsch.
736 NM/14 Bugzilla 622: Added timeout setting to transport filter.
737 Patch provided by Dean Brooks.
739 TK/05 Add native DKIM support (does not depend on external libraries).
741 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
742 Patch provided by Graeme Fowler.
744 NM/16 Bugzilla 851: Documentation example syntax fix.
746 NM/17 Changed NOTICE file to remove references to embedded PCRE.
748 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
751 NM/19 Bugzilla 745: TLS version reporting.
752 Patch provided by Phil Pennock.
754 NM/20 Bugzilla 167: bool: condition support.
755 Patch provided by Phil Pennock.
757 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
758 clients. Patch provided by Phil Pennock.
760 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
761 Patch provided by Brad "anomie" Jorsch.
763 NM/23 Bugzilla 687: Fix misparses in eximstats.
764 Patch provided by Heiko Schlittermann.
766 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
767 Patch provided by Heiko Schlittermann.
769 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
770 plus update to original patch.
772 NM/26 Bugzilla 799: Documentation correction for ratelimit.
774 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
775 Patch provided by David Brownlee.
777 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
779 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
781 NM/30 Bugzilla 888: TLS documentation bugfixes.
783 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
785 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
786 Unlike the original bugzilla I have changed all shell scripts in src tree.
788 NM/33 Bugzilla 898: Transport filter timeout fix.
789 Patch by Todd Rinaldo.
791 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
792 Patch by Serge Demonchaux.
794 NM/35 Bugzilla 39: Base64 decode bug fixes.
795 Patch by Jakob Hirsch.
797 NM/36 Bugzilla 909: Correct connect() call in dcc code.
799 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
801 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
803 NM/39 Bugzilla 911: Fixed MakeLinks build script.
809 TK/01 Add preliminary DKIM support. Currently requires a forked version of
810 ALT-N's libdkim that I have put here:
811 http://duncanthrax.net/exim-experimental/
813 Note to Michael Haardt: I had to rename some vars in sieve.c. They
814 were called 'true' and it seems that C99 defines that as a reserved
815 keyword to be used with 'bool' variable types. That means you could
816 not include C99-style headers which use bools without triggering
817 build errors in sieve.c.
819 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
820 as mailq or other aliases. Changed the --help handling significantly
821 to do whats expected. exim_usage() emits usage/help information.
823 SC/01 Added the -bylocaldomain option to eximstats.
825 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
827 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
829 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
835 PH/01 Another patch from the Sieve maintainer.
837 PH/02 When an IPv6 address is converted to a string for single-key lookup
838 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
839 dots are used instead of colons so that keys in lsearch files need not
840 contain colons. This was done some time before quoting was made available
841 in lsearch files. However, iplsearch files do require colons in IPv6 keys
842 (notated using the quote facility) so as to distinguish them from IPv4
843 keys. This meant that lookups for IP addresses in host lists did not work
844 for iplsearch lookups.
846 This has been fixed by arranging for IPv6 addresses to be expressed with
847 colons if the lookup type is iplsearch. This is not incompatible, because
848 previously such lookups could never work.
850 The situation is now rather anomolous, since one *can* have colons in
851 ordinary lsearch keys. However, making the change in all cases is
852 incompatible and would probably break a number of configurations.
854 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
857 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
858 conversion specification without a maximum field width, thereby enabling
859 a rogue spamd server to cause a buffer overflow. While nobody in their
860 right mind would setup Exim to query an untrusted spamd server, an
861 attacker that gains access to a server running spamd could potentially
862 exploit this vulnerability to run arbitrary code as the Exim user.
864 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
865 $primary_hostname instead of what libspf2 thinks the hosts name is.
867 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
868 a directory entry by the name of the lookup key. Previously, if a
869 symlink pointed to a non-existing file or a file in a directory that
870 Exim lacked permissions to read, a lookup for a key matching that
871 symlink would fail. Now it is enough that a matching directory entry
872 exists, symlink or not. (Bugzilla 503.)
874 PH/03 The body_linecount and body_zerocount variables are now exported in the
877 PH/04 Added the $dnslist_matched variable.
879 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
880 This means they are set thereafter only if the connection becomes
883 PH/06 Added the client_condition to authenticators so that some can be skipped
884 by clients under certain conditions.
886 PH/07 The error message for a badly-placed control=no_multiline_responses left
887 "_responses" off the end of the name.
889 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
891 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
892 (without spaces) instead of just copying the configuration text.
894 PH/10 Added the /noupdate option to the ratelimit ACL condition.
896 PH/11 Added $max_received_linelength.
898 PH/12 Added +ignore_defer and +include_defer to host lists.
900 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
901 way in which PCRE > 7.0 is built.
903 PH/14 Implemented queue_only_load_latch.
905 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
906 MAIL command. The effect was to mangle the value on 64-bit systems.
908 PH/16 Another patch from the Sieve maintainer.
910 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
912 PH/18 If a system quota error occurred while trying to create the file for
913 a maildir delivery, the message "Mailbox is full" was not appended to the
914 bounce if the delivery eventually timed out. Change 4.67/27 below applied
915 only to a quota excession during the actual writing of the file.
917 PH/19 It seems that peer DN values may contain newlines (and other non-printing
918 characters?) which causes problems in log lines. The DN values are now
919 passed through string_printing() before being added to log lines.
921 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
922 and InterBase are left for another time.)
924 PH/21 Added message_body_newlines option.
926 PH/22 Guard against possible overflow in moan_check_errorcopy().
928 PH/23 POSIX allows open() to be a macro; guard against that.
930 PH/24 If the recipient of an error message contained an @ in the local part
931 (suitably quoted, of course), incorrect values were put in $domain and
932 $local_part during the evaluation of errors_copy.
938 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
939 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
942 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
943 issue a MAIL command.
945 PH/02 In an ACL statement such as
947 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
949 if a client was not listed at all, or was listed with a value other than
950 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
951 the condition was not true (as it should be), so access was not denied.
952 The bug was that the ! inversion was incorrectly passed on to the second
953 item. This has been fixed.
955 PH/03 Added additional dnslists conditions == and =& which are different from
956 = and & when the dns lookup returns more than one IP address.
958 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
959 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
961 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
962 FSYNC, which compiles an option called disable_fsync that allows for
963 bypassing fsync(). The documentation is heavily laced with warnings.
965 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
967 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
968 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
969 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
970 including adding "make clean"; (3) Added -fPIC when compiling the test
971 dynamically loaded module, to get rid of a warning.
973 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
974 message fails, move_frozen_messages = true and ignore_bounce_errors_after
975 = 0s. The bug is otherwise harmless.
977 PH/07 There was a bug in the dovecot authenticator such that the value of
978 $auth1 could be overwritten, and so not correctly preserved, after a
979 successful authentication. This usually meant that the value preserved by
980 the server_setid option was incorrect.
982 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
984 PH/09 Installed PCRE release 7.0.
986 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
987 run for batched SMTP input. It is now run at the start of every message
988 in the batch. While fixing this I discovered that the process information
989 (output by running exiwhat) was not always getting set for -bs and -bS
990 input. This is fixed, and it now also says "batched" for BSMTP.
992 PH/11 Added control=no_pipelining.
994 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
995 patch, slightly modified), and move the expansion of helo_data till after
996 the connection is made in the smtp transport (so it can use these
999 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1001 PH/14 Added log_selector = +pid.
1003 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1005 PH/16 Add ${if forany and ${if forall.
1007 PH/17 Added dsn_from option to vary the From: line in DSNs.
1009 PH/18 Flush SMTP output before performing a callout, unless control =
1010 no_callout_flush is set.
1012 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1013 was true (the default) a successful delivery failed to delete the retry
1014 item, thus causing premature timeout of the address. The bug is now
1017 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1019 PH/21 Long custom messages for fakedefer and fakereject are now split up
1020 into multiline reponses in the same way that messages for "deny" and
1021 other ACL rejections are.
1023 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1024 with slight modification.
1026 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1027 draft, changing the syntax and factoring some duplicate code".
1029 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1030 for deliveries of the second and subsequent messages over the same SMTP
1033 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1034 ${reduce, with only minor "tidies".
1036 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1038 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1039 expansion side effects.
1041 PH/27 When a message times out after an over-quota error from an Exim-imposed
1042 quota, the bounce message says "mailbox is full". This message was not
1043 being given when it was a system quota that was exceeded. It now should
1046 MH/03 Made $recipients available in local_scan(). local_scan() already has
1047 better access to the recipient list through recipients_list[], but
1048 $recipients can be useful in postmaster-provided expansion strings.
1050 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1051 in the case of a MAIL command with additional options following the
1052 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1053 were accidentally chopped off.
1055 PH/29 SMTP synchronization checks are implemented when a command is read -
1056 there is a check that no more input is waiting when there shouldn't be
1057 any. However, for some commands, a delay in an ACL can mean that it is
1058 some time before the response is written. In this time, more input might
1059 arrive, invalidly. So now there are extra checks after an ACL has run for
1060 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1061 pipelining has not been advertised.
1063 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1065 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1066 returned in the SMTP response when smtp_return_error_details was false.
1067 This has been fixed.
1069 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1070 instead of the C I/O that was originally supplied, because problems were
1071 reported on Solaris.
1073 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1074 Exim which did not show up earlier: it was assuming that a call to
1075 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1076 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1077 was a macro that became an assignment, so it seemed to work. This has
1078 changed to a proper function call with a void return, hence the compile
1079 error. Exim's code has been fixed.
1081 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1084 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1086 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1088 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1089 because it turns out to be too big or there is a timeout) there is no
1090 "Completed" line in the log. When some messages of this type were
1091 selected by exigrep, they were listed as "not completed". Others were
1092 picked up by some special patterns. I have improved the selection
1093 criteria to be more general.
1095 PH/38 The host_find_failed option in the manualroute router can now be set
1096 to "ignore", to completely ignore a host whose IP address cannot be
1097 found. If all hosts are ignored, the behaviour is controlled by the new
1098 host_all_ignored option.
1100 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1101 homing or because of multiple MX records with /mx) generated more than
1102 one IP address, and the following item turned out to be the local host,
1103 all the secondary addresses of the first item were incorrectly removed
1104 from the list, along with the local host and any following hosts (which
1105 is what is supposed to happen).
1107 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1108 whoever called Exim into the -H file. In the case of the daemon it was
1109 behaving confusingly. When first started, it used values for whoever
1110 started the daemon, but after a SIGHUP it used the Exim user (because it
1111 calls itself on a restart). I have changed the code so that it now always
1114 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1115 message are rejected with the same error (e.g. no authentication or bad
1116 sender address), and a DATA command is nevertheless sent (as can happen
1117 with PIPELINING or a stupid MUA), the error message that was given to the
1118 RCPT commands is included in the rejection of the DATA command. This is
1119 intended to be helpful for MUAs that show only the final error to their
1122 PH/42 Another patch from the Sieve maintainer.
1124 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1125 Eximstats - Fixed some broken HTML links and added missing column headers
1127 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1128 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1130 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1132 PH/43 Yet another patch from the Sieve maintainer.
1134 PH/44 I found a way to check for a TCP/IP connection going away before sending
1135 the response to the final '.' that terminates a message, but only in the
1136 case where the client has not sent further data following the '.'
1137 (unfortunately, this is allowed). However, in many cases there won't be
1138 any further data because there won't be any more messages to send. A call
1139 to select() can be used: if it shows that the input is "ready", there is
1140 either input waiting, or the socket has been closed. An attempt to read
1141 the next input character can distinguish the two cases. Previously, Exim
1142 would have sent an OK response which the client would never have see.
1143 This could lead to message repetition. This fix should cure that, at
1144 least in a lot of common cases.
1146 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1147 advertised in response to EHLO.
1153 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1154 fixed by 4.65/MH/01 (is this a record?) are fixed:
1156 (i) An empty string was always treated as zero by the numeric comparison
1157 operators. This behaviour has been restored.
1159 (ii) It is documented that the numeric comparison operators always treat
1160 their arguments as decimal numbers. This was broken in that numbers
1161 starting with 0 were being interpreted as octal.
1163 While fixing these problems I realized that there was another issue that
1164 hadn't been noticed. Values of message_size_limit (both the global option
1165 and the transport option) were treated as octal if they started with 0.
1166 The documentation was vague. These values are now always treated as
1167 decimal, and I will make that clear in the documentation.
1173 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1174 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1177 MH/01 Don't check that the operands of numeric comparison operators are
1178 integers when their expansion is in "skipping" mode (fixes bug
1179 introduced by 4.64-PH/07).
1181 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1182 child addresses, Exim now panics and dies. Previously, because the count
1183 is held in a short int, deliveries were likely to be lost. As such a
1184 large number of recipients for a single message is ridiculous
1185 (performance will be very, very poor), I have chosen to impose a limit
1186 rather than extend the field.
1192 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1193 leftover -K file (the existence of which was triggered by #402).
1194 While we were at it, introduced process PID as part of the -K
1195 filename. This should rule out race conditions when creating
1198 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1199 processing considerably. Previous code took too long for large mails,
1200 triggering a timeout which in turn triggers #401.
1202 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1203 in the DK code in transports.c. sendfile() is not really portable,
1204 hence the _LINUX specificness.
1206 TF/01 In the add_headers option to the mail command in an Exim filter,
1207 there was a bug that Exim would claim a syntax error in any
1208 header after the first one which had an odd number of characters
1211 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1212 callout verification, Exim cached a "reject" for the entire domain. This
1213 is correct for most verifications, but it is not correct for a recipient
1214 verification with use_sender or use_postmaster set, because in that case
1215 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1216 case of MAIL FROM:<> rejection from other early rejections (e.g.
1217 rejection of HELO). When verifying a recipient using a non-null MAIL
1218 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1219 Whatever the result of the callout, the value of the domain cache is
1220 left unchanged (for any other kind of callout, getting as far as trying
1221 RCPT means that the domain itself is ok).
1223 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1226 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1227 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1228 session, the daemon ignores ECONNECT errors and logs others; it now
1229 ignores EPIPE as well.
1231 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1232 (quoted-printable decoding).
1234 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1235 later the small subsequent patch to fix an introduced bug.
1237 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1239 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1241 PH/08 An error is now given if message_size_limit is specified negative.
1243 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1244 to be given (somewhat) arbitrary names.
1246 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1249 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1250 miscellaneous code fixes
1252 PH/10 Added the log_reject_target ACL modifier to specify where to log
1255 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1256 hostname. This is wrong, because it relates to the incoming message (and
1257 probably the interface on which it is arriving) and not to the outgoing
1258 callout (which could be using a different interface). This has been
1259 changed to use the value of the helo_data option from the smtp transport
1260 instead - this is what is used when a message is actually being sent. If
1261 there is no remote transport (possible with a router that sets up host
1262 addresses), $smtp_active_hostname is used.
1264 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1265 tweaks were necessary in order to get it to work (see also 21 below):
1266 (a) The code assumed that strncpy() returns a negative number on buffer
1267 overflow, which isn't the case. Replaced with Exim's string_format()
1269 (b) There were several signed/unsigned issues. I just did the minimum
1270 hacking in of casts. There is scope for a larger refactoring.
1271 (c) The code used strcasecmp() which is not a standard C function.
1272 Replaced with Exim's strcmpic() function.
1273 (d) The code set only $1; it now sets $auth1 as well.
1274 (e) A simple test gave the error "authentication client didn't specify
1275 service in request". It would seem that Dovecot has changed its
1276 interface. Fortunately there's a specification; I followed it and
1277 changed what the client sends and it appears to be working now.
1279 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1282 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1283 address A is aliased to B and C, where B exists and C does not. Without
1284 -v the output is "A verified" because verification stops after a
1285 successful redirection if more than one address is generated. However,
1286 with -v the child addresses are also verified. Exim was outputting "A
1287 failed to verify" and then showing the successful verification for C,
1288 with its parentage. It now outputs "B failed to verify", showing B's
1289 parentage before showing the successful verification of C.
1291 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1292 look up a TXT record in a specific list after matching in a combined
1295 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1296 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1297 they consult the DNS. I had assumed they would set it the way they
1298 wanted; and indeed my experiments on Linux seem to show that in some
1299 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1300 To be on the safe side, however, I have now made the interface to
1301 host_find_byname() similar to host_find_bydns(), with an argument
1302 containing the DNS resolver options. The host_find_byname() function now
1303 sets these options at its start, just as host_find_bydns() does. The smtp
1304 transport options dns_qualify_single and dns_search_parents are passed to
1305 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1306 of host_find_byname() use the default settings of RES_DEFNAMES
1307 (qualify_single) but not RES_DNSRCH (search_parents).
1309 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1310 spool_read_header() do less string testing, by means of a preliminary
1311 switch on the second character of optional "-foo" lines. (This is
1312 overdue, caused by the large number of possibilities that now exist.
1313 Originally there were few.) While I was there, I also converted the
1314 str(n)cmp tests so they don't re-test the leading "-" and the first
1315 character, in the hope this might squeeze out yet more improvement.
1317 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1318 flag allowing group syntax was set by the header_syntax check but not
1319 turned off, possible causing trouble later; (2) The flag was not being
1320 set at all for the header_verify test, causing "group"-style headers to
1321 be rejected. I have now set it in this case, and also caused header_
1322 verify to ignore an empty address taken from a group. While doing this, I
1323 came across some other cases where the code for allowing group syntax
1324 while scanning a header line wasn't quite right (mostly, not resetting
1325 the flag correctly in the right place). These bugs could have caused
1326 trouble for malformed header lines. I hope it is now all correct.
1328 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1329 with the "reply" argument non-NULL. The code, however (which originally
1330 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1331 but it didn't always do it. This confused somebody who was copying the
1332 code for some other use. I have removed all the tests.
1334 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1335 feature that was used to support insecure browsers during the U.S. crypto
1336 embargo. It requires special client support, and Exim is probably the
1337 only MTA that supported it -- and would never use it because real RSA is
1338 always available. This code has been removed, because it had the bad
1339 effect of slowing Exim down by computing (never used) parameters for the
1340 RSA_EXPORT functionality.
1342 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1343 authenticator to fail if there's a tab character in the incoming data
1344 (there should never be unless someone is messing about, as it's supposed
1345 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1346 if the connection is using TLS or if the remote IP is the same as the
1347 local IP, and the "valid-client-cert option" if a client certificate has
1350 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1351 authenticators. This can be used for authorization after authentication
1352 succeeds. (In the case of plaintext, it servers for both authentication
1355 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1356 if any retry times were supplied.
1358 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1359 connection, where there is no client IP address to check. In this
1360 situation, the verify now always succeeds.
1362 PH/25 Applied John Jetmore's -Mset patch.
1364 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1366 PH/27 In a string expansion for a processed (not raw) header when multiple
1367 headers of the same name were present, leading whitespace was being
1368 removed from all of them, but trailing whitespace was being removed only
1369 from the last one. Now trailing whitespace is removed from each header
1370 before concatenation. Completely empty headers in a concatenation (as
1371 before) are ignored.
1373 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1374 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1376 PH/29 [Removed. This was a change that I later backed out, and forgot to
1377 correct the ChangeLog entry (that I had efficiently created) before
1378 committing the later change.]
1380 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1381 address errors (4xx response to RCPT) over the same connection as other
1382 messages routed to the same hosts. Such deliveries are always "forced",
1383 so retry times are not inspected. This resulted in far too many retries
1384 for the affected addresses. The effect occurred only when there were more
1385 hosts than the hosts_max_try setting in the smtp transport when it had
1386 the 4xx errors. Those hosts that it had tried were not added to the list
1387 of hosts for which the message was waiting, so if all were tried, there
1388 was no problem. Two fixes have been applied:
1390 (i) If there are any address or message errors in an SMTP delivery, none
1391 of the hosts (tried or untried) are now added to the list of hosts
1392 for which the message is waiting, so the message should not be a
1393 candidate for sending over the same connection that was used for a
1394 successful delivery of some other message. This seems entirely
1395 reasonable: after all the message is NOT "waiting for some host".
1396 This is so "obvious" that I'm not sure why it wasn't done
1397 previously. Hope I haven't missed anything, but it can't do any
1398 harm, as the worst effect is to miss an optimization.
1400 (ii) If, despite (i), such a delivery is accidentally attempted, the
1401 routing retry time is respected, so at least it doesn't keep
1402 hammering the server.
1404 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1405 in ${readsocket because some servers need this prod.
1407 PH/32 Added some extra debug output when updating a wait-xxx database.
1409 PH/33 The hint "could be header name not terminated by colon", which has been
1410 given for certain expansion errors for a long time, was not being given
1411 for the ${if def:h_colon_omitted{... case.
1413 PH/34 The spec says: "With one important exception, whenever a domain list is
1414 being scanned, $domain contains the subject domain." There was at least
1415 one case where this was not true.
1417 PH/35 The error "getsockname() failed: connection reset by peer" was being
1418 written to the panic log as well as the main log, but it isn't really
1419 panic-worthy as it just means the connection died rather early on. I have
1420 removed the panic log writing for the ECONNRESET error when getsockname()
1423 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1424 runs only) independently of the message's sender address. This meant
1425 that, if the 4xx error was in fact related to the sender, a different
1426 message to the same recipient with a different sender could confuse
1427 things. In particualar, this can happen when sending to a greylisting
1428 server, but other circumstances could also provoke similar problems.
1429 I have changed the default so that the retry time for these errors is now
1430 based a combination of the sender and recipient addresses. This change
1431 can be overridden by setting address_retry_include_sender=false in the
1434 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1435 remote server are returned as part of bounce messages. This was not
1436 happening for LMTP over a pipe (the lmtp transport), but now it is the
1437 same for both kinds of LMTP.
1439 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1440 in header lines that were in fact CNAMEs. This is no longer the case.
1442 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1443 and queue runs started by the daemon processed all messages. This has
1444 been fixed so that -R and -S can now usefully be given with -q<time>.
1446 PH/40 Import PCRE release 6.7 (fixes some bugs).
1448 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1450 PH/42 Give an error if -q is specified more than once.
1452 PH/43 Renamed the variables $interface_address and $interface_port as
1453 $received_ip_address and $received_port, to make it clear that these
1454 values apply to message reception, and not to the outgoing interface when
1455 a message is delivered. (The old names remain recognized, of course.)
1457 PH/44 There was no timeout on the connect() call when using a Unix domain
1458 socket in the ${readsocket expansion. There now is.
1460 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1461 be meaningful with "accept".
1463 SC/01 Eximstats V1.43
1464 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1466 SC/02 Eximstats V1.44
1467 Use a glob alias rather than an array ref in the generated
1468 parser. This improves both readability and performance.
1470 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1471 Collect SpamAssassin and rejection statistics.
1472 Don't display local sender or destination tables unless
1473 there is data to show.
1474 Added average volumes into the top table text output.
1476 SC/04 Eximstats V1.46
1477 Collect data on the number of addresses (recipients)
1478 as well as the number of messages.
1480 SC/05 Eximstats V1.47
1481 Added 'Message too big' to the list of mail rejection
1482 reasons (thanks to Marco Gaiarin).
1484 SC/06 Eximstats V1.48
1485 Mainlog lines which have GMT offsets and are too short to
1486 have a flag are now skipped.
1488 SC/07 Eximstats V1.49 (Alain Williams)
1489 Added the -emptyok flag.
1491 SC/08 Eximstats V1.50
1492 Fixes for obtaining the IP address from reject messages.
1494 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1495 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1496 whitesspace changes from 4.64-PH/27
1498 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1501 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1502 are found, allow negative numbers in numeric criteria)
1504 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1506 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1507 to match changes made in 4.64-PH/43
1509 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1511 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1512 "domain" is actually an IP address, and force a failure. This locks out
1513 those revolvers/nameservers that support "A-for-A" lookups, in
1514 contravention of the specifications.
1516 PH/48 When a host name was looked up from an IP address, and the subsequent
1517 forward lookup of the name timed out, the host name was left in
1518 $sender_host_name, contrary to the specification.
1520 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1521 restricted to single-key lookups, Exim was not diagnosing an error if
1522 * or *@ was used with a query-style lookup.
1524 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1526 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1527 long ago, but noone interested enough thought of it. Let's just say that
1528 the "1.1" means that there are some new functions that weren't there at
1529 some point in the past.
1531 PH/51 Error processing for expansion failure of helo_data from an smtp
1532 transport during callout processing was broken.
1534 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1535 tested/used via the -bh/-bhc/-bs options.
1537 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1538 bug, fixed in subsequent PCRE releases).
1540 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1541 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1543 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1549 SC/01 Use a glob alias rather than an array ref in eximstats generated
1550 parser. This improves both readability and performance.
1552 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1553 Don't display local sender or destination tables in eximstats unless
1554 there is data to show.
1555 Added average volumes into the eximstats top table text output.
1557 SC/03 Collect data on the number of addresses (recipients) as well
1558 as the number of messages in eximstats.
1560 TF/01 Correct an error in the documentation for the redirect router. Exim
1561 does (usually) call initgroups() when daemonizing.
1563 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1564 with consistent privilege compared to when running as a daemon.
1566 TF/03 Note in the spec that $authenticated_id is not set for local
1567 submissions from trusted users.
1569 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1570 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1572 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1573 by adding some example configuration directives to the default
1574 configuration file. A little bit of work is required to uncomment the
1575 directives and define how usernames and passwords are checked, but
1576 there is now a framework to start from.
1578 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1579 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1580 without this. I don't know how relevant this is to other LDAP libraries.
1582 PH/02 Add the verb name to the "unknown ACL verb" error.
1584 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1586 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1588 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1589 directory not expanded when it should be if an expanded home directory
1590 was set for the address (which is overridden by the transport).
1592 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1595 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1596 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1597 because it is too late at that time, and has no effect.
1599 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1600 security issue with \' (bugzilla #107). I could not use the
1601 PQescapeStringConn() function, because it needs a PGconn value as one of
1604 PH/08 When testing addresses using -bt, indicate those final addresses that
1605 are duplicates that would not cause an additional delivery. At least one
1606 person was confused, thinking that -bt output corresponded to deliveries.
1607 (Suppressing duplicates isn't a good idea as you lose the information
1608 about possibly different redirections that led to the duplicates.)
1610 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1611 systems where poll() doesn't work, in particular OS X.
1613 PH/10 Added more information to debugging output for retry time not reached.
1615 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1616 operations in malware.c.
1618 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1621 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1622 syslog_duplication set false, log lines that would normally be written
1623 both the the main log and to the reject log were not written to syslog at
1626 PH/14 In the default configuration, change the use of "message" in ACL warn
1627 statements to "add_header".
1629 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1630 not followed by a command (e.g. "seen endif").
1632 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1633 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1636 PH/17 Added extra conditions to the default value of delay_warning_condition
1640 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1641 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1642 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1645 The Auto-Submitted: and various List- headers are standardised, whereas I
1646 don't think Precedence: ever was.
1648 PH/18 Refactored debugging code in route_finduser() to show more information,
1649 in particular, the error code if getpwnam() issues one.
1651 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1652 This is apparently needed in addition to the PH/07 change above to avoid
1653 any possible encoding problems.
1655 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1656 but not after initializing Perl.
1658 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1659 output them only if debugging. By default they are written stderr,
1660 apparently, which is not desirable.
1662 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1665 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1668 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1670 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1671 authenticated or an ident call has been made. Suppress the default
1672 values for $authenticated_id and $authenticated_sender (but permit -oMai
1673 and -oMas) when testing with -bh.
1675 PH/24 Re-jigged the order of the tests in the default configuration so that the
1676 tests for valid domains and recipients precede the DNS black list and CSA
1677 tests, on the grounds that those ones are more expensive.
1679 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1680 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1681 This bug exists in every version of Exim that I still have, right back to
1684 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1685 However, an attempt to turn on case-sensitivity in a regex key by
1686 including (?-i) didn't work because the subject string was already
1687 lowercased, and the effects were non-intuitive. It turns out that a
1688 one-line patch can be used to allow (?-i) to work as expected.
1694 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1695 other effects) broke the use of negated acl sub-conditions.
1697 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1700 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1701 "Deny" causes Exim to reject the incoming connection with a 554 error.
1702 Unfortunately, if there is a major crisis, such as a disk failure,
1703 tcp-wrappers gives "deny", whereas what one would like would be some
1704 kind of temporary error. A kludge has been added to help with this.
1705 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1706 554 error is used if errno is still zero or contains ENOENT (which occurs
1707 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1710 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1712 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1713 errors. Otherwise a message that provokes a temporary error (when other
1714 messages do not) can cause a whole host to time out.
1716 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1717 addresses were routed directly to files or pipes from a redirect router.
1718 File deliveries just didn't batch; pipe deliveries might have suffered
1721 PH/06 A failure to get a lock for a hints database would erroneously always say
1722 "Failed to get write lock", even when it was really a read lock.
1724 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1725 of 0600. This has been changed to use the value of the lockfile_mode
1726 option (which defaults to 0600).
1728 PH/08 Applied small patch from the Sieve maintainer.
1730 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1731 folder from quota calculations, a direct delivery into this folder messed
1732 up the contents of the maildirsize file. This was because the regex was
1733 used only to exclude .Trash (or whatever) when the size of the mailbox
1734 was calculated. There was no check that a delivery was happening into an
1735 excluded directory. This bug has been fixed by ignoring all quota
1736 processing for deliveries into excluded directories.
1738 PH/10 Added the maildirfolder_create_regex option to appendfile.
1744 PH/01 The code for finding all the local interface addresses on a FreeBSD
1745 system running IPv6 was broken. This may well have applied to all BSD
1746 systems, as well as to others that have similar system calls. The broken
1747 code found IPv4 interfaces correctly, but gave incorrect values for the
1748 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1749 that it would not match correctly against @[] and not recognize the IPv6
1752 PH/02 The ipliteral router was not recognizing addresses of the form user@
1753 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1755 PH/03 Added disable_ipv6.
1757 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1758 decoded form, because it is most often used to construct To: headers
1759 lines in autoreplies, and the decoded form may well be syntactically
1760 invalid. However, $reply_address has leading white space removed, and all
1761 newlines turned into spaces so that the autoreply transport does not
1764 PH/05 If group was specified without a user on a router, and no group or user
1765 was specified on a transport, the group from the router was ignored.
1767 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1768 for visible compile-time settings that can be used to change these
1769 numbers, for those that want even more. Backwards compatibility with old
1770 spool files has been maintained. However, going back to a previous Exim
1771 release will lost any variables that are in spool files.
1773 PH/07 Two small changes when running in the test harness: increase delay when
1774 passing a TCP/IP connection to a new process, in case the original
1775 process has to generate a bounce, and remove special handling of
1776 127.0.0.2 (sic), which is no longer necessary.
1778 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1779 be the same on different OS.
1781 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1784 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1785 whether --show-vars was specified or not
1787 JJ/02 exipick: Added support for new ACL variable spool format introduced
1790 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1791 syntactically invalid From: or Reply-to: line, and a filter used this to
1792 generate an autoreply, and therefore failed to obtain an address for the
1793 autoreply, Exim could try to deliver to a non-existent relative file
1794 name, causing unrelated and misleading errors. What now happens is that
1795 it logs this as a hard delivery error, but does not attempt to create a
1798 PH/11 The exinext utility has a -C option for testing purposes, but although
1799 the given file was scanned by exinext itself; it wasn't being passed on
1800 when Exim was called.
1802 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1803 an end-of-file indication when reading a command response.
1805 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1806 compiled. In many other places in Exim, IPv6 addresses are always
1807 recognized, so I have changed this. It also means that IPv4 domain
1808 literals of the form [IPV4:n.n.n.n] are now always recognized.
1810 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1811 used if the router is not running as root, for example, when verifying at
1812 ACL time, or when using -bh. The debugging output from this situation was
1813 non-existent - all you got was a failure to exec. I have made two
1816 (a) Failures to set uid/gid, the current directory, or a process leader
1817 in a subprocess such as that created by queryprogram now generate
1818 suitable debugging ouput when -d is set.
1820 (b) The queryprogram router detects when it is not running as root,
1821 outputs suitable debugging information if -d is set, and then runs
1822 the subprocess without attempting to change uid/gid.
1824 PH/15 Minor change to Makefile for building test_host (undocumented testing
1827 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1828 additional section of a DNS packet that returns MX or SRV records.
1829 Instead, it always explicitly searches for A/AAAA records. This avoids
1830 major problems that occur when a DNS server includes only records of one
1831 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1832 fixed another bug: if SRV records were looked up and the corresponding
1833 address records were *not* found in the additional section, the port
1834 values from the SRV records were lost.
1836 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1837 using the correct key (the original address) when searching the retry
1838 rules in order to find which one to use for generating the retry hint.
1840 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1841 adding the default one. Similarly, if it contains a Reply-To: header, the
1842 errors_reply_to option, if set, is not used.
1844 PH/19 When calculating a retry time, Exim used to measure the "time since
1845 failure" by looking at the "first failed" field in the retry record. Now
1846 it does not use this if it is later than than the arrival time of the
1847 message. Instead it uses the arrival time. This makes for better
1848 behaviour in cases where some deliveries succeed, thus re-setting the
1849 "first failed" field. An example is a quota failure for a huge message
1850 when small messages continue to be delivered. Without this change, the
1851 "time since failure" will always be short, possible causing more frequent
1852 delivery attempts for the huge message than are intended.
1853 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1855 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1856 $1, $2, $3) because the numerical variables can be reset during some
1857 expansion items (e.g. "match"), thereby losing the authentication data.
1859 PH/21 Make -bV show the size of off_t variables so that the test suite can
1860 decide whether to run tests for quotas > 2G.
1862 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1863 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1864 filecount value is greater than 2G or if a quota value is greater than 2G
1865 on a system where the size of off_t is not greater than 4, a panic error
1868 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1869 never match. The debug and -bh output now contains an explicit error
1870 message indicating a malformed IPv4 address or mask.
1872 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1873 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1874 PH/23 above applies.
1876 PH/25 Do not write to syslog when running in the test harness. The only
1877 occasion when this arises is a failure to open the main or panic logs
1878 (for which there is an explicit test).
1880 PH/26 Added the /no_tell option to "control=freeze".
1882 PH/27 If a host name lookup failed very early in a connection, for example, if
1883 the IP address matched host_lookup and the reverse lookup yielded a name
1884 that did not have a forward lookup, an error message of the form "no IP
1885 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1886 could be logged. Now it outputs the IP address instead of "NULL".
1888 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1889 allows the sender and the authenticated sender to be set when
1890 submitting a message from within Exim. Since child_open_exim() is
1891 documented for local_scan(), the new function should be too.
1893 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1894 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1895 results in an empty string is now treated as unset.
1897 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1899 PH/31 Added sender_verify_fail logging option.
1901 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1902 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1903 tidied the source and removed it altogether.
1905 PH/33 When a queue run was abandoned because the load average was too high, a
1906 log line was always written; now it is written only if the queue_run log
1907 selector is set. In addition, the log line for abandonment now contains
1908 information about the queue run such as the pid. This is always present
1909 in "start" and "stop" lines but was omitted from the "abandon" line.
1911 PH/34 Omit spaces between a header name and the colon in the error message that
1912 is given when verify = headers_syntax fails (if there are lots of them,
1913 the message gets confusing).
1915 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1916 names, as there are now some PTR records that contain slashes. This check
1917 is only to protect against broken name servers that fall over on strange
1918 characters, so the fact that it applies to all lookups doesn't matter.
1920 PH/36 Now that the new test suite is complete, we can remove some of the
1921 special code in Exim that was needed for the old test suite. For example,
1922 sorting DNS records because real resolvers return them in an arbitrary
1923 order. The new test suite's fake resolver always returns records in the
1926 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1927 bounces) except when queue_only is set, to avoid logging races between
1928 the different processes.
1930 PH/38 Panic-die if .include specifies a non-absolute path.
1932 PH/39 A tweak to the "H" retry rule from its user.
1934 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1935 a label. They prevented compilation on older perls.
1937 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1938 a warning to be raised on newish perls.
1940 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1941 on queue. Changes to match documented behaviour of showing count of
1942 messages matching specified criteria.
1944 PH/40 Changed the default ident timeout from 30s to 5s.
1946 PH/41 Added support for the use of login_cap features, on those BSD systems
1947 that have them, for controlling the resources used by pipe deliveries.
1949 PH/42 The content-scanning code uses fopen() to create files in which to put
1950 message data. Previously it was not paying any attention to the mode of
1951 the files. Exim runs with umask(0) because the rest of the code creates
1952 files with open(), and sets the required mode explicitly. Thus, these
1953 files were ending up world-writeable. This was not a big issue, because,
1954 being within the spool directory, they were not world-accessible. I have
1955 created a function called modefopen, which takes an additional mode
1956 argument. It sets umask(777), creates the file, chmods it to the required
1957 mode, then resets the umask. All the relevant calls to fopen() in the
1958 content scanning code have been changed to use this function.
1960 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1961 to 24 hours. This avoids potential overflow problems when processing G
1962 and H retry rules. I suspect nobody ever tinkers with this value.
1964 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1966 PH/45 When the plaintext authenticator is running as a client, the server's
1967 challenges are checked to ensure they are valid base64 strings. By
1968 default, the authentication attempt is cancelled if an invalid string is
1969 received. Setting client_ignore_invalid_base64 true ignores these errors.
1970 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1971 they are received. Thus, the responses can be made to depend on the
1972 challenges. If an invalid string is ignored, an empty string is placed in
1975 PH/46 Messages that are created by the autoreply transport now contains a
1976 References: header, in accordance with RFCs 2822 and 3834.
1978 PH/47 Added authenticated_sender_force to the smtp transport.
1980 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1982 PH/49 Installed latest patch from the Sieve maintainer.
1984 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1985 was also set, the appendfile transport was unnecessarily scanning a
1986 directory of message files (e.g. for maildir delivery) to find the count
1987 of files (along with the size), even though it did not need this
1988 information. It now does the scan only if it needs to find either the
1989 size of the count of files.
1991 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1993 PH/52 Two bugs concerned with error handling when the smtp transport is
1996 (i) Exim was not creating retry information for temporary errors given
1997 for individual recipients after the DATA command when the smtp transport
1998 was used in LMTP mode. This meant that they could be retried too
1999 frequently, and not timed out correctly.
2001 (ii) Exim was setting the flag that allows error details to be returned
2002 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2003 recipients that were returned after the DATA command.
2005 PH/53 This is related to PH/52, but is more general: for any failing address,
2006 when detailed error information was permitted to be returned to the
2007 sender, but the error was temporary, then after the final timeout, only
2008 "retry timeout exceeded" was returned. Now it returns the full error as
2009 well as "retry timeout exceeded".
2011 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2012 do this, and (what is worse) MTAs that accept it.
2014 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2015 will now be deprecated.
2017 PH/56 New os.c-cygwin from the Cygwin maintainer.
2019 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2020 formats (previously only available in exim formats via -bpr, -bpru,
2021 and -bpra. Now also available in native and exiqgrep formats)
2023 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2024 with very large, slow to parse queues
2026 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2028 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2030 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2031 responses to authentication challenges, though it was showing the
2032 challenges; (ii) I've removed the CR characters from the debug output for
2035 PH/58 Allow for the insertion of a newline as well as a space when a string
2036 is turned into more than one encoded-word during RFC 2047 encoding. The
2037 Sieve code now uses this.
2039 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2040 data_4xx, lost_connection, tls_required.
2042 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2043 message was being sent as an SMTP response.
2045 PH/61 Add -l and -k options to exicyclog.
2047 PH/62 When verifying, if an address was redirected to one new address, so that
2048 verification continued, and the new address failed or deferred after
2049 having set something in $address_data, the value of $address_data was not
2050 passed back to the ACL. This was different to the case when no
2051 redirection occurred. The value is now passed back in both cases.
2053 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2054 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2055 use login_cap.h, so on its own it isn't the distinguishing feature. The
2056 new name refers directly to the setclassresources() function.
2058 PH/65 Added configuration files for NetBSD3.
2060 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2062 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2063 is preferred over IPv4.
2065 PH/68 The bounce_return_message and bounce_return_body options were not being
2066 honoured for bounces generated during the reception of non-SMTP messages.
2067 In particular, this applied to messages rejected by the ACL. This bug has
2068 been fixed. However, if bounce_return_message is true and bounce_return_
2069 body is false, the headers that are returned for a non-SMTP message
2070 include only those that have been read before the error was detected.
2071 (In the case of an ACL rejection, they have all been read.)
2073 PH/69 The HTML version of the specification is now built in a directory called
2074 spec_html instead of spec.html, because the latter looks like a path with
2075 a MIME-type, and this confuses some software.
2077 PH/70 Catch two compiler warnings in sieve.c.
2079 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2080 function verify_get_ident() calls ip_connect() to connect a socket, but
2081 if the "connect()" function timed out, ip_connect() used to close the
2082 socket. However, verify_get_ident() also closes the socket later, and in
2083 between Exim writes to the log, which may get opened at this point. When
2084 the socket was closed in ip_connect(), the log could get the same file
2085 descriptor number as the socket. This naturally causes chaos. The fix is
2086 not to close the socket in ip_connect(); the socket should be closed by
2087 the function that creates it. There was only one place in the code where
2088 this was missing, in the iplookup router, which I don't think anybody now
2089 uses, but I've fixed it anyway.
2091 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2092 well as to direct DNS lookups. Otherwise the handling of names in host
2093 lists is inconsistent and therefore confusing.
2099 PH/01 Two changes to the default runtime configuration:
2101 (1) Move the checks for relay_from_hosts and authenticated clients from
2102 after to before the (commented out) DNS black list checks.
2104 (2) Add control=submission to the relay_from_hosts and authenticated
2105 clients checks, on the grounds that messages accepted by these
2106 statements are most likely to be submissions.
2108 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2110 (1) Generate an error if the third argument for the ${prvs expansion is
2113 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2116 (3) Reset the variables that are obtained from the first argument of
2117 ${prvscheck and used in the second argument before leaving the code,
2118 because their memory is reclaimed, so using them afterwards may do
2121 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2122 one (it's much easier than Tom thought :-).
2124 (5) Because of (4), we can now allow for the use of $prvscheck_result
2125 inside the third argument.
2127 PH/03 For some reason, the default setting of PATH when running a command from
2128 a pipe transport was just "/usr/bin". I have changed it to
2131 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2132 anything to be listed in the output from -bV.
2134 PH/05 When a filter generated an autoreply, the entire To: header line was
2135 quoted in the delivery log line, like this:
2137 => >A.N.Other <ano@some.domain> <original@ddress> ...
2139 This has been changed so that it extracts the operative address. There
2140 may be more than one such address. If so, they are comma-separated, like
2143 => >ano@some.domain,ona@other.domain <original@ddress> ...
2145 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2146 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2147 not being looked up in the rDNS to get a host name, Exim was showing the
2148 IP address twice in Received: lines, even though the IP addresses were
2149 identical. For example:
2151 Received: from [1.2.3.4] (helo=[1.2.3.4])
2153 However, if the real host name was known, it was omitting the HELO data
2154 if it matched the actual IP address. This has been tidied up so that it
2155 doesn't show the same IP address twice.
2157 PH/07 When both +timestamp and +memory debugging was on, the value given by
2158 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2159 called by the debug printing, thereby overwriting the timestamp buffer.
2160 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2162 PH/08 When the original message was included in an autoreply transport, it
2163 always said "this is a copy of the message, including all the headers",
2164 even if body_only or headers_only was set. It now gives an appropriate
2167 PH/09 Applied a patch from the Sieve maintainer which:
2169 o fixes some comments
2170 o adds the (disabled) notify extension core
2171 o adds some debug output for the result of if/elsif tests
2172 o points to the current vacation draft in the documentation
2173 and documents the missing references header update
2177 o fixes a bug in processing the envelope test (when testing
2178 multiple envelope elements, the last element determinted the
2181 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2182 Electronic Mail") by including:
2184 Auto-submitted: auto-generated
2186 in the messages that it generates (bounce messages and others, such as
2187 warnings). In the case of bounce messages for non-SMTP mesages, there was
2188 also a typo: it was using "Auto_submitted" (underscore instead of
2189 hyphen). Since every message generated by Exim is necessarily in response
2190 to another message, thes have all been changed to:
2192 Auto-Submitted: auto-replied
2194 in accordance with these statements in the RFC:
2196 The auto-replied keyword:
2198 - SHOULD be used on messages sent in direct response to another
2199 message by an automatic process,
2201 - MUST NOT be used on manually-generated messages,
2203 - MAY be used on Delivery Status Notifications (DSNs) and Message
2204 Disposition Notifications (MDNs),
2206 - MUST NOT be used on messages generated by automatic or periodic
2207 processes, except for messages which are automatic responses to
2210 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2211 to the default Received: header definition.
2213 PH/12 Added log selector acl_warn_skipped (default on).
2215 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2216 variables because (a) they are in the wrong storage pool and (b) even if
2217 they were copied, it wouldn't work properly because of the caching.
2219 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2220 checking when decoding. Apparently there are clients that generate
2221 overlong encoded strings. Why am I not surprised?
2223 PH/15 If the first argument of "${if match_address" was not empty, but did not
2224 contain an "@" character, Exim crashed. Now it writes a panic log message
2225 and treats the condition as false.
2227 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2229 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2230 extension "envelope-auth". The code is finished and in agreement with
2231 other implementations, but there is no documentation so far and in fact,
2232 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2233 not changing the active code.
2235 Print executed "if" and "elsif" statements when debugging is used. This
2236 helps a great deal to understand what a filter does.
2238 Document more things not specified clearly in RFC3028. I had all this
2239 sorted out, when out of a sudden new issues came to my mind. Oops."
2241 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2244 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2245 canonical form (as documented). However, after a host name lookup from
2246 the IP address, check_host() was doing a simple string comparison with
2247 addresses acquired from the DNS when checking that the found name did
2248 have the original IP as one of its addresses. Since any found IPv6
2249 addresses are likely to be in abbreviated form, the comparison could
2250 fail. Luckily, there already exists a function for doing the comparison
2251 by converting both addresses to binary, so now that is used instead of
2252 the text comparison.
2254 PH/20 There was another similar case to PH/19, when a complete host name was
2255 given in a host list; looking up its IP address could give an abbreviated
2256 form, whereas the current host's name might or might not be abbreviated.
2257 The same fix has been applied.
2263 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2264 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2267 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2268 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2270 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2272 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2273 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2274 Also, for "Auto-Submitted", treat anything other than "no" as
2275 non-personal, in accordance with RFC 3834. (Previously it treated
2276 anything starting "auto-" as non-personal.)
2278 TF/01 The control=submission/name=... option had a problem with syntax
2279 errors if the name included a slash character. The /name= option
2280 now slurps the rest of the string, so it can include any characters
2281 but it must come last in the list of options (after /sender_retain
2284 PH/05 Some modifications to the interface to the fake nameserver for the new
2292 TK/01 Added the "success_on_redirect" address verification option. See
2293 NewStuff for rationale and an example.
2295 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2297 PH/02 Patch to exigrep to allow it to work on syslog lines.
2299 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2300 fread() to skip over the body file's header line, because in Cygwin the
2301 header line is locked and is inaccessible.
2303 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2304 co-exist for some time) to make it clear that it is the Exim ID that is
2305 referenced, not the Message-ID: header line.
2307 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2308 string_format() function, because snprintf() does not exist on all
2311 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2312 ${stat: expansion item.
2314 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2315 protocol synchronization error", to keep the pedants happy.
2317 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2318 well as for IRIX systems, when gcc is being used. See the host.c source
2321 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2323 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2326 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2327 transport and to the smtp transport in LMTP mode.
2329 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2331 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2332 run a filter in a subprocess. This could lead to confusion in subsequent
2333 lookups in the parent process. There should also be a search_tidyup() at
2334 the end of the subprocess.
2336 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2337 only if the host matched helo_try_verify_hosts, which caused the
2338 verification to occur when the EHLO/HELO command was issued. The ACL just
2339 tested the remembered result. Now, if a previous verification attempt has
2340 not happened, "verify = helo" does it there and then.
2342 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2344 TK/03 Fix log output including CR from clamd.
2346 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2347 exist provoked a memory error which could cause a segfault.
2349 PH/15 Installed PCRE 6.2
2351 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2353 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2354 of the problem. Specifically, suggested +O2 rather than +O1 for the
2357 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2359 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2360 record, the port was not correctly logged when the outgoing_port log
2361 selector was set (it logged the transort's default port).
2363 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2364 fallback_hosts, and "hosts" in the smtp transport.
2366 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2367 host errors such as "Connection refused".
2369 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2370 authentication with radiusclient 0.4.9:
2372 - Error returned from rc_read_config was caught wrongly
2373 - Username/password not passed on to radius server due to wrong length.
2375 The presumption is that some radiusclient API changes for 4.51/PH/17
2376 were not taken care of correctly. The code is still untested by me (my
2377 Linux distribution still has 0.3.2 of radiusclient), but it was
2378 contributed by a Radius user.
2380 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2381 expanding the "port" option of the smtp transport.
2383 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2384 while reading a MIME header. Thanks to Tom Hughes for a patch.
2386 PH/24 Include config.h inside local_scan.h so that configuration settings are
2389 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2390 that in an ACL for RCPT (for example), you can examine exactly what was
2393 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2394 commands, but it was not correctly comparing the address with the actual
2395 client host address. Thus, it would show the EHLO address in Received:
2396 header lines when this was not necessary.
2398 PH/27 Added the % operator to ${eval:}.
2400 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2401 it should be ignoring failures (because with -C, for example, it has lost
2402 privilege). It wasn't ignoring creation failures other than "already
2405 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2408 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2409 because an input line was too long, either on its own, or by virtue of
2410 too many continuations, the temporary file was not being removed, and the
2411 return code was incorrect.
2413 PH/31 Missing "BOOL" in function definition in filtertest.c.
2415 PH/32 Applied Sieve patches from the maintainer.
2417 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2419 PH/33 Added "verify = not_blind".
2421 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2422 Local/Makefile (with some defaults set). These are used in built scripts
2423 such as exicyclog, but they have never been used in the exim_install
2424 script (though there are many overriding facilities there). I have
2425 arranged that the exim_install script now takes note of these two
2428 PH/35 Installed configuration files for Dragonfly.
2430 PH/36 When a locally submitted message by a trusted user did not contain a
2431 From: header, and the sender address was obtained from -f or from an SMTP
2432 MAIL command, and the trusted user did not use -F to supply a sender
2433 name, $originator_name was incorrectly used when constructing a From:
2434 header. Furthermore, $originator_name was used for submission mode
2435 messages from external hosts without From: headers in a similar way,
2436 which is clearly wrong.
2438 PH/37 Added control=suppress_local_fixups.
2440 PH/38 When log_selector = +received_sender was set, and the addition of the
2441 sender made the log line's construction buffer exactly full, or one byte
2442 less than full, an overflow happened when the terminating "\n" was
2445 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2446 when the result of a list match is failure because a DNS lookup failed.
2448 PH/40 RM_COMMAND is now used in the building process.
2450 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2451 the "build-* directories that it finds.
2453 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2454 address in a domain literal was a prefix of an interface address.
2456 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2457 when verifying a sender address, unless rewrite_headers is false.
2459 PH/44 Wrote a long comment about why errors_to addresses are verified as
2460 recipients, not senders.
2462 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2463 the ratelimit ACL was added.
2465 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2467 PH/46 Added extra information about PostgreSQL errors to the error string.
2469 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2470 avoiding the need to install special zones in a real server. This is
2471 backwards compatible; if it can't find the fake resolver, it drops back.
2472 Thus, both old and new test suites can be run.
2474 TF/02 Added util/ratelimit.pl
2476 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2477 clock is set back in time.
2479 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2480 Candler <B.Candler@pobox.com>.
2482 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2483 OK for addresses that are the result of redirections.
2485 PH/48 A number of further additions for the benefit of the new test suite,
2486 including a fake gethostbyname() that interfaces to the fake DNS resolver
2489 TF/06 The fix for widen_domains has also been applied to qualify_single and
2490 search_parents which are the other dnslookup options that can cause
2493 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2496 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2498 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2499 service" if it encountered an MX record with an empty target hostname.
2500 The message is now "an MX or SRV record indicated no SMTP service".
2502 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2503 if the DNS of the sending site is misconfigured. This is quite a
2504 common situation. This change restores the behaviour of treating a
2505 helo verification defer as a failure.
2507 PH/51 If self=fail was set on a router, the bounce message did not include the
2508 actual error message.
2514 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2516 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2517 command itself ended in error, the underlying message about the transport
2518 filter timeout was being overwritten with the pipe command error. Now the
2519 underlying error message should be appended to the second error message.
2521 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2523 PH/02 Reduce the amount of output that "make" produces by default. Full output
2524 can still be requested.
2526 PH/03 The warning log line about a condition test deferring for a "warn" verb
2527 was being output only once per connection, rather than after each
2528 occurrence (because it was using the same function as for successful
2529 "warn" verbs). This seems wrong, so I have changed it.
2531 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2532 it should not have, which might have caused a crash in the right
2533 circumstances, but probably never did.
2535 PH/04 Installed a modified version of Tony Finch's patch to make submission
2536 mode fix the return path as well as the Sender: header line, and to
2537 add a /name= option so that you can make the user's friendly name appear
2540 TF/03 Added the control = fakedefer ACL modifier.
2542 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2543 Mark Lowes for thorough testing.
2545 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2547 TK/03 Merged latest SRS patch from Miles Wilton.
2549 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2550 with the definition in sysexits.h (which is #included earlier).
2551 Fortunately, Exim does not actually use EX_OK. The code used to try to
2552 preserve the sysexits.h value, by assumimg that macro definitions were
2553 scanned for macro replacements. I have been disabused of this notion,
2554 so now the code just undefines EX_OK before #including unistd.h.
2556 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2557 in the smtp transport. When a block could not be written in a single
2558 write() function, the timeout was being re-applied to each part-write.
2559 This seems wrong - if the receiver was accepting one byte at a time it
2560 would take for ever. The timeout is now adjusted when this happens. It
2561 doesn't have to be particularly precise.
2563 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2564 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2566 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2567 without a domain if the check to <postmaster@domain> fails.
2569 SC/01 Eximstats: added -xls and the ability to specify output files
2570 (patch written by Frank Heydlauf).
2572 SC/02 Eximstats: use FileHandles for outputing results.
2574 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2576 SC/04 Eximstats: fixed display of large numbers with -nvr option
2578 SC/05 Eximstats: fixed merging of reports with empty tables.
2580 SC/06 Eximstats: added the -include_original_destination flag
2582 SC/07 Eximstats: removed tabs and trailing whitespace.
2584 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2586 TK/06 MBOX spool code: Add real "From " MBOX separator line
2587 so the .eml file is really in mbox format (even though
2588 most programs do not really care). Patch from Alex Miller.
2590 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2591 The latter is generated from $received_to and is only set if the
2592 message has one envelope recipient. SA can use these headers,
2593 obviously out-of-the-box. Patch from Alex Miller.
2595 PH/08 The ${def test on a variable was returning false if the variable's
2596 value was "0", contrary to what the specification has always said!
2597 The result should be true unless the variable is empty.
2599 PH/09 The syntax error of a character other than { following "${if
2600 def:variable_name" (after optional whitespace) was not being diagnosed.
2601 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2602 accidental colon was present, for example, could give incorrect results.
2604 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2605 result is used (not including appendfile, where other changes are about
2608 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2609 This involved changing a lot of size variables from int to off_t. It
2610 should work with maildirs and everything.
2612 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2613 spamd dying while we are connected to it.
2615 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2618 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2619 function for BDB 4.3.
2621 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2623 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2624 the sources for PCRE have been re-arranged and more files are now
2627 PH/15 The code I had for printing potentially long long variables in PH/11
2628 above was not the best (it lost precision). The length of off_t variables
2629 is now inspected at build time, and an appropriate printing format (%ld
2630 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2631 to be "long long int" or "long int". This is needed for the internal
2632 formatting function string_vformat().
2634 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2635 the configuration file to be ":syslog", then the script "guesses" where
2636 the logs files are, rather than using the compiled in default. In our
2637 case the guess is not the same as the compiled default, so the script
2638 suddenly stopped working when I started to use syslog. The patch checks
2639 to see if log_file_path is "". If so, it attempts to read it from exim
2640 with no configuration file to get the compiled in version, before it
2641 falls back to the previous guessing code."
2643 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2644 implementing BATV in an Exim configuration. See NewStuff for the gory
2647 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2648 Makefile that are specific to HP-UX.
2650 PH/18 If the "use_postmaster" option was set for a recipient callout together
2651 with the "random" option, the postmaster address was used as the MAIL
2652 FROM address for the random test, but not for the subsequent recipient
2653 test. It is now used for both.
2655 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2656 patch removes a few documentation additions to RFC 3028, because the
2657 latest draft now contains them. It adds the new en;ascii-case comparator
2658 and a new error check for 8bit text in MIME parts. Comparator and
2659 require names are now matched exactly. I enabled the subaddress
2660 extension, but it is not well tested yet (read: it works for me)."
2662 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2663 rework some of the code of TK/09 above to avoid the hardwired use of
2664 "%lld" and "long long". Replaced the call to snprintf() with a call to
2667 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2668 records point to non-existent hosts", "retry timeout exceeded", and
2669 "retry time not reached for any host after a long failure period".
2671 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2672 experimental DomainKeys support:
2674 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2675 (2) On an error such as an illegally used "control", the wrong name for
2676 the control was given.
2678 These problems did NOT occur unless DomainKeys support was compiled.
2680 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2682 PH/24 Added ${if match_ip condition.
2684 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2685 databases so that it will be absolutely obvious if a crash occurs in the
2686 DB library. This is a regular occurrence (often caused by mis-matched
2689 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2690 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2691 picked up on a user's system that detects such things. There doesn't seem
2692 to be a gcc warning option for this - only an attribute that has to be
2693 put on the function's prototype. It seems that in Fedora Core 4 they have
2694 set this on a number of new functions. No doubt there will be more in due
2697 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2698 specify a transport. However, if an address that was verified by such a
2699 router was the subject of a callout, Exim crashed because it tried to
2700 read the rcpt_include_affixes from the non-existent transport. Now it
2701 just assumes that the setting of that option is false. This bug was
2702 introduced by 4.51/PH/31.
2704 PH/28 Changed -d+all to exclude +memory, because that information is very
2705 rarely of interest, but it makes the output a lot bigger. People tend to
2706 do -d+all out of habit.
2708 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2709 code in os-type was giving problems when libc.so lives in lib64, like on
2712 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2713 aren't the modern standard, and it seems that some systems' include files
2714 don't always have them. Exim was already checking for some of the newer
2715 ones like T_AAAA, and defining it itself. I've added checks for all the
2716 record types that Exim uses.
2718 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2719 not automatically generating a new one, as it is supposed to. This
2720 prevented TLS from working. If the file did exist, but contained invalid
2721 data, a new version was generated, as expected. It was only the case of a
2722 non-existent file that was broken.
2724 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2725 with a change in libdomainkeys > 0.64.
2727 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2728 from DNS. If the selector record carries the flag, it now has
2729 precedence over the domain-wide flag.
2731 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2733 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2734 the use of a port name that isn't defined in /etc/services) occurred, the
2735 message was deferred as in a normal delivery, and thus remained on the
2736 spool, instead of being failed because of the mua_wrapper setting. This
2737 is now fixed, and I tidied up some of the mua_wrapper messages at the
2740 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2741 the messages in a hash of arrays rather than using individual hashes.
2742 This is a bit cleaner and results in dramatic memory savings, albeit
2743 at a slight CPU cost.
2745 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2746 as requested by Marc Sherman.
2748 SC/10 Eximstats: added histograms for user specified patterns as requested
2751 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2753 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2754 fopen() in the content-scanning modules that did not already have it.
2760 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2761 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2763 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2765 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2767 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2768 to test Sieve filters that use "vacation".
2770 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2771 that changes the way the GnuTLS parameters are stored in the cache file.
2772 The new format can be generated externally. For backward compatibility,
2773 if the data in the cache doesn't make sense, Exim assumes it has read an
2774 old-format file, and it generates new data and writes a new file. This
2775 means that you can't go back to an older release without removing the
2778 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2779 work if there are any delivery delays because "one_time" forces the
2780 parent to be marked "delivered", so its unseen clone is never tried
2781 again. For this reason, Exim now forbids the simultaneous setting of
2784 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2785 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2786 entry if you want to know the details. The fix, however, neglected to
2787 consider the case when local delivery batching is involved. The test for
2788 "previously delivered" was not happening when checking to see if an
2789 address could be batched with a previous (undelivered) one; under
2790 certain circumstances this could lead to multiple deliveries to the same
2793 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2794 in its include files, and this causes problems building Exim.
2796 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2797 header_syntax) but Exim was just ignoring anything given after a slash.
2798 In particular, this caused confusion with an attempt to use "verify =
2799 reverse_host_lookup/defer_ok". An error is now given when options are
2800 supplied for verify items that do not have them. (Maybe reverse_host_
2801 lookup should have a defer_ok option, but that's a different point.)
2803 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2804 defined by RFC 821) to 2048, because there were problems with some AUTH
2805 commands, and RFC 1869 says the size should be increased for extended
2806 SMTP commands that take arguments.
2808 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2811 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2812 "unknown" error; now it says that the functionality isn't in the binary.
2814 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2815 an address' error message when a string expansion fails (syntax or
2816 whatever). Otherwise the password may appear in the log. Following change
2817 PH/42 below, there is no longer a chance of it appearing in a bounce
2820 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2822 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2823 of its addresses was ever tried. (Bugzilla bug #2.)
2825 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2826 the result incorrectly in the debug output. (It correctly added a newline
2827 to what was transported.)
2829 TF/01 Added $received_time.
2831 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2832 commented out examples of how to interface to a virus scanner and to
2833 SpamAssassin. Also added commented examples of av_scanner and
2834 spamd_address settings.
2836 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2837 and controls are allowed in which ACLs. There were a couple of minor
2838 errors. Some of the entries in the conditions table (which is a table of
2839 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2840 negation of where the condition IS allowed.
2842 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2844 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2845 header file does not have a version number, so I've had to invent a new
2846 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2847 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2848 radiusclient), but it was contributed by a Radius user.
2850 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2851 files or directories, for OpenSSL.
2853 PH/19 When an Exim process that is running as root has to create an Exim log
2854 file, it does so in a subprocess that runs as exim:exim so as to get the
2855 ownership right at creation (otherwise, other Exim processes might see
2856 the file with the wrong ownership). There was no test for failure of this
2857 fork() call, which would lead to the process getting stuck as it waited
2858 for a non-existent subprocess. Forks do occasionally fail when resources
2859 run out. I reviewed all the other calls to fork(); they all seem to check
2862 PH/20 When checking for unexpected SMTP input at connect time (before writing
2863 the banner), Exim was not dealing correctly with a non-positive return
2864 from the read() function. If the client had disconnected by this time,
2865 the result was a log entry for a synchronization error with an empty
2866 string after "input=" when read() returned zero. If read() returned -1
2867 (an event I could not check), uninitialized data bytes were printed.
2868 There were reports of junk text (parts of files, etc) appearing after
2871 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2873 PH/22 Added support for macro redefinition, and (re)definition in between
2874 driver and ACL definitions.
2876 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2877 forgetting to use the resulting value; it was using the unexpanded value.
2879 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2880 hadn't been configured. The fix is from Juergen Kreileder, who
2881 understands it better than I do:
2883 "Here's what I see happening with three configured cyrus_sasl
2884 authenticators configured (plain, login, cram-md5):
2886 On startup auth_cyrus_sasl_init() gets called for each of these.
2887 This means three calls to sasl_listmech() without a specified mech_list.
2888 => SASL tests which mechs of all available mechs actually work
2889 => three warnings about OTP not working
2890 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2892 With the patch, sasl_listmech() also gets called three times. But now
2893 SASL's mech_list option is set to the server_mech specified in the the
2894 authenticator. Or in other words, the answer from sasl_listmech()
2895 gets limited to just the mech you're testing for (which is different
2897 => the return list contains just 'plain' or 'login', 'cram-md5' or
2898 nothing depending on the value of ob->server_mech.
2900 I've just tested the patch: Authentication still works fine,
2901 unavailable mechs specified in the exim configuration are still
2902 caught, and the auth.log warnings about OTP are gone."
2904 PH/25 When debugging is enabled, the contents of the command line are added
2905 to the debugging output, even when log_selector=+arguments is not
2908 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2909 answer is "GNU", and only if the return is "GNU/something" is the answer
2912 PH/27 $acl_verify_message is now set immediately after the failure of a
2913 verification in an ACL, and so is available in subsequent modifiers. In
2914 particular, the message can be preserved by coding like this:
2916 warn !verify = sender
2917 set acl_m0 = $acl_verify_message
2919 Previously, $acl_verify_message was set only while expanding "message"
2920 and "log_message" when a very denied access.
2922 PH/28 Modified OS/os.c-Linux with
2924 -#ifndef OS_LOAD_AVERAGE
2925 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2927 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2928 nomenclature these days.)
2930 PH/29 Installed patch from the Sieve maintainer that adds the options
2931 sieve_useraddress and sieve_subaddress to the redirect router.
2933 PH/30 In these circumstances:
2934 . Two addresses routed to the same list of hosts;
2935 . First host does not offer TLS;
2936 . First host accepts first address;
2937 . First host gives temporary error to second address;
2938 . Second host offers TLS and a TLS session is established;
2939 . Second host accepts second address.
2940 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2941 and peerdn, if requested) that were in fact used only for the second
2944 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2945 attention to any local part prefix or suffix that was matched by the
2946 router that accepted the address. It now behaves in the same way as it
2947 does for delivery: the affixes are removed from the local part unless
2948 rcpt_include_affixes is set on the transport.
2950 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2951 timeout during the DATA phase of an incoming message.
2953 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2954 applied a patch sanctioned by the Sieve maintainer.
2956 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2957 the uid or gid is negative. A case of a negative gid caused this to be
2958 noticed. The fix allows for either to be negative.
2960 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2961 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2964 PH/36 The change PH/12 above was broken. Fixed it.
2966 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2967 the grounds that routing the same address twice would always produce the
2968 same answer. This might have been true once, but it is certainly no
2969 longer true now. Routing a child address may depend on the previous
2970 routing that produced that child. Some complicated redirection strategies
2971 went wrong when messages had multiple recipients, and made Exim's
2972 behaviour dependent on the order in which the addresses were given.
2974 I have moved the duplicate checking until after the routing is complete.
2975 Exim scans the addresses that are assigned to local and remote
2976 transports, and removes any duplicates. This means that more work will be
2977 done, as duplicates will always all be routed, but duplicates are
2978 presumably rare, so I don't expect this is of any significance.
2980 For deliveries to pipes, files, and autoreplies, the duplicate checking
2981 still happens during the routing process, since they are not going to be
2984 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2985 It corrects a timeout issue with spamd. This is Ian's comment: "The
2986 background is that sometimes spamd either never reads data from a
2987 connection it has accepted, or it never writes response data. The exiscan
2988 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2989 blindly assumes that writes won't block so it may never time out."
2991 PH/39 Allow G after quota size as well as K and M.
2993 PH/40 The value set for $authenticated_id in an authenticator may not contain
2994 binary zeroes or newlines because the value is written to log lines and
2995 to spool files. There was no check on this. Now the value is run through
2996 the string_printing() function so that such characters are converted to
2997 printable escape sequences.
2999 PH/41 $message_linecount is a new variable that contains the total number of
3000 lines in the message. Compare $body_linecount, which is the count for the
3003 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3004 bounce and delay warning messages, except in certain special cases, which
3007 (a) An SMTP error message from a remote host;
3008 (b) A message specified in a :fail: redirection;
3009 (c) A message specified in a "fail" command in a system filter;
3010 (d) A message specified in a FAIL return from the queryprogram router;
3011 (e) A message specified by the cannot_route_message router option.
3013 In these cases only, Exim does include the error details in bounce and
3014 warning messages. There are also a few cases where bland messages such
3015 as "unrouteable address" or "local delivery error" are given.
3017 PH/43 $value is now also set for the "else" part of a ${run expansion.
3019 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3020 being worked on, but at least Exim now implements the latest version to
3023 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3024 process to complete was treated as a delivery failure, a timeout while
3025 writing the message to the pipe was logged, but erroneously treated as a
3026 successful delivery. Such timeouts include transport filter timeouts. For
3027 consistency with the overall process timeout, these timeouts are now
3028 treated as errors, giving rise to delivery failures by default. However,
3029 there is now a new Boolean option for the pipe transport called
3030 timeout_defer, which, if set TRUE, converts the failures into defers for
3031 both kinds of timeout. A transport filter timeout is now identified in
3034 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3035 systems where "make" and "gmake" are different, calling "gmake" at top
3036 level broke things. I've arranged for the value of $(MAKE) to be passed
3037 from the Makefile to this script so that it can call the same version of
3041 A note about Exim versions 4.44 and 4.50
3042 ----------------------------------------
3044 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3045 changes of various kinds. As a consequence, a big documentation update was
3046 needed. This delayed the release for rather longer than seemed good, especially
3047 in the light of a couple of (minor) security issues. Therefore, the changes
3048 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3049 release. So 4.44 and 4.50 are in effect two different branches that both start
3052 I have left the 4.50 change log unchanged; it contains all the changes since
3053 4.43. The change log for 4.44 is below; many of its items are identical to
3054 those for 4.50. This seems to be the most sensible way to preserve the
3055 historical information.
3061 1. Minor wording change to the doc/README.SIEVE file.
3063 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3064 computation of the current number of files was incorrect.
3066 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3067 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3070 4. Give more explanation in the error message when the command for a transport
3071 filter fails to execute.
3073 5. There are several places where Exim runs a non-Exim command in a
3074 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3075 was being done only for the command run by the queryprogram router. It is
3076 now done for all such subprocesses. The other cases are: ${run, transport
3077 filters, and the commands run by the lmtp and pipe transports.
3079 6. Added CONFIGURE_GROUP build-time option.
3081 7. Some older OS have a limit of 256 on the maximum number of file
3082 descriptors. Exim was using setrlimit() to set 1000 as a large value
3083 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3084 systems. I've change it so that if it can't get 1000, it tries for 256.
3086 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3087 was an oversight, and furthermore, ever since the addition of extra
3088 controls (e.g. 4.43/32), the checks on when to allow different forms of
3089 "control" were broken. There should now be diagnostics for all cases when a
3090 control that does not make sense is encountered.
3092 9. Added the /retain_sender option to "control=submission".
3094 10. $recipients is now available in the predata ACL (oversight).
3096 11. Tidy the search cache before the fork to do a delivery from a message
3097 received from the command line. Otherwise the child will trigger a lookup
3098 failure and thereby defer the delivery if it tries to use (for example) a
3099 cached ldap connection that the parent has called unbind on.
3101 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3102 of $address_data from the recipient verification was clobbered by the
3103 sender verification.
3105 13. The value of address_data from a sender verification is now available in
3106 $sender_address_data in subsequent conditions in the ACL statement.
3108 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3110 15. Added a new option "connect=<time>" to callout options, to set a different
3113 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3114 was its contents. (It was OK if the option was not defined at all.)
3116 17. A "Completed" log line is now written for messages that are removed from
3117 the spool by the -Mrm option.
3119 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3120 information about exactly what failed.
3122 19. Added -dd to debug only the daemon process.
3124 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3125 handles timeouts, both on the server side and network timeouts. Renamed the
3126 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3128 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3129 It is now set to "smtps".
3131 22. $host_address is now set to the target address during the checking of
3132 ignore_target_hosts.
3134 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3135 being passed; this would have caused $sender_host_name to have been used if
3136 matching the list had actually called for a host name (not very likely,
3137 since this list is usually IP addresses). A host name is now passed as
3140 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3141 code that specifies a non-restarting handler (typically sigaction() in
3142 modern systems) in an attempt to fix a rare and obscure crash bug.
3144 25. Narrowed the window for a race in the daemon that could cause it to ignore
3145 SIGCHLD signals. This is not a major problem, because they are used only to
3146 wake it up if nothing else does.
3148 26. A malformed maildirsize file could cause Exim to calculate negative values
3149 for the mailbox size or file count. Odd effects could occur as a result.
3150 The maildirsize information is now recalculated if the size or filecount
3153 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3154 support for a long time. Removed HAVE_SYS_VFS_H.
3156 28. Installed the latest version of exipick from John Jetmore.
3158 29. In an address list, if the pattern was not a regular expression, an empty
3159 subject address (from a bounce message) matched only if the pattern was an
3160 empty string. Non-empty patterns were not even tested. This was the wrong
3161 because it is perfectly reasonable to use an empty address as part of a
3162 database query. An empty address is now tested by patterns that are
3163 lookups. However, all the other forms of pattern expect the subject to
3164 contain a local part and a domain, and therefore, for them, an empty
3165 address still always fails if the pattern is not itself empty.
3167 30. Exim went into a mad DNS loop when attempting to do a callout where the
3168 host was specified on an smtp transport, and looking it up yielded more
3169 than one IP address.
3171 31. Re-factored the code for checking spool and log partition space into a
3172 function that finds that data and another that does the check. The former
3173 is then used to implement four new variables: $spool_space, $log_space,
3174 $spool_inodes, and $log_inodes.
3176 32. The RFC2047 encoding function was originally intended for short strings
3177 such as real names; it was not keeping to the 75-character limit for
3178 encoded words that the RFC imposes. It now respects the limit, and
3179 generates multiple encoded words if necessary. To be on the safe side, I
3180 have increased the buffer size for the ${rfc2047: expansion operator from
3183 33. It is now permitted to omit both strings after an "if" condition; if the
3184 condition is true, the result is "true". As before, when the second string
3185 is omitted, a false condition yields an empty string. This makes it less
3186 cumbersome to write custom ACL and router conditions.
3188 34. Failure to deliver a bounce message always caused it to be frozen, even if
3189 there was an errors_to setting on the router. The errors_to setting is now
3192 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3193 canonical form (fully expanded) before being placed in
3194 $sender_host_address.
3196 36. The table in the code that translates DNS record types into text (T_A to
3197 "A" for instance) was missing entries for NS and CNAME. It is just possible
3198 that this could have caused confusion if both these types were looked up
3199 for the same domain, because the text type is used as part of Exim's
3200 per-process caching. But the chance of anyone hitting this buglet seems
3203 37. The dnsdb lookup has been extended in a number of ways.
3205 (1) There is a new type, "zns", which walks up the domain tree until it
3206 finds some nameserver records. It should be used with care.
3208 (2) There is a new type, "mxh", which is like "mx" except that it returns
3209 just the host names, not the priorities.
3211 (3) It is now possible to give a list of domains (or IP addresses) to be
3212 looked up. The behaviour when one of the lookups defers can be
3213 controlled by a keyword.
3215 (4) It is now possible to specify the separator character for use when
3216 multiple records are returned.
3218 38. The dnslists ACL condition has been extended: it is now possible to supply
3219 a list of IP addresses and/or domains to be looked up in a particular DNS
3222 39. Added log_selector=+queue_time_overall.
3224 40. When running the queue in the test harness, wait just a tad after forking a
3225 delivery process, to get repeatability of debugging output.
3227 41. Include certificate and key file names in error message when GnuTLS fails
3228 to set them up, because the GnuTLS error message doesn't include the name
3229 of the failing file when there is a problem reading it.
3231 42. Allow both -bf and -bF in the same test run.
3233 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3235 44. The "Exiscan patch" is now merged into the mainline Exim source.
3237 45. Sometimes the final signoff response after QUIT could fail to get
3238 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3239 < 0 before doing a fflush(). This bug looks as though it goes back to the
3240 introduction of TLS in release 3.20, but "sometimes" must have been rare
3241 because the tests only now provoked it.
3243 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3244 (this can affect the format of dates).
3246 47. exim_tidydb, when checking for the continued existence of a message for
3247 which it has found a message-specific retry record, was not finding
3248 messages that were in split spool directories. Consequently, it was
3249 deleting retry records that should have stayed in existence.
3251 48. Steve fixed some bugs in eximstats.
3253 49. The SPA authentication driver was not abandoning authentication and moving
3254 on to the next authenticator when an expansion was forced to fail,
3255 contradicting the general specification for all authenticators. Instead it
3256 was generating a temporary error. It now behaves as specified.
3258 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3259 (the order specifies the preference for clients). The order is now AES256,
3260 AES128, 3DES, ARCFOUR128.
3262 51. Small patch to Sieve code - explicitly set From: when generating an
3265 52. Exim crashed if a remote delivery caused a very long error message to be
3266 recorded - for instance if somebody sent an entire SpamAssassin report back
3267 as a large number of 550 error lines. This bug was coincidentally fixed by
3268 increasing the size of one of Exim's internal buffers (big_buffer) that
3269 happened as part of the Exiscan merge. However, to be on the safe side, I
3270 have made the code more robust (and fixed the comments that describe what
3273 53. Now that there can be additional text after "Completed" in log lines (if
3274 the queue_time_overall log selector is set), a one-byte patch to exigrep
3275 was needed to allow it to recognize "Completed" as not the last thing in
3278 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3279 patch that reportedly fixes this has been added. I am not expert enough to
3280 create a test for it. This is what the patch creator wrote:
3282 "I found a little strange behaviour of ldap code when working with
3283 Windows 2003 AD Domain, where users was placed in more than one
3284 Organization Units. When I tried to give exim partial DN, the exit code
3285 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3286 But simultaneously result of request was absolutely normal ldap result,
3287 so I produce this patch..."
3289 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3290 REFERENCE, so I have modified the code to exclude the patch when that macro
3293 55. Some experimental protocols are using DNS PTR records for new purposes. The
3294 keys for these records are domain names, not reversed IP addresses. The
3295 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3296 leaves it alone. Component reversal etc. now happens only for IP addresses.
3299 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3301 57. Double the size of the debug message buffer (to 2048) so that more of very
3302 long debug lines gets shown.
3304 58. The exicyclog utility now does better if the number of log files to keep
3305 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3307 59. Two changes related to the smtp_active_hostname option:
3309 (1) $smtp_active_hostname is now available as a variable.
3310 (2) The default for smtp_banner uses $smtp_active_hostname instead
3311 of $primary_hostname.
3313 60. The host_aton() function is supposed to be passed a string that is known
3314 to be a valid IP address. However, in the case of IPv6 addresses, it was
3315 not checking this. This is a hostage to fortune. Exim now panics and dies
3316 if the condition is not met. A case was found where this could be provoked
3317 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3318 components; fortuitously, this particular loophole had already been fixed
3319 by change 4.50/55 above.
3321 If there are any other similar loopholes, the new check in host_aton()
3322 itself should stop them being exploited. The report I received stated that
3323 data on the command line could provoke the exploit when Exim was running as
3324 exim, but did not say which command line option was involved. All I could
3325 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3326 running as the user.
3329 61. There was a buffer overflow vulnerability in the SPA authentication code
3330 (which came originally from the Samba project). I have added a test to the
3331 spa_base64_to_bits() function which I hope fixes it.
3334 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3335 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3337 63. The daemon start-up calls getloadavg() while still root for those OS that
3338 need the first call to be done as root, but it missed one case: when
3339 deliver_queue_load_max is set with deliver_drop_privilege. This is
3340 necessary for the benefit of the queue runner, because there is no re-exec
3341 when deliver_drop_privilege is set.
3343 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3344 This has been fixed.
3346 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3347 was in use, was not putting the data itself into the right store pool;
3348 consequently, it could be overwritten for a subsequent message in the same
3349 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3352 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3354 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3355 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3356 the function were treating the return as a boolean value, which happened to
3357 work because 0=false and not-0=true, but is not correct code.
3359 68. The host_aton() function was not handling scoped IPv6 addresses (those
3360 with, for example, "%eth0" on the end) correctly.
3362 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3363 negated items (that is, ~something) in unsigned ints. Some compilers
3364 apparently mutter when there is no cast.
3366 70. If an address verification called from an ACL failed, and did not produce a
3367 user-specific message (i.e. there was only a "system" message), nothing was
3368 put in $acl_verify_message. In this situation, it now puts the system
3371 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3372 session; change 4.31/43 added the unwanted input to the log line - except
3373 that it did not do this in the start of session case. It now does.
3375 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3376 This is wrong and can cause the other end to generate a synchronization
3377 error if it is another Exim or anything else that does the synchronization
3378 check. A QUIT command is no longer sent after a timeout.
3380 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3381 during host lookups.
3383 74. The defer_ok option of callout verification was not working if it was used
3384 when verifying addresses in header lines, that is, for this case:
3386 verify = header_sender/callout=defer_ok
3388 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3389 those file descriptors could be used for SMTP connections. If anything
3390 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3391 could be sent to the SMTP client, causing chaos. The daemon now opens
3392 stdin, stdout, and stderr to /dev/null when it puts itself into the
3395 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3396 log by default. The user can override this with suitable Perl magic.
3398 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3399 the log message when discard triggers, was not working for the DATA ACL or
3400 for the non-SMTP ACL.
3402 78. Error message wording change in sieve.c.
3404 79. If smtp_accept_max_per_host was set, the number of connections could be
3405 restricted to fewer than expected, because the daemon was trying to set up
3406 a new connection before checking whether the processes handling previous
3407 connections had finished. The check for completed processes is now done
3408 earlier. On busy systems, this bug wouldn't be noticed because something
3409 else would have woken the daemon, and it would have reaped the completed
3412 80. If a message was submitted locally by a user whose login name contained one
3413 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3414 It caused a spool format error. I have fixed the spool reading code. A
3415 related problem was that the "from" clause in the Received: line became
3416 illegal because of the space(s). It is now covered by ${quote_local_part.
3418 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3421 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3423 83. Patch from Sieve maintainer to fix unterminated string problem in
3424 "vacation" handling.
3426 84. Some minor changes to the Linux configuration files to help with other
3427 OS variants using glibc.
3429 85. One more patch for Sieve to update vacation handling to latest spec.
3432 ----------------------------------------------------
3433 See the note above about the 4.44 and 4.50 releases.
3434 ----------------------------------------------------
3440 1. Change 4.43/35 introduced a bug that caused file counts to be
3441 incorrectly computed when quota_filecount was set in an appendfile
3444 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3445 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3448 3. Give more explanation in the error message when the command for a transport
3449 filter fails to execute.
3451 4. There are several places where Exim runs a non-Exim command in a
3452 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3453 was being done only for the command run by the queryprogram router. It is
3454 now done for all such subprocesses. The other cases are: ${run, transport
3455 filters, and the commands run by the lmtp and pipe transports.
3457 5. Some older OS have a limit of 256 on the maximum number of file
3458 descriptors. Exim was using setrlimit() to set 1000 as a large value
3459 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3460 systems. I've change it so that if it can't get 1000, it tries for 256.
3462 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3463 was an oversight, and furthermore, ever since the addition of extra
3464 controls (e.g. 4.43/32), the checks on when to allow different forms of
3465 "control" were broken. There should now be diagnostics for all cases when a
3466 control that does not make sense is encountered.
3468 7. $recipients is now available in the predata ACL (oversight).
3470 8. Tidy the search cache before the fork to do a delivery from a message
3471 received from the command line. Otherwise the child will trigger a lookup
3472 failure and thereby defer the delivery if it tries to use (for example) a
3473 cached ldap connection that the parent has called unbind on.
3475 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3476 of $address_data from the recipient verification was clobbered by the
3477 sender verification.
3479 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3480 was its contents. (It was OK if the option was not defined at all.)
3482 11. A "Completed" log line is now written for messages that are removed from
3483 the spool by the -Mrm option.
3485 12. $host_address is now set to the target address during the checking of
3486 ignore_target_hosts.
3488 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3489 being passed; this would have caused $sender_host_name to have been used if
3490 matching the list had actually called for a host name (not very likely,
3491 since this list is usually IP addresses). A host name is now passed as
3494 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3495 code that specifies a non-restarting handler (typically sigaction() in
3496 modern systems) in an attempt to fix a rare and obscure crash bug.
3498 15. Narrowed the window for a race in the daemon that could cause it to ignore
3499 SIGCHLD signals. This is not a major problem, because they are used only to
3500 wake it up if nothing else does.
3502 16. A malformed maildirsize file could cause Exim to calculate negative values
3503 for the mailbox size or file count. Odd effects could occur as a result.
3504 The maildirsize information is now recalculated if the size or filecount
3507 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3508 support for a long time. Removed HAVE_SYS_VFS_H.
3510 18. Updated exipick to current release from John Jetmore.
3512 19. Allow an empty sender to be matched against a lookup in an address list.
3513 Previously the only cases considered were a regular expression, or an
3516 20. Exim went into a mad DNS lookup loop when doing a callout where the
3517 host was specified on the transport, if the DNS lookup yielded more than
3520 21. The RFC2047 encoding function was originally intended for short strings
3521 such as real names; it was not keeping to the 75-character limit for
3522 encoded words that the RFC imposes. It now respects the limit, and
3523 generates multiple encoded words if necessary. To be on the safe side, I
3524 have increased the buffer size for the ${rfc2047: expansion operator from
3527 22. Failure to deliver a bounce message always caused it to be frozen, even if
3528 there was an errors_to setting on the router. The errors_to setting is now
3531 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3532 canonical form (fully expanded) before being placed in
3533 $sender_host_address.
3535 24. Updated eximstats to version 1.33
3537 25. Include certificate and key file names in error message when GnuTLS fails
3538 to set them up, because the GnuTLS error message doesn't include the name
3539 of the failing file when there is a problem reading it.
3541 26. Expand error message when OpenSSL has problems setting up cert/key files.
3544 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3545 (this can affect the format of dates).
3547 28. exim_tidydb, when checking for the continued existence of a message for
3548 which it has found a message-specific retry record, was not finding
3549 messages that were in split spool directories. Consequently, it was
3550 deleting retry records that should have stayed in existence.
3552 29. eximstats updated to version 1.35
3553 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3554 1.35 - bugfix such that pie charts by volume are generated correctly
3556 30. The SPA authentication driver was not abandoning authentication and moving
3557 on to the next authenticator when an expansion was forced to fail,
3558 contradicting the general specification for all authenticators. Instead it
3559 was generating a temporary error. It now behaves as specified.
3561 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3562 (the order specifies the preference for clients). The order is now AES256,
3563 AES128, 3DES, ARCFOUR128.
3565 31. Small patch to Sieve code - explicitly set From: when generating an
3568 32. Exim crashed if a remote delivery caused a very long error message to be
3569 recorded - for instance if somebody sent an entire SpamAssassin report back
3570 as a large number of 550 error lines. This bug was coincidentally fixed by
3571 increasing the size of one of Exim's internal buffers (big_buffer) that
3572 happened as part of the Exiscan merge. However, to be on the safe side, I
3573 have made the code more robust (and fixed the comments that describe what
3576 33. Some experimental protocols are using DNS PTR records for new purposes. The
3577 keys for these records are domain names, not reversed IP addresses. The
3578 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3579 leaves it alone. Component reversal etc. now happens only for IP addresses.
3582 34. The host_aton() function is supposed to be passed a string that is known
3583 to be a valid IP address. However, in the case of IPv6 addresses, it was
3584 not checking this. This is a hostage to fortune. Exim now panics and dies
3585 if the condition is not met. A case was found where this could be provoked
3586 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3587 components; fortuitously, this particular loophole had already been fixed
3588 by change 4.50/55 or 4.44/33 above.
3590 If there are any other similar loopholes, the new check in host_aton()
3591 itself should stop them being exploited. The report I received stated that
3592 data on the command line could provoke the exploit when Exim was running as
3593 exim, but did not say which command line option was involved. All I could
3594 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3595 running as the user.
3598 35. There was a buffer overflow vulnerability in the SPA authentication code
3599 (which came originally from the Samba project). I have added a test to the
3600 spa_base64_to_bits() function which I hope fixes it.
3603 36. The daemon start-up calls getloadavg() while still root for those OS that
3604 need the first call to be done as root, but it missed one case: when
3605 deliver_queue_load_max is set with deliver_drop_privilege. This is
3606 necessary for the benefit of the queue runner, because there is no re-exec
3607 when deliver_drop_privilege is set.
3609 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3610 was in use, was not putting the data itself into the right store pool;
3611 consequently, it could be overwritten for a subsequent message in the same
3612 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3615 38. Sometimes the final signoff response after QUIT could fail to get
3616 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3617 < 0 before doing a fflush(). This bug looks as though it goes back to the
3618 introduction of TLS in release 3.20, but "sometimes" must have been rare
3619 because the tests only now provoked it.
3625 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3626 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3627 Now it returns an integer. A number of calls were still expecting a T/F
3628 return. Fortuitously, in all cases, the tests worked in OK situations,
3629 which is the norm. However, things would have gone wrong on any write
3630 failures on the smtp file descriptor. This function is used when sending
3631 messages over SMTP and also when doing verify callouts.
3633 2. When Exim is called to do synchronous delivery of a locally submitted
3634 message (the -odf or -odi options), it no longer closes stderr before doing
3637 3. Implemented the mua_wrapper option.
3639 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3641 5. Implemented the functions header_remove(), header_testname(),
3642 header_add_at_position(), and receive_remove_recipient(), and exported them
3645 6. If an ACL "warn" statement specified the addition of headers, Exim already
3646 inserted X-ACL-Warn: at the start if there was no header name. However, it
3647 was not making this test for the second and subsequent header lines if
3648 there were newlines in the string. This meant that an invalid header could
3649 be inserted if Exim was badly configured.
3651 7. Allow an ACL "warn" statement to add header lines at the start or after all
3652 the Received: headers, as well as at the end.
3654 8. Added the rcpt_4xx retry error code.
3656 9. Added postmaster_mailfrom=xxx to callout verification option.
3658 10. Added mailfrom=xxxx to the callout verification option, for verify=
3661 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3662 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3664 12. Inserted some casts to stop certain compilers warning when using pointer
3665 differences as field lengths or precisions in printf-type calls (mostly
3666 affecting debugging statements).
3668 13. Added optional readline() support for -be (dynamically loaded).
3670 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3671 same clock tick as a message's arrival, so that its received time was the
3672 same as the "first fail" time on the retry record, and that message
3673 remained on the queue past the ultimate address timeout, every queue runner
3674 would try a delivery (because it was past the ultimate address timeout) but
3675 after another failure, the ultimate address timeout, which should have then
3676 bounced the address, did not kick in. This was a "< instead of <=" error;
3677 in most cases the first failure would have been in the next clock tick
3678 after the received time, and all would be well.
3680 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3681 being recognized when the domain list was tested by the match_domain
3682 condition in an expansion string.
3684 16. Added the ${str2b64: operator.
3686 17. Exim was always calling setrlimit() to set a large limit for the number of
3687 processes, without checking whether the existing limit was already
3688 adequate. (It did check for the limit on file descriptors.) Furthermore,
3689 errors from getrlimit() and setrlimit() were being ignored. Now they are
3690 logged to the main and panic logs, but Exim does carry on, to try to do its
3691 job under whatever limits there are.
3693 18. Imported PCRE 5.0.
3695 19. Trivial typo in log message " temporarily refused connection" (the leading
3698 20. If the log selector return_path_on_delivery was set and an address was
3699 redirected to /dev/null, the delivery process crashed because it assumed
3700 that a return path would always be set for a "successful" delivery. In this
3701 case, the whole delivery is bypassed as an optimization, and therefore no
3704 21. Internal re-arrangement: the function for sending a challenge and reading
3705 a response while authentication was assuming a zero-terminated challenge
3706 string. It's now changed to take a pointer and a length, to allow for
3707 binary data in such strings.
3709 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3711 23. Exim was not respecting finduser_retries when seeking the login of the
3712 uid under which it was called; it was always trying 10 times. (The default
3713 setting of finduser_retries is zero.) Also, it was sleeping after the final
3714 failure, which is pointless.
3716 24. Implemented tls_on_connect_ports.
3718 25. Implemented acl_smtp_predata.
3720 26. If the domain in control=submission is set empty, Exim assumes that the
3721 authenticated id is a complete email address when it generates From: or
3722 Sender: header lines.
3724 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3725 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3726 chown and chgrp in /bin and hostname in /usr/bin.
3728 28. Exim was keeping the "process log" file open after each use, just as it
3729 does for the main log. This opens the possibility of it remaining open for
3730 long periods when the USR1 signal hits a daemon. Occasional processlog
3731 errors were reported, that could have been caused by this. Anyway, it seems
3732 much more sensible not to leave this file open at all, so that is what now
3735 29. The long-running daemon process does not normally write to the log once it
3736 has entered its main loop, and it closes the log before doing so. This is
3737 so that log files can straightforwardly be renamed and moved. However,
3738 there are a couple of unusual error situations where the daemon does write
3739 log entries, and I had neglected to close the log afterwards.
3741 30. The text of an SMTP error response that was received during a remote
3742 delivery was being truncated at 512 bytes. This is too short for some of
3743 the long messages that one sometimes sees. I've increased the limit to
3746 31. It is now possible to make retry rules that apply only when a message has a
3747 specific sender, in particular, an empty sender.
3749 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3750 it possible to be selective about when SMTP synchronization is enforced.
3752 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3754 32. Implemented hosts_connection_nolog.
3756 33. Added an ACL for QUIT.
3758 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3761 35. Added mailbox_size and mailbox_filecount to appendfile.
3763 36. Added control = no_multiline_responses to ACLs.
3765 37. There was a bug in the logic of the code that waits for the clock to tick
3766 in the case where the clock went backwards by a substantial amount such
3767 that the microsecond fraction of "now" was more than the microsecond
3768 fraction of "then" (but the whole seconds number was less).
3770 38. Added support for the libradius Radius client library this is found on
3771 FreeBSD (previously only the radiusclient library was supported).
3777 1. When certain lookups returned multiple values in the form name=value, the
3778 quoting of the values was not always being done properly. Specifically:
3779 (a) If the value started with a double quote, but contained no whitespace,
3781 (b) If the value contained whitespace other than a space character (i.e.
3782 tabs or newlines or carriage returns) it was not quoted.
3783 This fix has been applied to the mysql and pgsql lookups by writing a
3784 separate quoting function and calling it from the lookup code. The fix
3785 should probably also be applied to nisplus, ibase and oracle lookups, but
3786 since I cannot test any of those, I have not disturbed their existing code.
3788 2. A hit in the callout cache for a specific address caused a log line with no
3789 reason for rejecting RCPT. Now it says "Previous (cached) callout
3790 verification failure".
3792 3. There was an off-by-one bug in the queryprogram router. An over-long
3793 return line was truncated at 256 instead of 255 characters, thereby
3794 overflowing its buffer with the terminating zero. As well as fixing this, I
3795 have increased the buffer size to 1024 (and made a note to document this).
3797 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3798 when Exim is waiting for an SMTP response from a remote server, Exim
3799 restarts its select() call on the socket, thereby resetting its timeout.
3800 This is not a problem when such interrupts are rare. Somebody set up a cron
3801 job to run exiwhat every 2 minutes, which is less than the normal select()
3802 timeout (5 or 10 minutes). This meant that the select() timeout never
3803 kicked in because it was always reset. I have fixed this by comparing the
3804 time when an interrupt arrives with the time at the start of the first call
3805 to select(). If more time than the timeout has elapsed, the interrupt is
3806 treated as a timeout.
3808 5. Some internal re-factoring in preparation for the addition of Sieve
3809 extensions (by MH). In particular, the "personal" test is moved to a
3810 separate function, and given an option for scanning Cc: and Bcc: (which is
3811 not set for Exim filters).
3813 6. When Exim created an email address using the login of the caller as the
3814 local part (e.g. when creating a From: or Sender: header line), it was not
3815 quoting the local part when it contained special characters such as @.
3817 7. Installed new OpenBSD configuration files.
3819 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3820 try to make them clearer.
3822 9. Callout options, other than the timeout value, were being ignored when
3823 verifying sender addresses in header lines. For example, when using
3825 verify = header_sender/callout=no_cache
3827 the cache was (incorrectly) being used.
3829 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3830 only the Cygwin environment.
3832 11. When return_path_on_delivery was set as a log selector, if different remote
3833 addresses in the same message used different return paths and parallel
3834 remote delivery occurred, the wrong values would sometimes be logged.
3835 (Whenever a remote delivery process finished, the return path value from
3836 the most recently started remote delivery process was logged.)
3838 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3839 lines when AUTH and/or TLS are in use. This is the "received protocol"
3840 field. Exim used to use "asmtp" for authenticated SMTP, without any
3841 indication (in the protocol name) for TLS use. Now it follows the RFC and
3842 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3843 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3844 names appear in log lines as well as in Received: header lines.
3846 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3847 extensions, and comparison tests, and to fix some bugs.
3849 14. Changes to the "personal" filter test:
3851 (1) The test was buggy in that it was just doing the equivalent of
3852 "contains" tests on header lines. For example, if a user's address was
3853 anne@some.where, the "personal" test would incorrectly be true for
3855 To: susanne@some.where
3857 This test is now done by extracting each address from the header in turn,
3858 and checking the entire address. Other tests that are part of "personal"
3859 are now done using regular expressions (for example, to check local parts
3860 of addresses in From: header lines).
3862 (2) The list of non-personal local parts in From: addresses has been
3863 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3864 taken from the Sieve specification recommendations.
3866 (3) If the message contains any header line starting with "List-" it is
3867 treated as non-personal.
3869 (4) The test for "circular" in the Subject: header line has been removed
3870 because it now seems ill-conceived.
3872 15. Minor typos in src/EDITME comments corrected.
3874 16. Installed latest exipick from John Jetmore.
3876 17. If headers_add on a router specified a text string that was too long for
3877 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3878 of string_sprintf() is now avoided.
3880 18. $message_body_size was not set (it was always zero) when running the DATA
3881 ACL and the local_scan() function.
3883 19. For the "mail" command in an Exim filter, no default was being set for
3884 the once_repeat time, causing a random time value to be used if "once" was
3885 specified. (If the value happened to be <= 0, no repeat happened.) The
3886 default is now 0s, meaning "never repeat". The "vacation" command was OK
3887 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3888 (I found it when inspecting the code).
3890 20. There is now an overall timeout for performing a callout verification. It
3891 defaults to 4 times the callout timeout, which applies to individual SMTP
3892 commands during the callout. The overall timeout applies when there is more
3893 than one host that can be tried. The timeout is checked before trying the
3894 next host. This prevents very long delays if there are a large number of
3895 hosts and all are timing out (e.g. when the network connections are timing
3896 out). The value of the overall timeout can be changed by specifying an
3897 additional sub-option for "callout", called "maxwait". For example:
3899 verify = sender/callout=5s,maxwait=20s
3901 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3902 to the end before writing, but this should make it even safer).
3904 22. Exim was forgetting that it had advertised PIPELINING for the second and
3905 subsequent messages on an SMTP connection. It was also not resetting its
3906 memory on STARTTLS and an internal HELO.
3908 23. When Exim logs an SMTP synchronization error within a session, it now
3909 records whether PIPELINING has been advertised or not.
3911 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3912 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3913 rather than long int.
3915 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3917 26. Added the never_mail option to autoreply.
3923 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3924 crash if the getsockname() call failed; this can happen if a connection is
3925 closed very soon after it is established. The problem was simply in the
3926 order in which certain operations were done, causing Exim to try to write
3927 to the SMTP stream before it had set up the file descriptor. The bug has
3928 been fixed by making things happen in the correct order.
3934 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3935 before the connection was closed, thus losing the rejection response.
3937 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3938 some early Solaris releases, but causes trouble in current releases where
3939 socklen_t is defined.
3941 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3944 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3947 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3948 that releasing the top of it at the end releases what was used for sub-
3949 expansions (unless the block got too big). However, discard this block if
3950 the first thing is a variable or header, so that we can use its block when
3951 it is dynamic (useful for very large $message_headers, for example).
3953 6. Lookups now cache *every* query, not just the most recent. A new, separate
3954 store pool is used for this. It can be recovered when all lookup caches are
3955 flushed. Lookups now release memory at the end of their result strings.
3956 This has involved some general refactoring of the lookup sources.
3958 7. Some code has been added to the store_xxx() functions to reduce the amount
3959 of flapping under certain conditions.
3961 8. log_incoming_interface used to affect only the <= reception log lines. Now
3962 it causes the local interface and port to be added to several more SMTP log
3963 lines, for example "SMTP connection from", and rejection lines.
3965 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3967 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3969 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3970 could be overwritten at the end of the current message (or the start of a
3971 new message if it was set in a HELO ACL). The value is now preserved for
3972 the duration of the SMTP connection.
3974 12. If a transport had a headers_rewrite setting, and a matching header line
3975 contained an unqualified address, that address was qualified, even if it
3976 did not match any rewriting rules. The underlying bug was that the values
3977 of the flags that permit the existence of unqualified sender and recipient
3978 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3979 non-local messages, and by -bnq for local messages) were not being
3980 preserved with the message after it was received.
3982 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3983 "next input=" as part of the text comprising the host identity instead of
3984 the correct text. The code was using the same buffer for two different
3985 strings. However, depending on which order the printing function evaluated
3986 its arguments, the bug did not always show up. Under Linux, for example, my
3987 test suite worked just fine.
3989 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3990 doesn't work with some older versions of Perl. It has been changed to "my",
3991 which in any case is probably the better facility to use.
3993 15. A really picky compiler found some instances of statements for creating
3994 error messages that either had too many or two few arguments for the format
3997 16. The size of the buffer for calls to the DNS resolver has been increased
3998 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3999 for addresses that have a lot of PTR records. This alleviates a problem; it
4000 does not fully solve it.
4002 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4003 buffer now truncates the list and logs the incident, which is the same
4004 action as happens when Exim is looking up a host name and its aliases.
4005 Previously in this situation something unpredictable would happen;
4006 sometimes it was "internal error: store_reset failed".
4008 18. If a server dropped the connection unexpectedly when an Exim client was
4009 using GnuTLS and trying to read a response, the client delivery process
4010 crashed while trying to generate an error log message.
4012 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4013 string, for example:
4015 warn message = H1: something\nH2: something
4017 the text was added as a single header line from Exim's point of view
4018 though it ended up OK in the delivered message. However, searching for the
4019 second and subsequent header lines using $h_h2: did not work. This has been
4020 fixed. Similarly, if a system filter added multiple headers in this way,
4021 the routers could not see them.
4023 20. Expanded the error message when iplsearch is called with an invalid key to
4024 suggest using net-iplsearch in a host list.
4026 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4027 ACLs are no longer actually imposed (and a message to that effect is
4030 22. If a "gecos" field in a passwd entry contained escaped characters, in
4031 particular, if it contained a \" sequence, Exim got it wrong when building
4032 a From: or a Sender: header from that name. A second bug also caused
4033 incorrect handling when an unquoted " was present following a character
4034 that needed quoting.
4036 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4037 was not being matched caselessly.
4039 24. Arranged for all hyphens in the exim.8 source to be escaped with
4042 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4043 properly. Recipient callout cache records were still being keyed to include
4044 the sender, even when use_sender was set false. This led to far more
4045 callouts that were necessary. The sender is no longer included in the key
4046 when use_sender is false.
4048 26. Added "control = submission" modifier to ACLs.
4050 27. Added the ${base62d: operator to decode base 62 numbers.
4052 28. dnsdb lookups can now access SRV records.
4054 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4055 the configuration file.
4057 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4058 (-v) mode. This makes the output for a verbose queue run more intelligible.
4060 31. Added a use_postmaster feature to recipient callouts.
4062 32. Added the $body_zerocount variable, containing the number of binary zero
4063 bytes in the message body.
4065 33. The time of last modification of the "new" subdirectory is now used as the
4066 "mailbox time last read" when there is a quota error for a maildir
4069 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4071 35. Added +ignore_unknown as a special item in host lists.
4073 36. Code for decoding IPv6 addresses in host lists is now included, even if
4074 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4075 address was recognized as an IP address, but was then not correctly decoded
4076 into binary, causing unexpected and incorrect effects when compared with
4083 1. Very minor rewording of debugging text in manualroute to say "list of
4084 hosts" instead of "hostlist".
4086 2. If verify=header_syntax was set, and a header line with an unqualified
4087 address (no domain) and a large number of spaces between the end of the
4088 name and the colon was received, the reception process suffered a buffer
4089 overflow, and (when I tested it) crashed. This was caused by some obsolete
4090 code that should have been removed. The fix is to remove it!
4092 3. When running in the test harness, delay a bit after writing a bounce
4093 message to get a bit more predictability in the log output.
4095 4. Added a call to search_tidyup() just before forking a reception process. In
4096 theory, someone could use a lookup in the expansion of smtp_accept_max_
4097 per_host which, without the tidyup, could leave open a database connection.
4099 5. Added the variables $recipient_data and $sender_data which get set from a
4100 lookup success in an ACL "recipients" or "senders" condition, or a router
4101 "senders" option, similar to $domain_data and $local_part_data.
4103 6. Moved the writing of debug_print from before to after the "senders" test
4106 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4107 problems for message scanning, either using a data ACL, or using
4108 local_scan() because the Received: header was not generated till after they
4109 were called (in order to set the time as the time of reception completion).
4110 I have revised the way this works. The header is now generated after the
4111 body is received, but before the ACL or local_scan() are called. After they
4112 are run, the timestamp in the header is updated.
4118 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4119 before starting a queue runner without re-exec. This happened only when
4120 deliver_drop_privilege was set or when the Exim user was set to root. The
4121 effect of the bug was that timeouts during subsequent deliveries caused
4122 crashes instead of being properly handled. The handler is now left at its
4123 default (and expected) setting.
4125 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4126 message, again when deliver_drop_privilege is set or Exim is run as root.
4127 The bug described in (1) was not present in this case, but the tidying up
4128 of the other signals was missing. I have made the two cases consistent.
4130 3. The ignore_target_hosts setting on a manualroute router was being ignored
4131 for hosts that were looked up using the /MX notation.
4133 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4136 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4137 operated on the sender address. After changing the $sender_address to <>
4138 for the sender address verify, Exim was re-instated it as the original
4139 (before rewriting) address, but remembering that it had rewritten it, so it
4140 wasn't rewriting it again. This bug also had the effect of breaking the
4141 sender address verification caching when the sender address was rewritten.
4143 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4144 This has been changed so that if the ip literal address matches
4145 ignore_target_hosts, the router declines.
4147 7. Added expansion conditions match_domain, match_address, and match_local_
4148 part (NOT match_host).
4150 8. The placeholder for the Received: header didn't have a length field set.
4152 9. Added code to Exim itself and to exim_lock to test for a specific race
4153 condition that could lead to file corruption when using MBX delivery. The
4154 issue is with the lockfile that is created in /tmp. If this file is removed
4155 after a process has opened it but before that process has acquired a lock,
4156 there is the potential for a second process to recreate the file and also
4157 acquire a lock. This could lead to two Exim processes writing to the file
4158 at the same time. The added code performs the same test as UW imapd; it
4159 checks after acquiring the lock that its file descriptor still refers to
4160 the same named file.
4162 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4163 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4166 11. Added the smtp_active_hostname option. If used, this will typically be made
4167 to depend on the incoming interface address. Because $interface_address is
4168 not set up until the daemon has forked a reception process, error responses
4169 that can happen earlier (such as "too many connections") no longer contain
4172 12. If an expansion in a condition on a "warn" statement fails because a lookup
4173 defers, the "warn" statement is abandoned, and the next ACL statement is
4174 processed. Previously this caused the whole ACL to be aborted.
4176 13. Added the iplsearch lookup type.
4178 14. Added ident_timeout as a log selector.
4180 15. Added tls_certificate_verified as a log selector.
4182 16. Added a global option tls_require_ciphers (compare the smtp transport
4183 option of the same name). This controls incoming TLS connections.
4185 17. I finally figured out how to make tls_require_ciphers do a similar thing
4186 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4187 before starting the TLS session.
4189 18. Tabs are now shown as \t in -bP output.
4191 19. If the log selector return_path_on_delivery was set, Exim crashed when
4192 bouncing a message because it had too many Received: header lines.
4194 20. If two routers both had headers_remove settings, and the first one included
4195 a superfluous trailing colon, the final name in the first list and the
4196 first name in the second list were incorrectly joined into one item (with a
4197 colon in the middle).
4203 1. Added -C and -D options to the exinext utility, mainly to make it easier
4204 to include in the automated testing, but these could be helpful when
4205 multiple configurations are in use.
4207 2. The exinext utility was not formatting the output nicely when there was
4208 an alternate port involved in the retry record key, nor when there was a
4209 message id as well (for retries that were specific to a specific message
4210 and a specific host). It was also confused by IPv6 addresses, because of
4211 the additional colons they contain. I have fixed the IPv4 problem, and
4212 patched it up to do a reasonable job for IPv6.
4214 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4215 delivery, the log line now contains "pipelined" if PIPELINING was used.
4217 4. An SMTP transport process used to panic and die if the bind() call to set
4218 an explicit outgoing interface failed. This has been changed; it is now
4219 treated in the same way as a connect() failure.
4221 5. A reference to $sender_host_name in the part of a conditional expansion
4222 that was being skipped was still causing a DNS lookup. This no longer
4225 6. The def: expansion condition was not recognizing references to header lines
4226 that used bh_ and bheader_.
4228 7. Added the _cache feature to named lists.
4230 8. The code for checking quota_filecount in the appendfile transport was
4231 allowing one more file than it should have been.
4233 9. For compatibility with Sendmail, the command line option
4241 and sets the incoming protocol and host name (for trusted callers). The
4242 host name and its colon can be omitted when only the protocol is to be set.
4243 Note the Exim already has two private options, -pd and -ps, that refer to
4244 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4245 "s", but I don't think that's a major issue.
4247 10. A number of refactoring changes to the code, none of which should affect
4250 (a) The number of logging options was getting close to filling up the
4251 32-bit word that was used as a bit map. I have split them into two classes:
4252 those that are passed in the argument to log_write(), and those that are
4253 only ever tested independently outside of that function. These are now in
4254 separate 32-bit words, so there is plenty of room for expansion again.
4255 There is no change in the user interface or the logging behaviour.
4257 (b) When building, for example, log lines, the code previously used a
4258 macro that called string_cat() twice, in order to add two strings. This is
4259 not really sufficiently general. Furthermore, there was one instance where
4260 it was actually wrong because one of the argument was used twice, and in
4261 one call a function was used. (As it happened, calling the function twice
4262 did not affect the overall behaviour.) The macro has been replaced by a
4263 function that can join an arbitrary number of extra strings onto a growing
4266 (c) The code for expansion conditions now uses a table and a binary chop
4267 instead of a serial search (which was left over from when there were very
4268 few conditions). Also, it now recognizes conditions like "pam" even when
4269 the relevant support is not compiled in: a suitably worded error message is
4270 given if an attempt is made to use such a condition.
4272 11. Added ${time_interval:xxxxx}.
4274 12. A bug was causing one of the ddress fields not to be passed back correctly
4275 from remote delivery subprocesses. The field in question was not being
4276 subsequently used, so this caused to problems in practice.
4278 13. Added new log selectors queue_time and deliver_time.
4280 14. Might have fixed a bug in maildirsizefile handling that threw up
4281 "unexpected character" debug warnings, and recalculated the data
4282 unnecessarily. In any case, I expanded the warning message to give more
4285 15. Added the message "Restricted characters in address" to the statements in
4286 the default ACL that block characters like @ and % in local parts.
4288 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4289 Three changes have been made:
4291 (a) There was a serious bug; a negative response to MAIL caused the whole
4292 recipient domain to be cached as invalid, thereby blocking all messages
4293 to all local parts at the same domain, from all senders. This bug has
4294 been fixed. The domain is no longer cached after a negative response to
4295 MAIL if the sender used is not empty.
4297 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4300 (c) A new callout option, "use_sender" has been added for people who want
4301 the modified behaviour.
4307 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4310 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4311 indeed breaks things for older releases.
4313 3. Added additional logging to the case where there is a problem reading data
4314 from a filter that is running in a subprocess using a pipe, in order to
4315 try to track down a specific problem.
4317 4. Testing facility fudge: when running in the test harness and attempting
4318 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4319 getting "No route to host". Convert this to a timeout.
4321 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4324 6. Some OS don't have socklen_t but use size_t instead. This affects the
4325 fifth argument of getsockopt() amongst other things. This is now
4326 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4327 can be set for individual OS. I have set it for SunOS5, OSF1, and
4328 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4329 some earlier ones do not.
4331 7. Change 4.30/15 was not doing the test caselessly.
4333 8. The standard form for an IPv6 address literal was being rejected by address
4334 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4335 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4336 this, as well as the form without the "IPv6" on the front (but only when
4337 address literals are enabled, of course).
4339 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4341 10. Exim crashed if a message with an empty sender address specified by -f
4342 encountered a router with an errors_to setting. This could be provoked only
4343 by a command such as
4347 where an empty string was supplied; "<>" did not hit this bug.
4349 11. Installed PCRE release 4.5.
4351 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4352 remained set. It is now erased.
4354 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4355 times from message ids (which are base 36 rather than the normal 62).
4357 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4358 were being counted as actual protocol errors, and logged if the log
4359 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4360 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4361 and DATA following a set of rejected RCPTs do not count as protocol errors.
4362 In other words, Exim assumes they were pipelined, though this may not
4363 actually be the case. Of course, in all cases the client gets an
4364 appropriate error code.
4366 15. If a lookup fails in an ACL condition, a message about the failure may
4367 be available; it is used if testing the ACL cannot continue, because most
4368 such messages specify what the cause of the deferral is. However, some
4369 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4370 that caused an old message to be retained and used if a later statement
4371 caused a defer, replacing the real cause of the deferral.
4373 16. If an IP address had so many PTR records that the DNS lookup buffer
4374 was not large enough to hold them, Exim could crash while trying to process
4375 the truncated data. It now detects and logs this case.
4377 17. Further to 4.21/58, another change has been made: if (and only if) the
4378 first line of a message (the first header line) ends with CRLF, a bare LF
4379 in a subsequent header line has a space inserted after it, so as not to
4380 terminate the header.
4382 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4383 unnecessarily, used atoi() instead of strtol(), and didn't check the
4384 termination when getting file sizes from file names by regex.
4386 19. Completely re-implemented the support for maildirsize files, in the light
4387 of a number of problems with the previous contributed implementation
4388 (4.30/29). In particular:
4390 . If the quota is zero, the maildirsize file is maintained, but no quota is
4393 . If the maildir directory does not exist, it is created before any attempt
4394 to write a maildirsize file.
4396 . The quota value in the file is just a cache; if the quota is changed in
4397 the transport, the new value overrides.
4399 . A regular expression is available for excluding directories from the
4402 20. The autoreply transport checks the characters in options that define the
4403 message's headers; it allows continued headers, but it was checking with
4404 isspace() after an embedded newline instead of explicitly looking for a
4407 21. If all the "regular" hosts to which an address was routed had passed their
4408 expiry times, and had not reached their retry times, the address was
4409 bounced, even if fallback hosts were defined. Now Exim should go on to try
4412 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4413 equivalent code in the SMTP transport. Some hosts send humungous responses
4414 to HELO/EHLO, more than 1024 it seems.
4416 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4417 gives compiler warnings in some environments. I've now done it "properly",
4420 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4421 (because of problems with the built-in one) was declared to return uschar *
4422 instead of char *, causing compiler failure.
4424 25. Fixed a file descriptor leak when processing alias/forward files.
4426 26. Fixed a minor format string issue in dbfn.c.
4428 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4430 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4431 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4434 29. When a sender address is verified, it is cached, to save repeating the test
4435 when there is more than one recipient in a message. However, when the
4436 verification involves a callout, it is possible for different callout
4437 options to be set for different recipients. It is too complicated to keep
4438 track of this in the cache, so now Exim always runs a verification when a
4439 callout is required, relying on the callout cache for the optimization.
4440 The overhead is duplication of the address routing, but this should not be
4443 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4444 to be verified with callout=postmaster, and the main callout worked but the
4445 postmaster check failed, the verification correctly failed. However, if a
4446 subsequent RCPT command asked for sender verification *without* the
4447 postmaster check, incorrect caching caused this verification also to fail,
4450 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4451 it was not caching the DNS options (qualify_single, search_parents) that
4452 were used when the lookup failed. A subsequent lookup with different
4453 options therefore always gave the same answer, though there were cases
4454 where it should not have. (Example: a "domains = !$mx_any" option on a
4455 dnslookup router: the "domains" option is always processed without any
4456 widening, but the router might have qualify_single set.) Now Exim uses the
4457 cached value only when the same options are set.
4459 32. Added John Jetmore's "exipick" utility to the distribution.
4461 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4462 than a timeout (e.g. a certificate is required, and is not provided), an
4463 Exim server now closes the connection immediately. Previously it waited for
4464 the client to close - but if the client is SSL, it seems that they each
4465 wait for each other, leading to a delay before one of them times out.
4467 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4468 maintained 0.8.x compatibility because I don't think many are using it, and
4469 it is clearly obsolete.
4471 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4474 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4475 client certificate was expired. A simple patch fixes this, though I don't
4476 understand the full logic of why the verify callback is called multiple
4479 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4480 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4481 which causes problems with some clients (such as the Certicom SSL Plus
4482 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4483 disables the coutermeasure allowing Eudora to connect."
4485 38. Exim was not checking that a write() to a log file succeeded. This could
4486 lead to Bad Things if a log got too big, in particular if it hit a file
4487 size limit. Exim now panics and dies if it cannot write to a log file, just
4488 as it does if it cannot open a log file.
4490 39. Modified OS/Makefile-Linux so that it now contains
4492 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4494 The two -D definitions ensure that Exim is compiled with large file
4495 support, which makes it possible to handle log files that are bigger than
4498 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4499 instance) a domain was checked against a named list that involved a lookup,
4500 causing $domain_data to be set, then another domain was checked against the
4501 same list, then the first domain was re-checked, the value of $domain_data
4502 after the final check could be wrong. In particular, if the second check
4503 failed, it could be set empty. This bug probably also applied to
4506 41. The strip_trailing_dot option was not being applied to the address given
4507 with the -f command-line option.
4509 42. The code for reading a message's header from the spool was incrementing
4510 $received_count, but never initializing it. This meant that the value was
4511 incorrect (doubled) while delivering a message in the same process in which
4512 it was received. In the most common configuration of Exim, this never
4513 happens - a fresh exec is done - but it can happen when
4514 deliver_drop_privilege is set.
4516 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4517 it now includes up to 150 characters of the unexpected data in the log
4520 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4521 and building data strings. The size of both of these buffers was 10 000
4522 bytes - far larger than anybody would *ever* want, thought I. Needless to
4523 say, somebody hit the limit. I have increased the maximum line length to
4524 20 000 and the maximum data length of concatenated lines to 100 000. I have
4525 also fixed two bugs, because there was no checking on these buffers. Tsk,
4526 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4527 buffer is too small.
4529 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4530 lsearch lookups. Now it does.
4532 46. When parsing a route_list item in a manualroute router, a fixed-length
4533 buffer was used for the list of hosts. I made this 1024 bytes long,
4534 thinking that nobody would ever have a list of hosts that long. Wrong.
4535 Somebody had a whole pile of complicated expansion conditions, and the
4536 string was silently truncated, leading to an expansion error. It turns out
4537 that it is easier to change to an unlimited length (owing to other changes
4538 that have happened since this code was originally written) than to build
4539 structure for giving a limitation error. The length of the item that
4540 expands into the list of hosts is now unlimited.
4542 47. The lsearch lookup could not handle data where the length of text line was
4543 more than 4095 characters. Such lines were truncated, leading to shortened
4544 data being returned. It should now handle lines of any length.
4546 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4547 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4550 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4551 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4552 no longer happens in comment lines. A list of replacements is now placed
4553 at the head of all of the source files, except those whose only change is
4554 to replace PERL_COMMAND in the very first #! line.
4556 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4557 messages on the queue, with a bottom-up merge sort, using code contributed
4558 by Michael Haardt. This should make operations like -bp somewhat faster on
4559 large queues. It won't affect queue runners, except when queue_run_in_order
4562 51. Installed eximstats 1.31 in the distribution.
4564 52. Added support for SRV lookups to the dnslookup router.
4566 53. If an ACL referred to $message_body or $message_body_end, the value was not
4567 reset for any messages that followed in the same SMTP session.
4569 54. The store-handling optimization for building very long strings was not
4570 differentiating between the different store pools. I don't think this
4571 actually made any difference in practice, but I've tidied it.
4573 55. While running the routers to verify a sender address, $sender_address
4574 was still set to the sender address. This is wrong, because when routing to
4575 send a bounce to the sender, it would be empty. Therefore, I have changed
4576 it so that, while verifying a sender address, $sender_address is set to <>.
4577 (There is no change to what happens when verifying a recipient address.)
4579 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4580 target A or AAAA records (if not already returned) without resetting the
4581 qualify_single or search_parents options of the DNS resolver. These are
4582 inappropriate in this case because the targets of MX and SRV records must
4583 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4584 target that, when qualified, matched something in the local domain. These
4585 two options are now turned off when doing these lookups.
4587 57. It seems that at least some releases of Reiserfs (which does not have the
4588 concept of a fixed number of inodes) returns zero and not -1 for the
4589 number of available inodes. This interacted badly with check_spool_inodes,
4590 which assumed that -1 was the "no such thing" setting. What I have done is
4591 to check that the total number of inodes is greater than zero before doing
4592 the test of how many are available.
4594 58. When a "warn" ACL statement has a log_message modifier, the message is
4595 remembered, and not repeated. This is to avoid a lot of repetition when a
4596 message has many recipients that cause the same warning to be written.
4597 Howewer, Exim was preserving the list of already written lines for an
4598 entire SMTP session, which doesn't seem right. The memory is now reset if a
4599 new message is started.
4601 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4602 debugging output unless log_rewrite was also set.
4604 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4605 of (int)(handle) when we know that handle contains (void *)(-1).
4607 61. The Exim daemon panic-logs an error return when it closes the incoming
4608 connection. However "connection reset by peer" seems to be common, and
4609 isn't really an error worthy of noting specially, so that particular error
4612 62. When Exim is trying to find all the local interfaces, it used to panic and
4613 die if the ioctl to get the interface flags failed. However, it seems that
4614 on at least one OS (Solaris 9) it is possible to have an interface that is
4615 included in the list of interfaces, but for which you get a failure error
4616 for this call. This happens when the interface is not "plumbed" into a
4617 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4618 failure of the "get flags" call assumes that the interface is down.
4620 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4621 makes life easier for people who are doing arithmetic on fields extracted
4622 from dates, where you often get leading zeros that should not be
4623 interpreted as octal.
4625 64. Added qualify_domain to the redirect router, to override the global
4628 65. If a pathologically long header line contained very many addresses (the
4629 report of this problem mentioned 10 000) and each of them was rewritten,
4630 Exim could use up a very large amount of memory. (It kept on making new
4631 copies of the header line as it rewrote, and never released the old ones.)
4632 At the expense of a bit more processing, the header rewriting function has
4633 been changed so that it no longer eats memory in this way.
4635 66. The generation of the Received: header has been moved from the time that a
4636 message starts to be received, to the time that it finishes. The timestamp
4637 in the Received: header should now be very close to that of the <= log
4638 line. There are two side-effects of this change:
4640 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4641 logged header lines no longer include the local Received: line, because
4642 it has not yet been created. The same applies to a copy of the message
4643 that is returned to a non-SMTP sender when a message is rejected.
4645 (b) When a filter file is tested using -bf, no additional Received: header
4646 is added to the test message. After some thought, I decided that this
4649 This change does not affect the value of $received_for. It is still set
4650 after address rewriting, but before local_scan() is called.
4652 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4654 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4655 gave an unhelpful panic error message, and a defer error. I have managed to
4656 change this behaviour so that it now rejects any supplied certificate,
4657 which seems right, as the list of acceptable certificates is empty.
4659 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4660 gave an unhelpful defer error. I have not managed to make this reject any
4661 supplied certificates, but the error message it gives is "no certificate
4662 supplied", which is not helpful.
4664 70. exigrep's output now also includes lines that are not associated with any
4665 message, but which match the given pattern. Implemented by a patch from
4666 Martin Sluka, which also tidied up the Perl a bit.
4668 71. Recipient callout verification, like sender verification, was using <> in
4669 the MAIL FROM command. This isn't really the right thing, since the actual
4670 sender may affect whether the remote host accepts the recipient or not. I
4671 have changed it to use the actual sender in the callout; this means that
4672 the cache record is now keyed on a recipient/sender pair, not just the
4673 recipient address. There doesn't seem to be a real danger of callout loops,
4674 since a callout by the remote host to check the sender would use <>.
4675 [SEE ABOVE: changed after hitting problems.]
4677 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4678 temporary errors. However, in the case of such a code being given after
4679 the end of a data transmission (i.e. after ".") Exim was failing to write
4680 a retry record for the message. (Yes, there was some broken host that was
4681 actually sending 8xx at this point.)
4683 73. An unknown lookup type in a host list could cause Exim to panic-die when
4684 the list was checked. (An example that provoked this was putting <; in the
4685 middle of a list instead of at the start.) If this happened during a DATA
4686 ACL check, a -D file could be left lying around. This kind of configuration
4687 error no longer causes Exim to die; instead it causes a defer errror. The
4688 incident is still logged to the main and panic logs.
4690 74. Buglet left over from Exim 3 conversion. The message "too many messages
4691 in one connection" was written to the rejectlog but not the mainlog, except
4692 when address rewriting (yes!) was being logged.
4694 75. Added write_rejectlog option.
4696 76. When a system filter was run not as root (that is, when system_filter_user
4697 was set), the values of the $n variables were not being returned to the
4698 main process; thus, they were not subsequently available in the $sn
4701 77. Added +return_path_on_delivery log selector.
4703 78. A connection timeout was being treated differently from recipients deferred
4704 when testing hosts_max_try with a message that was older than the host's
4705 retry timeout. (The host should not be counted, thus allowing all hosts to
4706 be tried at least once before bouncing.) This may have been the cause of an
4707 occasionally reported bug whereby a message would remain on the queue
4708 longer than the retry timeout, but would be bounced if a delivery was
4709 forced. I say "may" because I never totally pinned down the problem;
4710 setting up timeout/retry tests is difficult. See also the next item.
4712 79. The ultimate address timeout was not being applied to errors that involved
4713 a combination of host plus message (for example, a timeout on a MAIL
4714 command). When an address resolved to a number of possible hosts, and they
4715 were not all tried for each delivery (e.g. because of hosts_max_try), a
4716 message could remain on the queue longer than the retry timeout.
4718 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4721 81. Fixed an obscure SMTP outgoing bug which required at least the following
4722 conditions: (a) there was another message waiting for the same server;
4723 (b) the server returned 5xx to all RCPT commands in the first message so
4724 that the message was not completed; (c) the server dropped the connection
4725 or gave a negative response to the RSET that Exim sends to abort the
4726 transaction. The observed case was a dropped connection after DATA that had
4727 been sent in pipelining mode. That is, the server had advertised PIPELINING
4728 but was not implementing it correctly. The effect of the bug was incorrect
4729 behaviour, such as trying another host, and this could lead to a crash.
4735 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4736 and daemon.c were passed as pointers to ints; they should have been
4737 pointers to socklen_t variables (which are typically unsigned ints).
4739 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4742 3. Fixed a really odd bug that affected only the testing scheme; patching a
4743 certain fixed string in the binary changed the value of another string that
4744 happened to be identical to the end of the original first string.
4746 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4747 name", it returns that address as the IP address. On some operating
4748 systems (e.g. Solaris), it also passes back the IP address string as the
4749 "host name". However, on others (e.g. Linux), it passes back an empty
4750 string. Exim wasn't checking for this, and was changing the host name to an
4751 empty string, assuming it had been canonicized.
4753 5. Although rare, it is permitted to have more than one PTR record for a given
4754 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4755 all the names associated with an address, because they do in Solaris.
4756 However, it seems that they do not in Linux for data that comes from the
4757 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4758 I found this out when I moved to a new Linux workstation and tried to run
4759 the Exim test suite.
4761 To get round this problem I have changed the code so that it now does its
4762 own call to the DNS to look up PTR records when searching for a host name.
4763 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4764 addresses that are only in /etc/hosts are still found.
4766 This behaviour is, however, controlled by an option called host_lookup_
4767 order, which defaults to "bydns:byaddr". If people want to use the other
4768 order, or indeed, just use one or the other means of lookup, they can
4769 specify it in this variable.
4771 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4772 some operating systems, this comes back from gethostbyaddr() as an empty
4773 string, and this is what Exim used to test for. However, it seems that in
4774 other systems, "." is yielded. Exim now tests for this case too.
4776 7. The values of check_spool_space and check_log_space are now held internally
4777 as a number of kilobytes instead of an absolute number of bytes. If a
4778 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4779 kilobyte. This means that much larger values can be stored.
4781 8. Exim monitor: an attempt to get the action menu when not actually pointing
4782 at a message produces an empty menu entitled "No message selected". This
4783 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4784 no entries in it ("Shell widget menu has zero width and/or height"). So I
4785 have added a single, blank menu entry in this case.
4787 9. Added ${quote_local_part.
4789 10. MIME decoding is now applied to the contents of Subject: header lines when
4792 11. Now that a reference to $sender_host_address automatically causes a reverse
4793 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4794 host lookup before query-style lookups in lists that might use this
4795 variable. This has therefore been abolished, and the "net-" prefix is no
4796 longer necessary for query-style lookups.
4798 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4799 have been a typo for LFLAGS, so it has been changed.
4801 13. The install script calls Exim with "-C /dev/null" in order to find the
4802 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4803 to be output. Howeve, since Exim outputs its version number before the
4804 error, it didn't break the script. It just looked ugly. I fixed this by
4805 always allowing "-C /dev/null" if the caller is root.
4807 14. Ignore overlarge ACL variable number when reading spool file - insurance
4808 against a later release with more variables having written the file.
4810 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4811 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4812 this, as well as the form without the "IPv6" on the front.
4814 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4815 OS/Makefile-Darwin file.
4817 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4818 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4820 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4821 variable. However, some OS use ints for inodes. I've added cast to long int
4822 to get rid of the compiler warning.
4824 19. I had forgotten to lock out "/../" in configuration file names when
4825 ALT_CONFIG_PREFIX was set.
4827 20. Routers used for verification do not need to specify transports. However,
4828 if such a router generated a host list, and callout was configured, Exim
4829 crashed, because it could not find a port number from the (non-existent)
4830 transport. It now assumes port 25 in this circumstance.
4832 21. Added the -t option to exigrep.
4834 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4835 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4836 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4837 them is set, LOOKUP_LSEARCH is forced.
4839 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4841 24. Added sender and host information to the "rejected by local_scan()" log
4842 line; previously there was no indication of these.
4844 25. Added .include_if_exists.
4846 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4847 Linux. It turns out that not all file systems support this. Apparently some
4848 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4849 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4850 not supported on the file descriptor, is now ignored when Exim is trying to
4851 sync a directory. This applies only to Linux.
4853 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4855 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4856 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4859 29. Added support for maildirsize files from supplied patch (modified a bit).
4861 30. The use of :fail: followed by an empty string could lead Exim to respond to
4862 sender verification failures with (e.g.):
4864 550 Verification failed for <xxx>
4865 550 Sender verify failed
4867 where the first response line was missing the '-' that indicates it is not
4868 the final line of the response.
4870 31. The loop for finding the name of the user that called Exim had a hardwired
4871 limit of 10; it now uses the value of finduser_retries, which is used for
4872 all other user lookups.
4874 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4877 33. Exim was neglecting to zero errno before one call of strtol() when
4878 expanding a string and expecting an integer value. On some systems this
4879 resulted in spurious "integer overflow" errors. Also, it was casting the
4880 result into an int without checking.
4882 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4883 did not work. The code looks as if it has *never* worked, though it appears
4884 to have been documented since at least releast 1.62. I have made it work.
4886 35. The "timeout_DNS" error in retry rules, also documented since at least
4887 1.62, also never worked. As it isn't clear exactly what this means, and
4888 clearly it isn't a major issue, I have abolished the feature by treating it
4889 as "timeout", and writing a warning to the main and panic logs.
4891 36. The display of retry rules for -brt wasn't always showing the error code
4894 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4895 timeout_connect_A, timeout_connect_MX.
4897 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4898 to the empty sender.
4900 39. The daemon was not analysing the content of -oX till after it had closed
4901 stderr and disconnected from the controlling terminal. This meant that any
4902 syntax errors were only noted on the panic log, and the return code from
4903 the command was 0. By re-arranging the code a little, I've made the
4904 decoding happen first, so such errors now appear on stderr, and the return
4905 code is 1. However, the actual setting up of the sockets still happens in
4906 the disconnected process, so errors there are still only recorded on the
4909 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4910 connections (as happens on some IP stacks) was logged at start up time as
4911 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4912 it from "IPv6 and IPv4", which means that two separate sockets are being
4915 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4916 says whether AF_INET or AF_INET6 was passed as an argument.
4918 42. Exiwhat output was messed up when time zones were included in log
4921 43. Exiwhat now gives more information about the daemon's listening ports,
4922 and whether -tls-on-connect was used.
4924 44. The "port" option of the smtp transport is now expanded.
4926 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4927 silently ignored. Now an error message is written to the main and panic
4930 46. There's a new ACL modifier called "logwrite" which writes to a log file
4931 as soon as it is encountered.
4933 47. Added $local_user_uid and $local_user_gid at routing time.
4935 48. Exim crashed when trying to verify a sender address that was being
4938 49. Exim was recognizing only a space character after ".include". It now also
4939 recognizes a tab character.
4941 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4942 extracting the relevant information from the specification. The man page no
4943 longer contains scrambled data for the -d option, and I've added a section
4944 at the front about calling Exim under different names.
4946 51. Added "extra_headers" argument to the "mail" command in filter files.
4948 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4951 53. Installed eximstats 1.29.
4953 54. Added transport_filter_timeout as a generic transport option.
4955 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4956 Cc: header lines. This was required by RFC 822, but it not required by RFC
4959 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4960 incoming messages that did not have them. Now it does so only if the
4961 message originates locally, that is, if there is no associated remote host
4962 address. When Resent- header lines are present, this applies to the Resent-
4963 lines rather than the non-Resent- lines.
4965 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4966 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4968 58. Messages for configuration errors now include the name of the main
4969 configuration file - useful now that there may be more than one file in a
4970 list (.included file names were always shown).
4972 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4973 for those rare installations that do not start the daemon as root or run it
4974 setuid root. I've cut out the call to initgroups() if the daemon is not
4977 60. The Exim user and group can now be bound into the binary as text strings
4978 that are looked up at the start of Exim's processing.
4980 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4982 62. Added $mailstore_basename variable.
4984 63. Installed patch to sieve.c from Michael Haardt.
4986 64. When Exim failed to open the panic log after failing to open the main log,
4987 the original message it was trying to log was written to stderr and debug
4988 output, but if they were not available (the usual case in production), it
4989 was lost. Now it is written to syslog before the two lines that record the
4990 failures to open the logs.
4992 65. Users' Exim filters run in subprocesses under the user's uid. It is
4993 possible for a "deliver" command or an alias in a "personal" command to
4994 provoke an address rewrite. If logging of address rewriting is configured,
4995 this fails because the process is not running as root or exim. There may be
4996 a better way of dealing with this, but for the moment (because 4.30 needs
4997 to be released), I have disabled address rewrite logging when running a
4998 filter in a non-root, non-exim process.
5004 1. The buildconfig auxiliary program wasn't quoting the value set for
5005 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5006 not defined. This bug was masked in 4.22 by the effect that was fixed in
5009 2. Some messages that were rejected after a message id was allocated were
5010 shown as "incomplete" by exigrep. It no longer does this for messages that
5011 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5013 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5014 have allow_domain_literals set, the ID did not get logged in the <= line.
5015 Domain literals are now always recognized in Message-ID: header lines.
5017 4. The first argument for a ${extract expansion item is the key name or field
5018 number. Leading and trailing spaces in this item were not being ignored,
5019 causing some misleading effects.
5021 5. When deliver_drop_privilege was set, single queue runner processes started
5022 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5023 same command in the process it spins off) were not dropping privilege.
5025 6. When the daemon running as "exim" started a queue runner, it always
5026 re-executed Exim in the spun-off process. This is a waste of effort when
5027 deliver_drop_privilege is set. The new process now just calls the
5028 queue-runner function directly.
5034 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5037 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5038 ignored. Though the use of -oP was forcing the writing of a pid file, it
5039 was always written to the default place.
5041 3. If the message "no IP address found for host xxxx" is generated during
5042 incoming verification, it is now followed by identification of the incoming
5043 connection (so you can more easily find what provoked it).
5045 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5047 5. Added some features to "harden" Exim a bit more against certain attacks:
5049 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5050 be put in Local/Makefile. This is like the never_users runtime option,
5051 but it cannot be overridden. The default setting is "root".
5053 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5054 prefix string with which any file named in a -C command line option
5057 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5058 is retained for -C and -D only if the caller of Exim is root. Without
5059 it, the exim user may also use -C and -D and retain privilege.
5061 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5062 command line option is disabled.
5064 6. Macro names set by the -D option must start with an upper case letter, just
5065 like macro names defined in the configuration file.
5067 7. Added "dereference=" facility to LDAP.
5069 8. Two instances of the typo "uknown" in the source files are fixed.
5071 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5072 the Configure-Makefile script screwed up while processing it.
5074 10. Incorporated PCRE 4.4.
5076 11. The SMTP synchronization check was not operating right at the start of an
5077 SMTP session. For example, it could not catch a HELO sent before the client
5078 waited for the greeting. There is now a check for outstanding input at the
5079 point when the greeting is written. Because of the duplex, asynchronous
5080 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5081 way, but not yet received, when the check is performed.
5083 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5084 on TCP/IP sockets, because this apparently causes some broken clients to
5087 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5088 unchanged) from the Cygwin maintainer.
5090 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5091 was supported instead of testing for mbx. Effectively a typo.
5093 15. The spa authenticator server code was not checking that the input it
5094 received was valid base64.
5096 16. The debug output line for the "set" modifier in ACLs was not showing the
5097 name of the variable that was being set.
5099 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5101 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5102 lookup on its own. Something else was needed to trigger the lookup. For
5103 example, a match in host_lookup or the need for a host name in a host list.
5104 Now, if $sender_host_name is referenced and the host name has not yet been
5105 looked up, a lookup is performed. If the lookup fails, the variable remains
5106 empty, and $host_lookup_failed is set to "1".
5108 19. Added "eqi" as a case-independent comparison operator.
5110 20. The saslauthd authentication condition could segfault if neither service
5111 nor realm was specified.
5113 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5114 error message that was logged was misleading, and incoming SMTP
5115 connections were dropped. The message is now more accurate, and temporary
5116 errors are given to SMTP connections.
5118 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5119 (except RSET) with a 421 error, until QUIT is received. However, it was
5120 failing to send a response to QUIT.
5122 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5123 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5124 ACL is now run after the helo_try_verify_hosts code.
5126 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5127 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5128 case-independent, so other case variants are also recognized). Apparently
5129 some systems use these upper case variants.
5131 25. If more than two messages were waiting for the same host, and a transport
5132 filter was specified for the transport, Exim sent two messages over the
5133 same TCP/IP connection, and then failed with "socket operation on non-
5134 socket" when it tried to send the third.
5136 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5138 27. The extern definition of crypt16() in expand.c was not being excluded when
5139 the OS had its own crypt16() function.
5141 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5142 as a preferred synonym for return_size_limit, both as an option and as an
5145 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5147 30. Changed the default configuration ACL to relax the local part checking rule
5148 for addresses that are not in any local domains. For these addresses,
5149 slashes and pipe symbols are allowed within local parts, but the sequence
5150 /../ is explicitly forbidden.
5152 31. SPA server authentication was not clearing the challenge buffer before
5155 32. log_message in a "warn" ACL statement was writing to the reject log as
5156 well as to the main log, which contradicts the documentation and doesn't
5157 seem right (because no rejection is happening). So I have stopped it.
5159 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5160 However, I am unable to do any testing of this.
5162 34. Fixed an infelicity in the appendfile transport. When checking directories
5163 for a mailbox, to see if any needed to be created, it was accidentally
5164 using path names with one or more superfluous leading slashes; tracing
5165 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5167 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5168 contents are added to the log line that is written for every discarded
5169 recipient. (Previously a log_message setting was ignored.)
5171 36. The ${quote: operator now quotes the string if it is empty.
5173 37. The install script runs exim in order to find its version number. If for
5174 some reason other than non-existence or emptiness, which it checks, it
5175 could not run './exim', it was installing it with an empty version number,
5176 i.e. as "exim-". This error state is now caught, and the installation is
5179 38. An argument was missing from the function that creates an error message
5180 when Exim fails to connect to the socket for saslauthd authentication.
5181 This could cause Exim to crash, or give a corrupted message.
5183 39. Added isip, isip4, and isip6 to ${if conditions.
5185 40. The ACL variables $acl_xx are now saved with the message, and can be
5186 accessed later in routers, transports, and filters.
5188 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5189 strings in the file are not string-expanded.
5191 42. If a MAIL command specified a SIZE value that was too large to fit into an
5192 int variable, the check against message_size_limit failed. Such values are
5193 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5194 day this will have to be increased, but I don't think I want to be around
5195 when emails are that large.
5202 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5203 iconv() is not standard in FreeBSD.
5205 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5206 IPv6 enabled. The observed symptom was a segmentation fault on return from
5207 the function os_common_find_running_interfaces() in src/os.c.
5209 3. In the check_special_case() function in daemon.c I had used "errno" as an
5210 argument name, which causes warnings on some systems. This was basically a
5211 typo, since it was named "eno" in the comments!
5213 4. The code that waits for the clock to tick (at a resolution of some fraction
5214 of a second) so as to ensure message-id uniqueness was always waiting for
5215 at least one whole tick, when it could have waited for less. [This is
5216 almost certainly not relevant at current processor speeds, where it is
5217 unlikely to ever wait at all. But we try to future-proof.]
5219 5. The function that sleeps for a time interval that includes fractions of a
5220 second contained a race. It did not block SIGALRM between setting the
5221 timer, and suspending (a couple of lines later). If the interval was short
5222 and the sigsuspend() was delayed until after it had expired, the suspension
5223 never ended. On busy systems this could lead to processes getting stuck for
5226 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5227 process, before it forks any delivery processes. The open lookup caching
5228 mechanism meant that the open file or database connection was passed into
5229 the delivery process. The problem was that delivery processes always tidy
5230 up cached lookup data. This could cause a problem for the next delivery
5231 process started by the queue runner, because the external queue runner
5232 process does not know about the closure. So the next delivery process
5233 still has data in the lookup cache. In the case of a file lookup, there was
5234 no problem because closing a file descriptor in a subprocess doesn't affect
5235 the parent. However, if the lookup was caching a connection to a database,
5236 the connection was closed, and the second delivery process was likely to
5237 see errors such as "PGSQL: query failed: server closed the connection
5238 unexpectedly". The problem has been fixed by closing all cached lookups
5239 in a queue runner before running a delivery process.
5241 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5242 seem to have the "const" qualifier which it has on other OS. I've
5245 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5246 *of the same type* (client or server) with the same public name that an
5247 error should be diagnosed.
5249 9. When Exim looked up a host name for an IP address, but failed to find the
5250 original IP address when looking up the host name (a safety check), it
5251 output the message "<ip address> does not match any IP for NULL", which was
5252 confusing, to say the least. The bug was that the host name should have
5253 appeared instead of "NULL".
5255 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5256 user that is built into the binary, and the -C or -D options is used, root
5257 privilege is dropped before the configuration file is read. In addition,
5258 logging is switched to stderr instead of the normal log files. If the
5259 configuration then re-defines the Exim user, the unprivileged environment
5260 is probably not what is expected, so Exim logs a panic warning message (but
5263 However, if deliver_drop_privilege is set, the unprivileged state may well
5264 be exactly what is intended, so the warning has been cut out in that case,
5265 and Exim is allowed to try to write to its normal log files.
5271 1. smtp_return_error_details was not giving details for temporary sender
5272 or receiver verification errors.
5274 2. Diagnose a configuration error if two authenticators have the same public
5277 3. Exim used not to create the message log file for a message until the first
5278 delivery attempt. This could be confusing when incoming messages were held
5279 for policy or load reasons. The message log file is now created at the time
5280 the message is received, and an initial "Received" line is written to it.
5282 4. The automatically generated man page for command line options had a minor
5283 bug that caused no ill effects; however, a more serious problem was that
5284 the procedure for building the man page automatically didn't always
5285 operate. Consequently, release 4.20 contains an out-of-date version. This
5286 shouldn't happen again.
5288 5. When building Exim with embedded Perl support, the script that builds the
5289 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5290 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5292 6. The freeze_tell option was not being used for messages that were frozen on
5293 arrival, either by an ACL or by local_scan().
5295 7. Added the smtp_incomplete_transaction log selector.
5297 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5298 was accepting AUTH without a new EHLO.
5300 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5301 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5304 10. Logging of TCP/IP connections (when configured) now happens in the main
5305 daemon process instead of the child process, so that the TCP/IP connection
5306 count is more accurate (but it can never be perfect).
5308 11. The use of "drop" in a nested ACL was not being handled correctly in the
5309 outer ACL. Now, if condition failure induced by the nested "drop" causes
5310 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5311 or "require"), the connection is dropped.
5313 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5314 that yield "discard" can now be used with an "accept" or a "discard" verb,
5315 but an error is generated for any others (because I can't see a useful way
5316 to define what should happen).
5318 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5319 are now processed in the same way as lines in the Exim configuration file.
5320 In particular, continuation lines are supported.
5322 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5324 15. Added -ti meaning -t -i.
5326 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5327 domains, and warn by logging if others are found.
5329 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5330 returned from GIFCONF when Exim is trying to find the list of interfaces on
5331 a host. The code in os.c has been modified to copy each ifreq to an aligned
5332 structure in all cases.
5334 Also, in some cases, the returned ifreq's were being copied to a 'struct
5335 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5336 means the last couple of bytes of an IPv6 address could be chopped if the
5337 ifreq contained only a normal sockaddr (14 bytes storage).
5339 18. Named domain lists were not supported in the hosts_treat_as_local option.
5340 An entry such as +xxxx was not recognized, and was treated as a literal
5343 19. Ensure that header lines added by a DATA ACL are included in the reject log
5344 if the ACL subsequently rejects the message.
5346 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5347 MD5 (which is deprecated).
5349 21. When testing a filter file using -bf, Exim was writing a message when it
5350 took the sender from a "From " line in the message, but it was not doing so
5351 when it took $return_path from a Return-Path: header line. It now does.
5353 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5354 with a valid header line field name (a series of printing characters
5355 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5357 23. Changed "disc" in the source to "disk" to conform to the documentation and
5358 the book and for uniformity.
5360 24. Ignore Sendmail's -Ooption=value command line item.
5362 25. When execve() failed while trying to run a command in a pipe transport,
5363 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5364 could be confused with a return value of 69 from the command itself. This
5365 has been changed to 127, the value the shell returns if it is asked to run
5366 a non-existent command. The wording for the related log line suggests a
5367 non-existent command as the problem.
5369 26. If received_header_text expands to an empty string, do not add a Received:
5370 header line to the message. (Well, it adds a token one on the spool, but
5371 marks it "old" so that it doesn't get used or transmitted.)
5373 27. Installed eximstats 1.28 (addition of -nt option).
5375 28. There was no check for failure on the call to getsockname() in the daemon
5376 code. This can fail if there is a shortage of resources on the system, with
5377 ENOMEM, for example. A temporary error is now given on failure.
5379 29. Contrary to the C standard, it seems that in some environments, the
5380 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5381 program. Exim now does this explicitly; it affects the formatting of
5382 timestamps using strftime().
5384 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5385 complaints. I've now initialized all the variables, to avoid this.
5387 32. Header lines added by a system filter were not being "seen" during
5388 transport-time rewrites.
5390 33. The info_callback() function passed to OpenSSL is set up with type void
5391 (*)(SSL *, int, int), as described somewhere. However, when calling the
5392 function (actually a macro) that sets it up, the type void(*)() is
5393 expected. I've put in a cast to prevent warnings from picky compilers.
5395 34. If a DNS black list lookup found a CNAME record, but there were no A
5396 records associated with the domain it pointed at, Exim crashed.
5398 35. If a DNS black list lookup returned more than one A record, Exim ignored
5399 all but the first. It now scans all returned addresses if a particular IP
5400 value is being sought. In this situation, the contents of the
5401 $dnslist_value variable are a list of all the addresses, separated by a
5404 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5405 to accept a host name and all its aliases if the forward lookup for any of
5406 them yielded the IP address of the incoming connection. Now it accepts only
5407 those names whose forward lookup yields the correct IP address. Any other
5408 names are discarded. This closes a loophole whereby a rogue DNS
5409 administrator could create reverse DNS records to break through a
5410 wildcarded host restriction in an ACL.
5412 37. If a user filter or a system filter that ran in a subprocess used any of
5413 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5414 the wrong values were passed to the pipe command ($thisaddress had the
5415 value of $0, $0 had the value of $1, etc). This bug was introduced by
5416 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5418 38. Improved the line breaking for long SMTP error messages from ACLs.
5419 Previously, if there was no break point between 40 and 75 characters, Exim
5420 left the rest of the message alone. Two changes have been made: (a) I've
5421 reduced the minimum length to 35 characters; (b) if it can't find a break
5422 point between 35 and 75 characters, it looks ahead and uses the first one
5423 that it finds. This may give the occasional overlong line, but at least the
5424 remaining text gets split now.
5426 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5427 file descriptors might be low, and that setting 1000 would always raise it.
5428 It turns out that in some environments, the limit is already over 1000 and
5429 that lowering it causes trouble. So now Exim takes care not to decrease it.
5431 40. When delivering a message, the value of $return_path is set to $sender_
5432 address at the start of routing (routers may change the value). By an
5433 oversight, this default was not being set up when an address was tested by
5434 -bt or -bv, which affected the outcome if any router or filter referred to
5437 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5438 log only when "message" or "log_message" are set. However, if one of the
5439 conditions was an address verification, or a call to a nested ACL, the
5440 messages generated by the underlying test were being passed through. This
5441 no longer happens. The underlying message is available in $acl_verify_
5442 message for both "message" and "log_message" expansions, so it can be
5443 passed through if needed.
5445 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5446 new expansion $bh_ to give the encoded byte string without charset
5447 translation. Translation happens only if iconv() is available; HAVE_ICONV
5448 indicates this at build time. HEADERS_CHARSET gives the charset to
5449 translate to; headers_charset can change it in the configuration, and
5450 "headers charset" can change it in an individual filter file.
5452 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5453 that creates RFC 2047 encoded "words" labels them as that charset instead
5454 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5455 expansion operator; (ii) when Exim creates a From: line for a local
5456 message; (iii) when a header line is rewritten to include a "phrase" part.
5458 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5459 buggy, causing it to skip the first lines of messages whose message ID
5460 ended in 'D'. This would not have bitten before Exim release 4.14, because
5461 message IDs were unlikely to end in 'D' before then. The effect was to have
5462 incorrect size information for certain domains.
5464 45. #include "config.h" was missing at the start of the crypt16.c module. This
5465 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5468 46. If there was a timeout during a "random" callout check, Exim treated it as
5469 a failure of the random address, and carried on sending RSET and the real
5470 address. If the delay was just some slowness somewhere, the response to the
5471 original RCPT would be taken as a response to RSET and so on, causing
5472 mayhem of various kinds.
5474 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5475 when I implemented it. It didn't allow for the fact that some option values
5476 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5477 the right test for positive values.
5479 48. Domain names in DNS records are case-independent. Exim always looks them up
5480 in lower case. Some resolvers return domain names in exactly the case they
5481 appear in the zone file, that is, they may contain uppercase letters. Not
5482 all resolvers do this - some return always lower case. Exim was treating a
5483 change of case by a resolver as a change of domain, similar to a widening
5484 of a domain abbreviation. This triggered its re-routing code and so it was
5485 trying to route what was effectively the same domain again. This normally
5486 caused routing to fail (because the router wouldn't handle the domain
5487 twice). Now Exim checks for this case specially, and just changes the
5488 casing of the domain that it ultimately uses when it transmits the message
5491 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5494 50. If a filter generated a file delivery with a non-absolute name (possible if
5495 no home directory exists for the router), the forbid_file option was not
5498 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5499 the existing equality matching.
5501 52. Exim was using ints instead of ino_t variables in some places where it was
5502 dealing with inode numbers.
5504 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5505 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5506 called TMPDIR, and if it finds it is different, it changes its value.
5508 54. The smtp_printf() function is now made available to local_scan() so
5509 additional output lines can be written before returning. There is also an
5510 smtp_fflush() function to enable the detection of a dropped connection.
5511 The variables smtp_input and smtp_batched_input are exported to
5514 55. Changed the default runtime configuration: the message "Unknown user"
5515 has been removed from the ACL, and instead placed on the localuser router,
5516 using the cannot_route_message feature. This means that any verification
5517 failures that generate their own messages won't get overridden. Similarly,
5518 the "Unrouteable address" message that was in the ACL for unverifiable
5519 relay addresses has also been removed.
5521 56. Added hosts_avoid_esmtp to the smtp transport.
5523 57. The exicyclog script was not checking for the esoteric option
5524 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5525 will work only if exicyclog is run under the appropriate euid.
5527 58. Following a discussion on the list, the rules by which Exim recognises line
5528 endings on incoming messages have been changed. The -dropcr and drop_cr
5529 options are now no-ops, retained only for backwards compatibility. The
5530 following line terminators are recognized: LF CRLF CR. However, special
5531 processing applies to CR:
5533 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5534 nor a local message in the state where . is a terminator.
5536 (ii) If a bare CR is encountered in a header line, an extra space is added
5537 after the line terminator so as not to end the header. The reasoning
5538 behind this is that bare CRs in header lines are most likely either
5539 to be mistakes, or people trying to play silly games.
5541 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5542 was being incorrectly given as 18 bytes larger than it should have been.
5543 This is a VOB (very old bug).
5545 60. This may never have affected anything current, but just in case it has:
5546 When the local host is found other than at the start of a list of hosts,
5547 the local host, those with the same MX, and any that follow, are discarded.
5548 When the list in question was part of a longer list of hosts, the following
5549 hosts (not currently being processed) were also being discarded. This no
5550 longer happens. I'm not sure if this situation could ever has previously
5553 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5556 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5557 header. This is something that is recommended in a new Internet Draft, and
5558 is something that is documented as being done by Sendmail. There are two
5559 possible values. For messages generated by the autoreply transport, Exim
5562 Auto-Submitted: auto-replied
5564 whereas for all other generated messages (e.g. bounces) it adds
5566 Auto-Submitted: auto-generated
5568 63. The "personal" condition in filters now includes a test for the
5569 Auto-Submitted: header. If it contains the string "auto-" the message it
5570 not considered personal.
5572 64. Added rcpt_include_affixes as a generic transport option.
5574 65. Added queue_only_override (default true).
5576 66. Added the syslog_duplication option.
5578 67. If what should have been the first header line of a message consisted of
5579 a space followed by a colon, Exim was mis-interpreting it as a header line.
5580 It isn't of course - it is syntactically invalid and should therefore be
5581 treated as the start of the message body. The misbehaviour could have
5582 caused a number of strange effects, including loss of data in subsequent
5583 header lines, and spool format errors.
5585 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5586 client host had authenticated. This control can now be exercised by an ACL
5587 for more flexibility.
5589 69. By default, callouts do not happen when testing with -bh. There is now a
5590 variant, -bhc, which does actually run the callout code, including
5591 consulting and updating the callout cache.
5593 70. Added support for saslauthd authentication, courtesy of Alexander
5596 71. If statvfs() failed on the spool or log directories while checking their
5597 size for availability, Exim confusingly gave the error "space shortage".
5598 Furthermore, in debugging mode it crashed with a floating point exception.
5599 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5600 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5601 serious problem, Exim now writes to the main and panic logs when this
5602 happens, with details of the failure. It then refuses to accept the
5603 incoming message, giving the message "spool directory problem" or "log
5604 directory problem" with a 421 code for SMTP messages.
5606 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5607 0, 1, and 2 exist, because some OS complain for execs without them (see
5608 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5609 descriptors. However, the code omitted to check that the open succeeded,
5610 causing mysterious errors if for some reason the permissions on /dev/null
5611 got screwed. Now Exim writes a message to the main and panic logs, and
5612 bombs out if it can't open /dev/null.
5614 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5615 interact so that it is all more flexible. It is supposed to remain
5616 backwards compatible. Also added extra_local_interfaces.
5618 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5619 to bomb out with an assertion failure - to the client this appears as a
5620 connection drop. This problem occurs in the part of the code that was taken
5621 from the Samba project. Fortunately, the assertion is in a very simple
5622 function, so I have fixed this by reproducing the function inline in the
5623 one place where it is called, and arranging for authentication to fail
5624 instead of killing the process with assert().
5626 75. The SPA client code was not working when the server requested OEM rather
5627 than Unicode encoding.
5629 76. Added code to make require_files with a specific uid setting more usable in
5630 the case where statting the file as root fails - usually a non-root-mounted
5631 NFS file system. When this happens and the failure is EACCES, Exim now
5632 forks a subprocess and does the per-uid checking as the relevant uid.
5634 77. Added process_log_path.
5636 78. If log_file_path was not explicitly set, a setting of check_log_space or
5637 check_log_inodes was ignored.
5639 79. If a space check for the spool or log partitions fails, the incident is now
5640 logged. Of course, in the latter case the data may get lost...
5642 80. Added the %p formatting code to string_format() so that it can be used to
5643 print addresses in debug_print(). Adjusted all the address printing in the
5644 debugging in store.c to use %p rather than %d.
5646 81. There was a concern that a line of code in smtp_in.c could overflow a
5647 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5648 initially expressed, the concern was not well-founded, because trailing
5649 spaces are removed early. However, if the trailing spaces were followed by
5650 a NULL, they did not get removed, so the overflow was possible. Two fixes
5653 (a) I re-wrote the offending code in a cleaner fashion.
5654 (b) If an incoming SMTP command contains a NULL character, it is rejected
5657 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5658 runs initgroups(), so that if the Exim user is in any additional groups,
5659 they will be used during message reception.
5665 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob